The Practical Real-Time Enterprise: Facts and Perspectives

The Practical Real-Time Enterprise Facts and Perspectives Bernd Kuhlin ´ Heinz Thielmann Editors The Practical Real-...

Author: Bernd Kuhlin | Heinz Thielmann

43 downloads 1884 Views 3MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

The Practical Real-Time Enterprise Facts and Perspectives

Bernd Kuhlin ´ Heinz Thielmann Editors

The Practical Real-Time Enterprise Facts and Perspectives With 127 Figures

12

Bernd Kuhlin Siemens AG Siemens Communications Enterprise Systems Com ESY Hofmannstraûe 51 81359 Mçnchen Germany [email protected] Professor Dr.-Ing. Heinz Thielmann Fraunhofer Institute for Secure Information Technology (SIT) Rheinstraûe 75 64295 Darmstadt Germany [email protected]

All proceeds from this book are donated to the Siemens UNICEF initiative ªWe help childrenº. A central part of this project is the education campaign of UNICEF in Afghanistan ± thus being an important investment in the future of this country, enabling more than three million children to attend school and get further education and teaching material.

Cataloging-in-Publication Data Library of Congress Control Number: 2004118033

ISBN 3-540-21995-1 Springer Berlin Heidelberg New York This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer-Verlag. Violations are liable for prosecution under the German Copyright Law. Springer is a part of Springer Science+Business Media springeronline.com ° Springer Berlin ´ Heidelberg 2005 Printed in Germany The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Hardcover-Design: design & production, Heidelberg SPIN 11007814

42/3130-5 4 3 2 1 0 ± Printed on acid-free paper

V

Acknowledgement The irrefutable trend towards real-time communication within companies – as well as within the realms of global networking and unlimited availability – does not merely change the working environment of our information society alone. This modern form of communication is also exerting an increasing influence on our everyday lives. In future it will push back the distinct boundaries between professional and private life, smoothing the transition. Particularly because of these social consequences and the exceptional technical complexity, experts from a great many disciplines are currently devoting themselves to this subject. And the subject – as befits an interdisciplinary discussion in the best sense of the words – is also quite controversial. The aim of this book of contemporary articles on real-time communication in enterprises is to provide all interested parties with an overview of the global scenario and to graphically portray the various positions. The multifaceted and informative whole is the sum of the individual parts. These range from the preparation to the realization and use of real-time-based communication solutions, culminating in the perspective of the near and distant future. Here broad and pioneering indications of the way ahead are described and proposed. At the same time, special value has been placed on the scientific basis and practical implementation. It would not have been possible to successfully examine these aspects if the authors from the many different companies, working environments and technical areas had not spontaneously and willingly donated their contributions to this book. As a result, we offer our cordial thanks to the authors who, in addition to all their vocational obligations, have volunteered in such an exemplary way to direct their knowledge and their insights of real-time communication into the current chapters of this book. They, together the numerous helpers in the background, have brought this book into being and made the subject exhaustively available to a wider audience for the first time. In particular we would like to thank Peter Kasprzyk who, as chief editor, never lost sight of the broader picture or his love for detail. The successful completion of this work is especially due to his untiring efforts. We would also like to thank Dr. Michael Meyer for his advice regarding the content and his supportive coordination within the circle of authors. Munich, January 2005

Bernd Kuhlin

Prof. Heinz Thielmann

VII

Foreword

Speed as a factor for success

Our modern industrial society lives life in the fast lane. The catchwords “faster”, “shorter”, “more powerful” reflect what we experience in almost all aspects of our lives. Whether at home or at work, we are constantly on the move and in a rush. In our private lives we find rapid exchange of information most entertaining and we are fascinated by the wide range of information that pours in on us from all around the world, mainly via the new media. It gives us the feeling of being a part of the action everywhere and all the time. Seldom are we aware that the only reason this flood of information, often referred to as “overstimulation”, does not lead to overkill is that we manage to organize our time effectively. There are many parallels to this in the business world. Here too, a great deal of time pressure is exerted from outside; goals are set ever higher and deadlines become tighter. In other words, demands on our time demand faster reaction. Crucial information travels around the globe – across all time zones – in a matter of seconds. In fact, instead of CET or CEST, it would make sense to have a single time zone for the worldwide network called GST for Global Simultaneous Time. In business more so than in private life, we are almost constantly online. We try to process information as it comes in. But real-time processing is often hampered by technical problems or a lack of chronological coordination (i.e. different time zones) with our business partners. The real-time enterprise – really makes it on time!

Of course, the changes in how we deal with time and information have completely different implications for the business environment than for our private lives. We can't just switch off during office hours. This means designing business processes that enable the fastest possible exchange of information, a requirement that demands innovative communication concepts that optimally exploit existing technical methods of data exchange. Our companies turn into true real-time businesses, i.e. into Real-Time Enterprises. Process-oriented employees, effective time management and innovative communication technology are the catalysts that let companies achieve consistently high levels of efficiency. Hence, optimum efficiency

VIII Foreword

does not inevitably demand a consolidation of performance that is widely regarded as stressful or disagreeable – on the contrary! Identifying challenges as opportunities

It is a fact that today we have to meet the challenges for which we ourselves are largely responsible due to ever more efficient communication technology. Companies can increasingly separate their sense of location from purely geographical considerations and instead base it more on their business needs. This results in virtual locations and our management task lies in effectively integrating this virtuality into the business world in real-time. Our competitors around the world are moving along the same lines. In this era of globalization, another essentially new demand is for related events to occur around the world at almost the same time, independent of where the parties involved may actually be. To achieve the required synchronicity, the technological aspect of communication must be given greater prominence. Globalization is accompanied by social adjustments that are felt to be painful and hard at first – something that we are experiencing first hand at the moment. Serious consideration reveals, however, that there is no acceptable alternative. Only a company that can affirm itself globally over the long-term will be recognized as a business leader. Successful companies bring prosperity to their societies or preserve prosperity already there. However, this triggers a Best Practice competition in which societies themselves will have to compare favorably on a global scale. Therefore, the only national economies that in the long run will be able to sustain their position as business leaders are those that are proactively involved in globalization and that regard real-time not as a social threat but as an opportunity – as an opportunity in globally changing times that will serve as a reference model to all other national economies in the ceaseless benchmarking of their cultures. Real-time enterprise – a real chance for the future

Our society is at the beginning of an evolutionary process whose outcome is still wide open. At present, the only certainty is that major changes will be necessary if we hope to maintain our global position. The fields on which companies will have to focus their activities have been identified and are described in detail in this book by renowned representatives from the fields of industry, consultation and science according to subject. The key contribution of the editors lies in the careful compilation of select individ-

Foreword

IX

ual expert opinions to present an overall insight that provides the reader with significant food for thought while promoting a motivational spirit of optimism.

Thomas Ganswindt

XI

Content

The Real-Time Enterprise Facts, practices and prospects..................................................................... 1 I. Business processes in modern companies......................................... 11 Dieter Spath, Thomas Renner, Anette Weisbecker

Inter-company business processes and e-collaboration ....................... 13 Arnold Picot, Thomas Hess

Business process management in real-time companies ....................... 29 Michael Meyer

Real value in real time.......................................................................... 45 Tom Davies, Barry Gilbert, Jeff Swartz

Competitive response: a new lens for evaluating company performance .......................................................................... 57 II. Typical examples from industries..................................................... 71 Susan J. Unger

Building the Real-Time Enterprise at DaimlerChrysler ...................... 73 Bernd Voigt

Real-time-oriented service architectures in the airline sector.............. 81 Claus E. Heinrich, Alexander Zeier

The Extended Enterprise – Economic integration in real-time ........................................................................................... 95 Michael-W. Hartmann

Strategic perspectives for the hotel industry ...................................... 113

XII

Content

III. Technological networking .............................................................. 123 Ralf Steinmetz, Manuel Görtz, Harald Müller

Technical networking.......................................................................... 125 Bernd Kuhlin

Designing more productive business processes with convergent networks................................................................... 145 Jörg Eberspächer, Josef Glasmann

QoS Architectures and Resource Management in the Intranet .......... 165 IV. Technical process support .............................................................. 191 Shai Agassi

Business process evolution in real-time ............................................. 193 Friedrich Kopitsch

I&C Technologies for a Real-Time Enterprise (RTE)........................ 201 V. Enterprise security ........................................................................... 237 Claudia Eckert

Secure mobile business solutions for real-time enterprises ............... 239 Walter Fumy, Jörg Sauerbrey

Identity & Access Management Faster ROI and improved security through efficient assignment of rights and access control............................................. 259 Martin Schallbruch

Real-time business requires security, trust and availability ............... 275 VI. Outsourcing..................................................................................... 285 Clemens Jochum

Intelligent IT sourcing in the financial industry: background, preconditions and requirements of future IT organization design ............................................................. 287

Content XIII Jürgen Frischmuth, Christian Oecking

Outsourcing as a strategic management decision .............................. 297 Johann Csecsinovits

Less costs, more functionality............................................................ 311 VII. Challenges facing qualification and management..................... 319 Dieter Spath, Rita Nøstdal, Martina Göhring

Visual management ............................................................................ 321 Ferri Abolhassan

From CIO to Chief Process Officer ................................................... 333 Rob House

How to be switched on – without being switched off........................ 343 VIII. The steps in evolving into an “E-Enterprise”........................... 349 Steve Blood

The RTE: it starts with early warnings .............................................. 351 Wolfgang König, Tim Weitzel

Towards the E-Enterprise: standards, networks and co-operation strategies ................................ 359 Andreas Pfeifer

The evolution to real-time architecture .............................................. 385 Ralf Reichwald, Christof M. Stotko, Frank T. Piller

Distributed mini-factory networks as a form of real-time enterprise: concept, flexibility potential and case studies.................. 403 Frank E. Gillett, Thomas Mendel

Organic IT: cut IT costs, speed up business ...................................... 435 Jörg Luther

Living and working in a global network ............................................ 453 Andy Mattes

Delivering the promise. Making it pervasive ..................................... 467

XIV Content

IX. The visions....................................................................................... 479 Claus Weyrich

Knowledge-based companies – objectives and requirements ............ 481 Hans-Jörg Bullinger, Rolf Ilg

Living and working in a networked world: ten trends ............................................................................................ 497 Andreas Neef, Klaus Burmeister

Swarm organization – a new paradigm for the E-enterprise of the future ................................................................... 509 Juniors Group

Young professionals look to the communication of tomorrow .......... 519 Georg Berner

Information and communication in 20XX ......................................... 531 List of authors......................................................................................... 545

The Real-Time Enterprise Facts, practices and prospects

Introduction Hans was furious. He had been waiting for the bid for a full two hours. He had explained all the plans to John – his Australian business partner – in detail on the whiteboard in the virtual conference. And the understanding had been that it would be taken care of in half an hour. A few online invitations to tender. Then check on the capacity of the plant in China. Doublecheck the spot price for uncut South African diamonds. Calculate everything and send it to the Writing Wizard. That was one new item he was particularly proud of. The autonomic AI software processed the information you input to create complete bid documents, including translating them into the language of the party requesting the bid – and it was done. Quickly. After all, customers weren't known for their patience. But now this delay. If the analysts got wind of it, the bottom would fall out of the share price. To heck with just-in-REAL-time! So better not even think about e-voting at the daily shareholders' meeting after market close. He would probably be out of a job before sundown. Just last week, a major industrial group – previously a very successful one – had involuntarily demonstrated what could happen in such a case. It had worked on a bid for a new hospital complex in Nairobi for two interminable days. The outcome was inevitable: the Managing Board was replaced immediately, but investor and customer confidence was so badly shaken that many business partners pulled out anyway. The consequence: the in-and-out business transactions so vital for liquidity dropped to almost zero within three days. That was the end: bankruptcy! At least the liquidation went quickly. In just five hours, all assets were transferred to the creditors by the Virtual Receiver and Liquidator. And that for a group with annual sales of 350 billion euros! It took your breath away. Hans thought back to the first years of the 21st century. Just as the mobile phone boom was ending and right in the middle of the

2

The Real-Time Enterprise

communications market disaster. People had gotten used to day-long delays, even if they just wanted to call someone. That's the way things were then. Even insurance companies with the latest e-business systems sometimes needed 24 hours and more to generate an offer. And then another two days to fully process a new policy. Car buyers – what an anachronism, even then – accepted delivery times of several months. What a joke – but somehow comforting, too. Whatever. Hans didn't have time to let his thoughts run wild about “real-time antiquity.” Not now and in such a precarious situation. But he wouldn't be where he was today if he were the type to let any little problem knock him out of the race. His strategy was already set. Agent. Share purchases. Good distribution. Majority in the shareholders' meeting. Job secure. Inquiry to the online forum – clarification of the bid. Win a little time. Bring in two alternatives – Mumu in Kenya and Ralph from Ontario. Super quick. Idea: the cause of the delay lies with John, whose network infrastructure provider is on strike. The need for speed

SPEED, SPEED, SPEED: the senseless and ill-informed mantra of yesterday's e-business hype or the brutal reality of an economy based on the global division of labor? We, the authors of this book, are firmly convinced that it is the latter. If today's technology and its development trends are extrapolated into the future, even cautiously and conservatively, we soon arrive at scenarios like the one described above, in which speed and quick response times are fundamental values and financial necessities of entrepreneurial behavior. We are still at the dawn of this development, but there has already been a lasting cultural change, as is illustrated by the two examples below. Just a few years ago, the generally recognized and accepted time for handling business correspondence was one to two weeks. Today, companies are publicly flamed in the Internet if they do not reply to e-mails within a working day. Customers experiencing such delays often defect right away, and they tell their friends – and anyone else who reads their postings. Another example is the sharp swings in the share prices of individual companies. If a leading, well-established company adjusts its sales forecast for the next quarter upwards by one percentage point, say, you often see jumps of 5-7% in its stock price. That can boost the company's market capitalization by several billion euros. The relationship between the two – 1% more sales and several billion euros of additional capitalization – has no basis in economic terms, but it happens nonetheless. We can smile or swear about these trends, but at the end of the day, they

The Real-Time Enterprise

3

are facts that have to be faced, and the challenges resulting from them have to be met. This book is intended to help you do just that. It is conceived as a guide, as a collection of ideas and proven concepts for tuning companies to enable greater speed and increase their competitiveness, make them more attractive to customers and thus ensure that they operate more efficiently. The real-time enterprise

THE “REAL-TIME ENTERPRISE” – just what is it? An initial definition seems useful at this point, since an innovative term can soon become blurred if not founded on a clear-cut concept. The basic idea is simple: to become quicker. But quicker in what way? In our view, a real-time enterprise has three main abilities: 1. Internal and external data is integrated quickly and in real-time in a wellorganized company data pool. Examples of this are customer feedback on products, sales figures from the branch network, or inventory levels for spare parts on customer premises. 2. Analyses of information in the company data pool can be obtained in real time, across function boundaries and at the touch of a button. Seen in this light, quarterly reports of sales figures for individual products, for example, are no longer in keeping with the times and may even represent a dangerous information lag, depending on the industry and commodity. 3. The number of working steps performed in batch mode is shifting dramatically in favor of immediate completion in real time. The issue of communications – or Real-Time Communications – plays a special role here. Studies have shown that processing times sometimes double when necessary communication events are handled in batch mode in the business process – e.g. e-mail or messages on answering machines – and not in real time. In other words, when an activity cannot be completed and lies around for days because an urgently needed partner cannot be contacted. This situation must be clearly distinguished from one involving an excessive workload, which may likewise prevent the immediate processing of individual business transactions. The necessity of acquiring these three abilities has implications for the processrelated, technical and organizational aspects of a business that are dealt with in detail in this book.

4

The Real-Time Enterprise

Content overview TRANSITION TO A REAL-TIME ENTERPRISE: A COMPLEX AND LONG-TERM DEVELOPMENT. The transition to a real time enterprise is neither simple nor immediate. It involves long-term changes, especially in internal and cross-company work processes, in internal and cross-company division of labor, in the technologies deployed in the evolution to a real time enterprise and in the management of the process, and in the way the enterprise itself is managed. The nine chapters of this book are made up of contributions from innovative practitioners, leading thinkers from academia and research as well as service and technology providers. They offer a wealth of ideas and the knowledge vital in effectively meeting the challenges of the “need for speed” and fully leveraging the potential it offers. I. Business processes in the modern enterprise

A global footprint, the international division of labor and changes in customer and consumer expectations are putting enormous pressure on companies to change the way they do business. The coordination and collaboration of specialized participants along the value chain and across company and national boundaries, increasingly volatile and well-informed customers with decreasing brand loyalty, coupled with the accelerating pace of change and new technological possibilities form a quartet of challenges for which answers are urgently needed. All this is also overlaid by management's continuing need to keep the company's results in line with the capital market's rising expectations regarding results and to optimize processes to increase productivity and leverage employed assets to the best possible effect. Of course, this situation does not affect stock corporations alone. Government agencies, for example, and facilities in the healthcare sector are under considerable pressure to cut costs and are looking for ways to provide their services at far lower cost. This is resulting in a tightening spiral in which the entire economic system is growing leaner, shedding excess ballast and increasing in pace, a situation that poses a very complex task even for the traditionally flexible small and medium-sized sector and selfemployed professionals. II. Typical examples from industry

Globalization, increasing division of labor and the pressure to cut costs in

The Real-Time Enterprise

5

business processes are not really new trends. However, innovation cycles are becoming faster and faster, especially in information and communications technology. This means that the process optimization “toolbox” is filled with more and more new possibilities at a lightning pace. Typical examples chosen from various industry segments demonstrate proven approaches and concepts for using the elements of this toolbox profitably for business processes, enabling “real time business” to take shape thanks to quicker handling. This is seen particularly clearly in the case of financial services providers at the customer interface and in the capital markets, in the automotive industry in supply management and development, in the chemical and pharmaceutical sectors in development and logistics, in retailing in recording and analyzing actual customer behavior and in the logistics chain, and finally in transport in planning, maintenance processes and traffic control. However, these industries still face major challenges regarding further productivity increases, as will also be shown. III. Technological networking

Enormous sums of money have been invested to automate business processes in the past decades, and the boosts in speed and productivity in a wide range of industries have been impressive. However, the annual rates of increase in speed and productivity are falling, while the financial efforts that enterprises need to make to achieve these goals are growing. What are the reasons behind this? More extensive process automation would mean increased networking of technology and processes – both among the already optimized individual functional areas within an enterprise and with external partners on the supplier and purchaser side. However, this often does not become a reality because the cost of integration at the infrastructure level (due to the heterogeneous nature of the network landscapes and installed device base) is so high as to be financially untenable. Still, the Internet has brought with it new possibilities and new energy in this context in recent years. In addition, the increasing transformation of voice communications from the technical point of view to an IT service (Voice over IP, voice/data convergence) is another driver of lower costs for the further networking needed for process automation. Finally, the greater use of standards in hardware and software products and in communication protocols and the publication of interfaces form an innovative basis for making networking affordable again.

6

The Real-Time Enterprise

IV. Technical process support

The possibility of a shared, open cross-company network is in turn the foundation for an end-to-end application landscape, allowing processes to be extensively automated and sped up across functional areas and enterprise boundaries. Workflow applications, for example in order processing, procurement, human resources and the finance department, have been a standard part of many enterprises for years. However, the potential represented by the data resources thus created remains untapped in most cases. All too often, it is still humans who – with painstaking detail to attention and their analytical skills – use this pool of data to create analyses that can be used as a basis for decision-making. That is no longer in keeping with the times, given the increasing pressure to accelerate the speed at which business is handled and strategic decisions made. A first important step here is to “format” the data. Content management systems create uniform information structures throughout the enterprise, allowing data to be evaluated with the aid of analytically enabled software modules. Many CRM systems and help desk applications are already equipped with such features. The consolidation of communications landscapes across all media and applications is another area with great potential for improving processes. Today's communications landscape is very fragmented. There is a wide range of media and applications – e-mail, instant messaging, SMS, voice in the wireline, wireless and utility radio networks, separate voice messaging systems, ERP messaging systems – that are not all networked with each other and are usually divorced from the work process as well. If a communications event – clarification of a question, a decision, approval or the like – arises during a rigidly defined work process (e.g. invoicing), it encounters the fragmented communications landscape outlined above, one that does not ensure that the communications partner can be reached quickly. (People have to phone around to contact someone or e-mails are not read because recipients are not at their desk.) The result is that the carefully planned business process plunges into an undefined and uncontrolled pending status. The innovative technological concept of Real-Time Communications promises a solution to this problem, since it ensures consolidation of the communications landscape as well as full integration of communications functions in applications that support business processes. The first practical applications in the areas of logistics, production, development and sales prove the benefits of Real-Time Communications, yet also demonstrate considerable potential for further improvements, in particular in terms of time, quality and the costs of business processes, especially through greater integration of communications.

The Real-Time Enterprise

7

V. Enterprise security

Security in enterprises is growing more and more complex. First, the number of potential targets for attack has increased sharply in recent years, since more information is stored in enterprises' electronic data pools. At the same time, the number of channels of attack is growing as a result of increased networking both within companies and with their external business partners. Finally, the number of attackers and the tools available to them, as repeatedly documented by media reports, represent a growing challenge for security officers. In this respect, it is not only necessary to consider the potential damage that can be inflicted on your own company, but also the liability for damages incurred by external partners as a result of gaps in your security system. Security should therefore be a management issue, one in which end-to-end concepts based on security policies are implemented in a focused way. The software and media industries have been faced for years with the additional challenge of piracy, a crime that has increased particularly as a result of the Internet and now causes serious financial loss running in the billions. VI. Outsourcing

Moving tasks that are not part of an enterprise's core competence outside the company, to external partners, has great potential for making internal business processes leaner, cutting costs and enabling greater flexibility in response to market fluctuations. This possibility has been greatly exploited in past years. However, more and more outsourcing projects have been reversed in the very recent past. Studies even claim that 70% of all outsourcing projects can be regarded as failures. One key reason may be that increased division of labor with external partners increases the complexity of the overall coordination of the enterprise and the network surrounding it. The problems that arise within an enterprise when activities are coordinated among the various functional areas (e.g. sales and production planning) can soon multiply if a large number of external partners are added. However, outsourcing should nevertheless not be rejected out of hand. Those projects that have been successful from the point of view of users and service providers prove that it is possible to achieve results from outsourcing in line with expectations. Moreover, innovative platforms and, above all, Real Time Communications offer the technological basis for successfully mastering the coordination of value added networks, even when many partners are involved.

8

The Real-Time Enterprise

VII. Challenges facing qualification and management

The trend toward the real time enterprise is accompanied by increased automation and accelerated business and decision-making processes. Despite all the automation, however, there are still many points where human action is needed. Since the automated part takes virtually no time whatsoever, the duration of a work process is determined largely by these points of human action. As a result, they are extremely important for the speed of an enterprise, but also increasingly complex. While the data entry person used to be the main interface between the environment and the machine, it is now (and will be increasingly so in future) the specialist with broad-ranging training who has to cut to the core of very complex matters and make the right decisions in a very short time. This trend will even result in completely new job profiles. At the same time, management will also be under strong pressure to change. Speed, even in wide-reaching decisions, is a must, and that means saying farewell to committees and management by consensus. Hierarchies must be kept very transparent and open. Points where decisions are unnecessarily filtered should be omitted or eliminated. In addition, the role and intrinsic value of occupations in the information and communications industry, or in the information and communications units of an enterprise, will increase sharply, since such persons form the backbone of the digitized real time enterprise and are thus a strategic success factor. VIII. The steps in evolving into an “E-Enterprise”

A company cannot turn into an e-enterprise or a real time enterprise overnight. First of all, the necessary requirements must be met on the infrastructure side by means of open systems and network architectures. However, this not only relates to infrastructures at your own organization, but also those of external infrastructure service providers – especially to wireless carriers, since this is where the separation from the enterprise's IT and applications is greatest. On this basis, open application platforms and process support tools can be introduced step-by-step. It is then possible to engineer real time processes in the second step, entailing consequences for the corporate culture and workforce qualification. Standardization activities, on both the user and the provider side, are of particular importance as flanking measures, since they ensure the financially vital basis for acceptable integration and process automation costs. This also covers the increasing number of communications devices that can be used to access elec-

The Real-Time Enterprise

9

tronic processes and the enterprise's data pool. In particular, growing mobility is spawning more and more types of access devices: mobile scanners, mobile phone/PC hybrids, networked sensors in process monitoring, merchandise tags with communication abilities, etc. This brings with it the potential danger of a slide into the same heterogeneous and fragmented situation as exists today, with all the associated problems.

IX. The visions

Living and working in real time or at top speed in a networked world will bring about a radical change at all levels of society in the coming decades. If today's technologies and their foreseeable developments are extrapolated into the future, even conservatively, we soon arrive at scenarios like the one outlined at the start of this chapter. It then becomes clear what implications this will have for organizational structures and processes in enterprises and the public sector. The increasing penetration of technology, coupled with global networking, will entail a wide range of new challenges. One example that gives a taste of things to come can be seen in the recent and widely publicized major power failures in a number of countries. If at some point the worldwide data network forms the backbone of all social, political and economic activity, even minor malfunctions would have serious consequences.

Successful Enterprises. The future holds a great deal of promise for businesses, but the challenges will be as great as the opportunities. Those companies and other organizations that follow the road sketched in the nine chapters of this book and become real time enterprises will master those challenges. They will be the winners in the marketplace of the future.

I. Business processes in modern companies

The need to optimize business processes is on going. Information and communication technology will normally play a key role. In the case of the real-time enterprise it plays a pivotal role.

Dieter Spath, Thomas Renner, Anette Weisbecker

Inter-company business processes and e-collaboration

Real-time companies can only be realized if inter-company processes are standardized, supported by suitable IT technologies and integrated in the business' value network. The tendency in the past of focusing primarily on internal processes is insufficient. This article describes major characteristics of inter-company business processes in the context of real-time companies and presents criteria for evaluating and selecting processes that are to be integrated. The contribution that inter-company e-business solutions, business integration technologies and standards can make toward achieving “real-time capability” is shown. Finally, an example of a procedure for developing a company into a real-time business is presented at the end of the article.

Global corporate cooperation Companies in competition

The demands placed on companies today are constantly on the rise. To be a successful player in global competition, new products must be developed rapidly and manufactured efficiently and according to customer-specific requirements. They must be sold globally in a clearly defined target market. In some industries such as in semiconductor memory devices, the supplier with the greatest profit will be the one who is first on the market with a new, more powerful product. In some sectors, it is impossible for a company to make a profit if it is too late in bringing products to market. At the same time, we have experienced considerable changes in the market in the last few years. While the years 1997 to the beginning of 2000 were still characterized by a belief in almost unlimited growth – especially in the “New Economy” – companies in a great number of sectors experienced a sharp drop in sales thereafter. Companies that are slow in recog-

14

Dieter Spath, Thomas Renner, Anette Weisbecker

nizing changes in the market and in the competition and who are therefore poorly prepared to react as a “breathing” business, run a high risk of disappearing from the market entirely. Reducing costs through acceleration

Speed is a major key to success in many sectors. In addition to the primarily “strategic” aspect of speed already described, there is also a direct link between speed and costs. Companies that accelerate their processes scrutinize each individual process step. Processes will be rapid only if, firstly, all non-vital process steps are eliminated and, secondly, processes are automated to the greatest extent possible – especially through IT support. When processes are simplified and automated with a view to accelerating them, this almost always leads to an automatic reduction in process costs. Optimizing processes across organizations and companies

Companies and organizations are well aware of the significance of methods for process optimization and of IT technologies for automation. A great deal of progress has been made in the last fifteen years. Systems for “Enterprise Resource Planning” (ERP) automate many standard processes and provide decision-makers with up-to-date information. Workflow systems control processes for recurring, clearly structured tasks. Document management systems achieve rapid process flows and immediate access to information and knowledge. Nevertheless, much potential still goes unused. Even today, many companies – for reasons that are at first understandable – only consider processes and IT systems within their own organization. They form “local optima” and thus give away potential that could otherwise be exploited through a close integration of partners – suppliers, development partners, sales partners and customers. It is now necessary to take the next step and move from “local optima” to “global optima” by optimizing the overall process, including the partial processes implemented by partners. Methods and IT solutions with which processes can be controlled and optimized across companies or organizations must be created and made available; a rapid and smooth exchange of data and information must be achieved across company boundaries; rather than focusing on just the one company, the entire “business network” must be optimized.

Inter-company business processes and e-collaboration

15

Real-time and inter-company solutions through e-business

E-business solutions are an important element of a real-time company. Implementing e-business solutions helps accelerate business processes and provide new IT applications that support inter-company processes. A survey of over 900 German companies taken for the “eBusiness Economy Barometer 2003” [Kelkar 2003] showed that 72 percent of companies using e-business were of the opinion that “e-business” has a “large” or “very large” impact on increasing process speeds. Of all the e-business advantages (Figure 1), this was considered to be the most significant effect, i.e. the greatest benefit of e-business. Company objectives (user) Cost savings Increase in sales

16% 7%

Customer requirements Opening up of new markets

28%

20%

36%

24%

Increase in speed Increase in flexibility

36%

25% 48%

12%

31%

39% 19%

Quality improvement of 2% 11% products very large

26% 30%

22%

large

13% 7% 2% 13%

5%

11%

4%

29% 40%

medium

7%

19%

39%

20% 9%

14%

13% 24%

small

very small

n=[504;515]

Figure 1: Impact of E-Business on company objectives

Some available e-business solutions – such as e-procurement systems, online tendering and auctions, partner portals, online shops and collaborative development platforms – offer solutions for specific areas of intercompany process optimization. However, they also point to the significant challenges companies face today. These include the standardization required for processes and interfaces and the “e-business capability” of business partners – topics that will be dealt with in detail further down in this article.

16

Dieter Spath, Thomas Renner, Anette Weisbecker

Rationalization and real-time in the service sector

The use of e-business solutions for rationalization and acceleration is not limited to the area of material goods or the manufacturing industry alone. In November 2003, Alan Greenspan declared that growth of American productivity was “surprisingly strong” – in the third quarter 2003 it was five percent over the previous year's level. Experts are talking of a “return of the new economy”. The reason for this productivity growth is primarily seen in the fact that rationalization has now also become possible in the service sector. Global networks and the use of e-business solutions make it possible to relocate routine tasks in the service sector to the “back office” or to partner companies in other countries. Certain process steps become wholly unnecessary with e-business solutions. Examples are the recording of receipts in the financial sector and the full data availability and processing capability on mobile terminals such as those used by courier services drivers [cf. Fischermann 2003].

Inter-company process from the perspective of integration

Characteristics

Inter-company and intra-company processes have different characteristics. Table 1 shows the main differences between three process types. The criteria were specifically selected to evaluate their suitability for automation and their real-time capability. The three process types under consideration are: • intra-company processes, • outsourced processes, • inter-company processes. The “outsourced” process type was deliberately taken up in a separate category in light of the ongoing outsourcing and insourcing discussion. It could – depending on the particular characteristics – also be viewed as a special case of an intra-company process (e. g. 100% IT subsidiary of a company) or a true inter-company process (e. g. no personal involvement and a large number of customers of the outsourcing service provider).

Inter-company business processes and e-collaboration

17

Table 1: Characteristics of inter-company processes Process type

Intra-company

Function of participant

Responsible party

Duration of business relationship

-

Outsourced

Inter-company

Service provider (e.g. Partner (e.g. supplier, for IT, purchasing, development, logistics, facility management, sales...) production Long-term

Medium-term to longterm

Contribution to added value

Core of added value; Small contribution or Varying degrees of contribution; special strategic; unique sell- none; economy of scale knowledge or economy ing points of scale

Major process characteristics

Contribution to added Service level; costs; interfaces value; costs; lead time; quality

Productivity; costs; reliability; interfaces

Process transparency

High

Medium

Low

Ability to control and modify processes

High

Medium

Low

IT support

Any; homogeneous company systems

Frequently used inter- Frequently used interfaces; heterogeneous faces; heterogeneous systems systems

Implementability Can be easily imple- Can be implemented Problematic (heterogeneous requirements and of general IT-supmented systems) ported workflow Standardization of processes

Individual processes can be defined as required

Standardization of interfaces

Interfaces can be defined as required

E-business capability

Can be defined by company

Process standards required

Process standards required

Cross-company inter- Cross-company interface standards face standards required required Can be influenced

Strongly depends on capabilities of partner

18

Dieter Spath, Thomas Renner, Anette Weisbecker

The goal is fast and efficient process implementation for all process types. An important factor in acceleration and cost reduction is far-reaching IT support and automation, in addition to the development of a process design that is as simple as possible. This in turn requires standardized processes, integrated information systems, consitent master data and common semantics [cf. Alt 2004]. All of these requirements are far easier to meet for intra-company processes than for inter-company processes, for two basic reasons: the heterogeneity of IT systems and the heterogeneity of the requirements of the various business partners with regard to processes and data interfaces (including the exact significance of individual data fields). The heterogeneity of processes and data interfaces in the inter-company area must be solved by means of cross-company standardization. It is the primary factor in the success of inter-company process automation and acceleration and is thus a necessary prerequisite for the real-time company. It will be described in greater detail further on this chapter. The second problem area – the heterogeneity of IT systems in different companies – will continue to be an issue in the foreseeable future. Although large companies appear to be limiting their ERP software to only a few software providers, the problem of varying system versions and company-specific system configurations and system extensions remains. The solution for this problem area is called Business Integration Software and will be discussed in a separate section. Selection criteria for integration

Not least because of the challenges noted above, inter-company integration of processes and IT systems is today still in its infancy. The significance of inter-company integration depends on various criteria that characterize the processes. Integration and the associated automation should be set as a goal primarily if one or more of the following criteria are met: • large number of business processes; • high frequency of repetition of processes per business partner, i.e. a large number of processes with a manageable number of partners; • large volume of data or large documents; • high cost of individual processes, e. g. from manual data entry; • high timing requirements, i.e. time-critical process; • manageable complexity of interfaces and processes.

Inter-company business processes and e-collaboration

19

When implementing the inter-company integration on the basis of these selection criteria, it is wise to begin with low complexity processes. Furthermore, it is important to consider the “e-business capability” of the partner to be integrated – not a few integration projects have failed because the partner’s existing IT infrastructure and software prevented implementation of the project. The strength of the position of the partner to be integrated also plays an important role. For example, if a large exclusive supplier lacks the willingness to integrate, successful implementation of the project is likewise endangered.

Beyond acceleration – Cost advantages from e-business solutions

Examples with an impact and with benefits

Which e-business solutions are relevant to the real-time company? Examples of e-business application solutions are described below for various business areas – purchasing, sales, logistics, service and finance. In addition, there is a multitude of individual solutions that would exceed the scope of this article. Therefore, examples have been selected that primarily illustrate inter-company processes in the sense of “e-collaboration”. The following applications will be examined: • • • • • • •

E-procurement, electronic tendering and online auctions in purchasing, Electronic marketplaces, Sales and service portals, Collaborative product development, Supply chain management and logistics associations, E-learning, Business intelligence solutions.

A short, detailed description of these applications will be presented in the following section. What effect do these applications have within a company or on the company's products with regard to time, quality and costs? The applications have already been selected on the premise that the respective solution must accelerate processes. In what areas does acceleration take place and is the assumption true that acceleration frequently leads to cost reductions or that

20

Dieter Spath, Thomas Renner, Anette Weisbecker

the solution under consideration has several positive effects? Table 2 shows the effects of the implementation of e-business solutions. The individual sub-criteria into which the main criteria were broken down were selected for the following reasons: the acceleration of decision-making processes is an important criterion for real-time companies, and the speed of the development process is an important criterion for a company to be able to compete successfully and an indicator of its innovative capability. Table 2: Effect and benefits of selected e-business solutions

Customer relationship (satisfaction)

Company products

Processes

Decisions

Quality Company products

Procured goods

Warehouse

Processes

Costs

Other processes

Development processes

Time

Decision-making processes

Benefits

E-procurement Electronic tendering Online auctions Electronic marketplaces Portals Collaborative product development Supply chain mgmt. Logistics networks E-learning Business intelligence Considerable positive effect

Medium positive effect

Small/no positive effect

The costs of warehouse stock are listed separately from process costs because they benefit from a reduction in working capital requirements. The costs of procured goods act directly on the income statement and indirectly on the manufacturing costs of company products – another important factor in competition. In addition to the quality of the decision-making process (again, in direct relation to the real-time company), the effect on quality was also evaluated

Inter-company business processes and e-collaboration

21

with respect to process quality, the quality of company products and the potential for “indirectly” increasing customer satisfaction – e. g. through improved quality of service or faster and simpler access to information. In addition, a higher product quality also “directly” promotes customer satisfaction. The evaluation shown in Table 2 is based on project experience at Fraunhofer IAO. The respective benefit does not always set in immediately, but may require an additional intermediate step. For example, customer problems can be detected very early on using “Business Intelligence”, but an additional step is necessary to eliminate them. In the table, it is assumed that this step has taken place. In addition to process acceleration, a positive effect on costs is observable in almost all cases. At the same time, the quality of processes and decisions improves. The “magic triangle” – time, quality and costs – is severed. An improvement of one parameter usually does not take place at the expense of another target value; rather, all three parameters can in many cases be improved simultaneously. Detailed, quantitative benefit calculations must be performed individually on the basis of company-specific characteristics. Although many companies have already started and partially implemented individual activities, it can still be said that the potential for acceleration, rationalization and cost reduction is far from having been fully exploited. This is also so because the new solutions should not merely “electrify” existing processes, but processes should be reworked and simplified during the implementation phase – a step that is often omitted due to time restrictions or (inappropriate) attempts at cost saving. As a result, the benefit of the new solutions falls far short of what is actually possible. Solution examples in detail

The e-business application solutions evaluated above will now be described briefly to make the individual evaluations easier to understand and to provide readers less familiar with this subject matter with additional information. e-procurement with “Desktop Purchasing”: Companies procure indirect goods via their own online catalogues in the Intranet, in the marketplace or from suppliers. The catalogues contain company-specific final prices and the product spectrum released for the company. They are based on previously negotiated framework agreements. The entire procurement process is controlled and supported electronically. Orders are transferred to the sup-

22

Dieter Spath, Thomas Renner, Anette Weisbecker

plier electronically and are frequently delivered within 24 hours. Procurement times are cut dramatically and stock is reduced. Electronic tendering: The tendering of requirements by electronic means accelerates the tendering process, increases the number of possible bidders and permits seamless, fully electronic processing and storage. An automated comparison of bids for individual lots and an electronic archive system take advantage of additional rationalization potential. In the public sector, offers become legally effective by means of electronic signatures. Encoding and special functions ensure adherence to regulations governing awarding of a contract and directives such as VOL and VOB (general terms and conditions applicable to delivery and construction contracts in Germany). Online auctions: “Reverse auctions” are the most widespread. A clearly defined requirement is electronically “negotiated” with a number of suppliers within a short period. This means that the suppliers essentially make price offers that are shown to all competitors anonymously. This price pressure ensures low procurement costs. At the same time, the negotiation process is highly transparent and is completed within a short period of time. Electronic marketplaces: Typically, marketplaces connect suppliers and buyers or development partners. They are a platform for providing functionalities for several companies or for cross-company cooperation. They facilitate integration for n:m relationships – for example, enabling several suppliers to be electronically connected to several customers. In some cases – such as with marketplaces shared by buying companies – they ensure standardized data processing and cross-supplier data access, thus assisting in creating market transparency. Sales and service portals: Portals makes accessible a series of functionalities for sales partners, customers and company branch offices at other locations. For example, current product and price information is made available, order data are collected and checked for further processing, the current order status is brought in and service offers are provided. Electronic product and spare parts data are provided in a standard format, allowing them to be adopted directly into the partner's engineering or procurement system for subsequent processing (see also Gurzki, 2003). Collaborative product development: Companies use shared platforms for rapid product development or production optimization. Online platforms offer shared project rooms, conversion services and discussion forums. Video conferencing and wideband Internet services enable direct communication or accelerated, joint development across country borders according to the “follow-the-sun” principle. Supply chain management: Electronic control of supply chains is already

Inter-company business processes and e-collaboration

23

highly advanced in certain sectors, such as in the automotive industry. It is impossible to realize just-in-time or just-in-sequence concepts without IT support. However, these concepts have by far not been introduced in all sectors. Moreover, there is still much unused potential in the area of multistage supply chain control in many branches, including in the automotive sector. Electronically supported logistics networks: Where logistics services are provided entirely by a single company, electronic support is usually available and status queries can be placed. However, as soon as several logistics partners work together in the chain or if special subcontractors are integrated, exact electronic tracking of consignments becomes a problem today. The concept of online freight exchanges has only been successful in certain limited areas. It is weakened, for example, by the fact that direct connection of company IT systems is highly complex due to a lack of standardization. True market transparency and full electronic process support are not yet available. E-learning: Online-supported systems for training and continuing education not only support intra-company learning processes, such as software or product training, but are also used for inter-company partner and customer training. Collaborative learning platforms offer functions for carrying out and presenting group-based learning sessions and for group viewing and annotation of learning materials. For example, E-learning considerably accelerates the process of introducing a new product on the market. Business intelligence: Software solutions consolidate and analyze data from operative systems with structured data bases and make them available for decision-making support in the form of reports – in real-time whenever possible. A higher level of usability is achieved by means of additional visualization and predefined events and alarms. Today, data are aggregated almost exclusively from internal company systems. In the future, an increase in the use of data from partner systems should be aimed for and can be expected.

Required standards and technologies

Standardization, a factor for success

A comprehensive real-time company can only be achieved with the aid of standards. If it is still possible to establish “internal company standards”, “true standardization” is essential to permit the electronic integration of

24

Dieter Spath, Thomas Renner, Anette Weisbecker

business partners. Company-specific solutions generate considerable additional effort on the part of business partners or prevent integration entirely.

Business processes

Business documents, data, structures

Classification and feature systems

Transport (E-Mail, Web, FTP, queuing system...)

Data transmission (modem, LAN, VPN, Internet; IPX, NETBEUI, TCP/IP, ...)

Figure 2: Examples of standards for inter-company business processes

With respect to inter-company business transactions, standards are necessary in the following specific areas, in addition to the standards that already exist for data transmission and transport, security and authentication (overview in Fig. 2): • Business documents and data: For the purpose of electronic data exchange, the contents of business documents must be standardized in detail to facilitate rapid and automatic additional processing. Examples of important business documents are orders, invoices and delivery notes, as well as offers, dispatch notifications, logistics data, etc. Standards offering solutions include EDI-FACT, ANSI X.12, openTRANS, xCBL, cXML, CIDX and others. Many of these standards are based on XML (Extensible Markup Language). XML itself provides the “language” but not a clearly defined “vocabulary” (semantics), which is also a prerequisite for communication. • Inter-company business processes: It is insufficient for electronic business transactions to merely describe the documents. Additional basic rules must be defined for how a business partner is to act upon receiving a document, for example, if a receipt of delivery is sufficient or if an order confirmation with an exact date of delivery is expected. By defining several consecutive process steps, for example, the overall standard-

Inter-company business processes and e-collaboration

25

ized process of a complete order transaction is generated. Standards for inter-company business processes include RosettaNet, ebXML and Biztalk. Some of these standards also specify individual processes themselves. • Classification: In the area of catalog-based product and service data that are exchanged between companies, standardized classifications are used for grouping products. Feature systems supplement the classification and enable a standardized description of product features. Examples of classification systems are eCl@ss, UNSPSC and ETIM. Many standards for business documents, processes and classifications have been developed in the last years and are now being extended and optimized. Figure 3 shows the prevalence of e-business standards in Germany in early 2003. The survey was made as part of the “eBusiness Economy Barometer” [Kelkar 2003]. User EDIFACT

52%

cXML

16%

BMEcat Datanorm, ELDANORM ebXML

19% 12%

eCI@ss openTRANS UNSPSC RosettaNet PRICAT

16%

9%

3% 2%

ETIM

3% 4%

proficl@ss

1% 3%

OAGIS

1% 2%

47%

22%

53% 53%

28%

61%

34%

6% 4%

xCBL

43% 47%

33%

5% 5% 3% 2%

27%

34%

9%

62%

27%

63%

31%

64%

28%

65%

28%

68%

28%

Currently in use

11% 34%

38%

12%

5%

25%

32%

11% 3%

7%

11%

17%

70% Planned

Not planned

Standard unknown

n=[377;447]

Figure 3: Prevalence of E-business standards in Germany, early 2003

26

Dieter Spath, Thomas Renner, Anette Weisbecker

Business integration software

The standardization of a technical exchange format for documents and the definition of processes is insufficient to create a real-time company. An additional step is still required: the direct integration of IT systems, i.e. the use of IT solutions to implement the defined processes. A number of standard software products are now available to facilitate this task. With the “Business Integration” software or “Enterprise Application Integration (EAI)” software, which is more strongly oriented toward intra-company processes, customized developments and adaptations can be largely avoided. Typical functions of these products are: data transformation, process management, authentication and encoding (security) and the provision of interfaces to standard products and messaging systems. Basic functions, selection criteria and a market overview are available in [Lebender 2003]. In addition to the “Business Integration” software described above in a narrower sense, there are additional software products that support crosscompany cooperation and system integration. These include portal software, groupware, workflow systems, application sharing, video conferencing and online project planning tools. These solutions focus more on human-human or human-machine communication. On the road to the real-time company

The previous sections showed that real-time companies are efficient companies. They do not become real-time for the sake of becoming real-time, but rather to make faster and better decisions, accelerate innovation, produce at lower cost and to maneuver on the market and compete with greater agility. Real-time companies look beyond intra-company processes and generate innovative solutions for the inter-company environment. They orient themselves to their strategic objectives, to customer requirements and to changes in the marketplace. e-business solutions are important building blocks in the support of inter-company processes. They bring about acceleration, cost reductions and quality improvement. The goal of becoming a real-time company makes a significant contribution towards securing and augmenting a company's success. The “realtime” objective is pursued within clearly defined limits in specifically selected areas. How then can a cost-effective, dynamic, real-time company be created?

Inter-company business processes and e-collaboration

27

What are the most important areas and what is their potential? How should a company proceed? Based on project experience at the Fraunhofer IAO, a procedural model was developed whose basic characteristics are shown in Figure 4. 2a

4a

Process determination and evaluation in specific areas – internal and external aspects

Detailed conceptual design and implementation of solution element

1

3

4b

Identification of realtime requirements and potential – strategy and market aspects

Basic conceptual design Technology platform(s) Implementation plan

Detailed conceptual design and implementation of solution element

2b Tools and methods for exploiting potential – mirroring on focal areas

4c Detailed conceptual design and implementation of solution element

Figure 4: Procedural model for transforming a business into a real-time company

In the first phase, the primary goals and guidelines in creating a real-time company are developed and the areas and processes to be included are defined. This is accomplished per the company strategy and market characteristics (customer, supplier, competitor). In the second phase, the individual areas and processes are first examined using a portfolio methodology. A benefit assessment is performed that leads to an initial prioritization. Tools and methods for exploiting the potential in the defined focal regions are examined and evaluated. Then a cost/benefit analysis is performed and a final prioritization is carried out. After the initial release by company management, the basic conceptual design is created and, if required, the basic technology platform(s) is selected and the implementation plan is laid out using the methods of change management. After another release, the individual measures and solution

28

Dieter Spath, Thomas Renner, Anette Weisbecker

elements are implemented in stages on the basis of clearly defined milestones. The implementation of measures is monitored for success on the basis of characteristic values that are established prior to implementation. It is well worth drafting a company-specific project to bring about acceleration in the sense of a real-time company. Given this new view of their processes, companies are able to recognize problems and challenges that up to now have remained hidden. The new and improved solutions they generate not only extend far beyond acceleration, but may also have a stronger impact on other target areas – solutions and effects that in the end generate a competitive advantage and measurably increase a company's success. Bibliography Alt, Rainer; Österle, Hubert: Real-time Business – Lösungen, Bausteine und Potenziale des Business Networking. Springer-Verlag, Berlin Heidelberg, 2004. Fischermann, Thomas: Aufschwung paradox. In: Die Zeit 31.12.2003, p. 22, Zeitverlag, Hamburg. Gurzki, Thorsten; Özcan, Nergis: Kunden-, Lieferanten- und Mitarbeiterportale in der betrieblichen Praxis. Fraunhofer IRB Verlag, Stuttgart, 2003. Kelkar, Oliver; Renner, Thomas et. al.: eBusiness-Konjunkturbarometer – Investitionschancen für die deutsche Wirtschaft 2003/2004. In: eBusiness Konjunkturbarometer 2003, p. 6 ff. Wegweiser-Verlag, Berlin, 2003. Lebender, Markus; Ondrusch, Nicole et. al.: Business Integration Software – Werkzeuge, Anbieter, Lösungen. Fraunhofer IRB Verlag, Stuttgart, 2003. Scheer, August Wilhelm; Abolhassan, Ferri et. al. (eds.): Real-Time Enterprise – Mit beschleunigten Managementprozessen Zeit und Kosten sparen. SpringerVerlag, Berlin Heidelberg, 2003.

Arnold Picot, Thomas Hess

Business process management in real-time companies

For many years the acceleration of processes through the improved provision of information in science and in practice has been intensively discussed and put to the test. The vision of a “real-time company” as has been sketched out in the last few months has intensified these considerations even further. It is a tantalizing vision: A reduction in transitions between media of the most varied types ought to significantly accelerate business and management processes. The goal is a company that, for example, takes into account minor revisions desired by the customer shortly before starting production, is able to determine the price for a service on the basis of information of previous customer purchasing behavior, or reacts immediately to short-term peaks in demand. Long gone will be the days where a manager makes incorrect disposition decisions because he only knows the sales figures from the previous week. Similarly, it should no longer be possible for the product management of an airline company to set too low a price for the remaining contingent of a flight solely because the latest booking figures are not known. Whether such a real-time company can be realized depends to a critical extent on the efficiency of the technologies available for data integration (see Holten 2002). For example, whether the customer has to make a telephone call or can access integrated supplier data directly from his web browser has a substantial effect on processing speeds. Only if the trade chain immediately forwards their purchasing data to the manufacturer can the expeditor make real-time decisions. A significant step was reached in the 90s with the proliferation of integrated databases and the resulting establishment of integrated commercial application systems. However, it had already become plain by the mid 90s that it would be possible to open up the systems to business partners. Up to now, this opening has been implemented in two further development steps: initially, the focus was placed on selected suppliers and customers, and then, in recent times, it became increasingly flexible towards the most diverse business partners in

30

Arnold Picot, Thomas Hess

the form of networks. It was, however, also clear at the beginning of the 90s that the approach of treating information and communication technologies and the associated application systems as isolated entities frequently leads to “electrification” of existing operational solutions (see Picot/Nippa 1995; Koch/Hess 2003). Many practical examples had shown that only through the modification of intra-company and inter-company business processes could the potential for new application systems be exploited. In the 90s wide discussions were held on the subject of business process management. Today, the approach of focusing on business processes has become standard, at least from a practical point of view. For this reason, although the subsequent article orients itself on the three technological development steps outlined above, the central focus is placed not on the technologies themselves, but on their potential for the design of business processes. It is important to start by saying that this involves successive stages of overlapping levels of development rather than consecutive and self-contained phases.

Step 1: Internal business processes It was Michael Hammer and Thomas Davenport who, in 1990, pointed out the potential for improvement of business processes, especially through new information and communication technologies. Hammer and Davenport and many subsequent authors made the observation that, until that point in time, companies had a plethora of data stocks in a great variety of application systems that were mostly dedicated to particular functions. The execution of business processes often required data to be exchanged between these systems via complicated interfaces, a procedure that frequently took a long time, was cost-intensive and was sometimes even impossible as the result of conversion and consistency problems. The objective was to keep these distributed and uncoordinated data in a few integrated databases or, in the extreme case, even in only one database. Frequently such integrated databases formed part of standard application systems that integrated functions as well as data. Thus, data and functions did not have to be repeatedly implemented. For purposes of analysis, operational data were increasingly stored separately from dispositive data in so-called data warehouses. However, the center of interest was always the processes within a company which – for the first time – were defined and designed independently of the boundaries of company operations. This included standardizable processes in both the product and customer-oriented sectors (such as order processing

Business process management in real-time companies

31

in an industrial firm or claim settlement in an insurance company) as well as processes in the support sector (such as in product development or payroll accounting). Not infrequently, significant improvements were achieved and, for the first time, time and quality were systematically used as a measure of success in addition to costs. Occasionally, systems based on key data were implemented for continuous process improvement. In contrast, a process-oriented alignment of the company's organizational structure, as was frequently demanded in the heyday of this first step, takes place relatively rarely in practice. Still, it is not unusual to encounter the establishment of cross-sector process responsibility as well as a corresponding controlling process. The redesign of internal processes is part of today's standard repertoire of operational practice. Meanwhile, methodical know-how is available in many companies and can also be found in many educational books for organizational theory (see Koch/Hess 2003). Currently new requirements have arisen through the development of so-called Smart Devices. Smart Devices are physical devices that are equipped with a microchip capable of communication. So equipped, these devices can forward status messages without changing to another medium and, consequently, the status message of a pallet is available at any location as soon as it has been created. Interesting potential is also promised by the systematic evaluation of data that automatically arise from E-Commerce bidding or are available at service providers of every type. For example, a marketing campaign could be much more efficiently steered if the reaction of customers in a test cluster were known. In practice, a large number of examples can be found for the successful reorganization of internal processes through the use of modern information and communication technologies. Thus, as early as the beginning of the 80s, the American car manufacturer Ford was able to decrease the number of employees in the billing department by 75% through the use of an integrated database, thereby achieving a clear reduction in the number of errors (see Hammer 1990). Before this reorganization, the billing department at Ford was mostly occupied with the investigation of discrepancies between purchasing department paperwork, goods received and supplier invoices. Use of an integrated database that could be accessed by all departments participating in the goods purchasing process dramatically reduced these costly checking tasks (see Figure 5). The purchasing department saves all necessary goods order information in the database. Moreover, delivered goods are only accepted by the goods inwards department if there is concurrence between the goods delivered

32

Arnold Picot, Thomas Hess

and open orders in the database. The application system introduced at Ford then issues the resulting supplier checks. After reorganization, the billing department is now only responsible for sending these checks to the supplier. The reorganized ordering process is shown in Figure 6. This could be rationalized even further by, for example, introducing a credit note process that fully foregoes the generation and checking of separate goods invoices and is only balanced on the basis of order and shipping documents. Ford Purchasing department Order

Copy ? Invoice

Supplier

Billing department ? Confirmation

?

Goods inwards

Goods

Payment

Figure 5: Ordering process before the reorganization at Ford1

A second example of efficiently organized internal processes is demonstrated by Home Shopping Europe (HSE). HSE decided to improve its processes by using continuous feedback. Via its own television channel, this shopping broadcaster offers its customers the opportunity of convenient purchasing at home, independent of shop opening hours and overcrowded shopping stores. An assortment of products is presented on television and can be ordered by the customers by telephone. In 2002, around 1.5 million customers ordered products from HSE in this way, generating the company a net turnover of 288.2 million Euro. HSE uses 'up to the minute' information on customer behavior to make short-term decisions on the presentation of products from their assortment. Using live control, HSE evaluates a variety of different information sources, such as warehousing, order volumes and call center volumes. This information is used to make According to Picot/Dietl/Franck 2002, p. 308.

1

Business process management in real-time companies

33

decisions directly during the broadcast on how products currently on display will continue to be presented. For example, in addition to adjusting the length of a product presentation, HSE can refer to current customer behavior to decide for a short-term interruption of a presentation to prevent bottlenecks in the order process. Purchasing department Order

Entry in database

Incoming goods department

Billing department

Acceptance of goods

Check sending

Comparison of the delivery with database

Sending of the check

Integrated database

Figure 6: Ordering process after the reorganization at Ford2

Step 2: Inter-company business processes If customer and supplier business processes were at all considered in step 1, they were regarded as quintessentially unchangeable and thus regarded as a given. Occasionally a business partner was given access to the application systems of the company; today this might be via a simple web browser. However, the focus was always placed on tapping the potential for improvement in the company. Only after the implementation of step 2 are business processes that extend beyond company boundaries systematically considered. The central starting point for the improvement of inter-company processes in step 2 is again data integration, i.e. improving the information supply for individual process steps and cutting back redundant and perhaps inconsistent data stocks. Only in exceptional cases is it still possible to support this integration with the establishment of an integrated inter-company dataAccording to Picot/Dietl/Franck 2002, p. 309.

2

34

Arnold Picot, Thomas Hess

base. This approach can be found time and again in supply chain systems, which are currently being tested with varying degrees of success. A loose form of data integration is usually more practical. A simple approach is the use of specific interfaces between the databases of participating companies. Crucial for these interfaces are suitable application systems for the transformation of data, i.e. that convert the memory format of company A into the memory format of company B. Among other things, specifications on the basis of the Extensible Markup Language (XML) are available for this purpose. The triggers of a transformation (a particular point in time, an event, etc.) have yet to be clarified. Manufacturers of standard commercial application systems have satisfied this trend and have arranged appropriate interfaces. Incidentally it must be noted that it has actually only been possible to realize highly integrated forms of comprehensive processes in their unadulterated form in rather small centrally-managed companies. If a company has a separate center or consists of legally independent units, this usually means that the data stocks of these units cannot be held in an integrated company-wide database. By definition, an inter-company business process comprises two or more independent companies that work together but follow independent goals. Consequently, process design becomes a matter of negotiation whose outcome depends heavily on the negotiating power of the participants. A good example of this is the supplier network of the automotive industry, where for a long time now manufacturers have not only drawn up clear guidelines for the interfaces, but have also specified details of how processes are to be configured at the suppliers. Due to the specificity of the investments made, a coordinated process design often leads to an actual reinforcement of existing power relations. Occasionally, drawbacks that are no longer acceptable to the supplier are counterbalanced in part by adjustment payments, or so-called 'side payments'. A loose form of data integration in inter-company cooperations can, for example, be effected through Electronic Data Interchange (EDI) (see, e.g. Kilian, Picot et al. 1994). Karstadt Warenhaus AG has used this standard for communication with its suppliers since 1984 (see Buxmann 2001, p. 69). With 60,000 employees (in 2002), a turnover of 6.3 billion Euros, 198 department stores and 31 sporting goods shops, Karstadt is the largest department store concern in Europe. EDI is used by Karstadt for exchanging information with its business partners. Between 60 and 70 suppliers used this standard in the first ten years following its introduction. Since then, however, the number of suppliers using EDI has grown to between 500 and 600 companies. EDI generates annual costs of about 80,000 Euros

Business process management in real-time companies

35

through its use. This is in addition to one-off costs for its introduction in the form of internal personnel costs for 24 man months, costs for hardware and software to the amount of around 75,000 Euros and costs for external services to the amount of 25,000 Euros. These expenses are offset, however, by considerable savings. The assumption is that about six Euros can be saved per invoice. With a billing revenue of 3.5 million in 1997, the standardized IT communication support results in savings of 21 million Euros. Savings can also be generated in procurement (to the amount of 5.4 million Euros in 1998) and in logistics (to between 5 and 10 million Euros in 1998). In addition, a time saving of five days is assumed if an order is placed for out of stock warehouse articles via EDI (see Figure 7). Without EDI: ordering takes several days

Karstadt

Supplier

With EDI: ordering takes a few seconds

EDI standard Karstadt

Supplier

Figure 7: Ordering process with and without EDI at Karstadt

Effective utilization of integrated application systems in inter-company cooperations can also be illustrated by the introduction of an integrated application system for the establishment of a 'real-time supply chain' at Lucent Technologies (see Trebilcock 2003). Lucent Technologies develops and markets systems, software and services for communication networks of different service providers. The company with 35,000 employees generated a turnover of 8.5 billion US$ for fiscal year 2003. With the introduction of a 'real-time supply chain', the company pursued three major objectives. Primarily, the system was to create a cross-company network of suppliers, warehouses, logisticians and customers. Thus, an essential requirement of the system was that it had to provide facilities to quickly inform Lucent Technologies and companies downstream in the supply chain of any vari-

36

Arnold Picot, Thomas Hess

ances in deliveries to permit them to react promptly. Furthermore, the application system was to enable a connection between suppliers and customers that was simple and uncomplicated. By making it possible to track the delivery process in real-time and to connect downstream companies, the system guarantees that schedules can be readily coordinated in the event of discrepancies in delivery times. The results of the new system are conclusive. Delivery reliability increased, leading to a significantly higher level of customer satisfaction. A further plus was that the 200 warehouses for North America alone could be reduced to a mere 15. Also, instead of the original 1,700 logistics service technicians, only two were required after the system was introduced, which meant the costs for logistics alone could be reduced by around 20%.

Step 3: Business process networks In the descriptions above, emphasis was on possible and potential designs for intra-company and inter-company business processes. It is logical to connect the design possibilities already outlined in a third step. This results in so-called business process networks, which can be perceived as flexible, task-related networks of internal and external units or modules. The starting point is the overall value added process or business process, comprising all activities required for the creation of products and services. In the first and second steps, it was explicitly or implicitly assumed that internal and external processes were separate, and the focus was on the design of intra-company and/or inter-company processes. This separation will now be abolished. The objective now is the uniform design and optimization of value added and cross-company business processes in the sense of a general infrastructure for flexible process design, where no difference will initially be made between internal and external business processes. This step of integration is made possible by information and communication technology infrastructures which, due to their capability of exchanging information, data and documents in real-time, allow similar forms of labor division for inter-company and intra-company processes. Thus, required resources and expertise are more crucial for handling subprocesses than location, existing capacities, legal-economic boundaries, etc. Under certain conditions, it ultimately makes no difference from which location or which company internal company data and documents are accessed. A precondition for the optimization of comprehensive business processes is differentiation into individual self-contained subprocesses that can be

Business process management in real-time companies

37

transacted by an organizational unit, i.e. individual locations, departments, companies. In this context reference is frequently made to modules, i.e. self-contained integrated units that are autonomously and wholly responsible for the execution of a differentiated (sub) process (see Picot et al. 2003). These modules can comprise the named organizational units as well as teams, project groups, manufacturing or marketing islands, or other forms of segmentation. Several questions arise regarding the design and optimization of integral business processes: 1. Is the individual subprocess executed internally or externally? 2. How is the internal or external execution actually carried out? 3. At which location is the subprocess executed? First of all, it is important to clarify whether an individual subprocess is to be executed internally or externally to the company. The transaction cost theory and the approach of core competencies suggest that the subprocesses that are executed within a company should be those that are specific to the company and that are strategically significant or that reflect the core competencies of the company (see Picot et al. 2003). The potentials and design possibilities outlined in point 2 of this article can thus be achieved. In contrast, if specificity and strategic importance are low, such as in the implementation of complementary competencies, close cooperation with other companies is advisable, e.g. in conjunction with supply chain management as mentioned in point 3. If the subprocesses are standardized, processing should take place entirely via electronic markets. Examples of such company structures can be found primarily in the automotive and textile industries. For example, PUMA concentrates on the core competencies 'Design' and 'Sales' and outsources all remaining subprocesses to cooperation partners or various market partners. Thanks to information and communication support infrastructures, production data such as cutting instructions, cutting templates or transport orders can be transferred in real-time to the respective partner companies. The question of whether processes should be executed within or outside a company does not yet reveal anything about the concrete organizational design of these subprocesses. There are a multitude of design options in this regard. The internal company design begs the question, for example, of whether the decentralized execution of subprocesses in modules makes sense for a customer-related process. For reasons of efficiency, it is often advisable to execute certain subprocesses internally or externally in higher,

38

Arnold Picot, Thomas Hess

specialized functional areas. These are then involved in the task implementation on a case-by-case basis (see Picot et al. 2003) using the information and communication support infrastructure. An example of this is a clerk in an insurance company who, in processing a claim for damages, involves internal company specialist departments such as the legal department, assessors office, etc. On the basis of appropriate IT and communication support networks – in particular the Intranet or Extranet – these departments can be integrated in near real-time during processing and the necessary information can be promptly called up and worked in. Because this eliminates potentially costly waiting time due to inquiries, processing times can be reduced and recognized attributes of bundling such as specialization, learning effects and, most notably, 'Economies of Scale' can be realized. Various options are available for the design of inter-company cooperations. These range from pure, flexible cooperation relationships – such as are observed in virtual companies – up to long-term, close cooperation relationships such as occur in project groups, in network organizations such as Benetton or in the aforementioned supply chains (see e.g. Franz 2003) and that are eligible for the different tasks. The same is true for market-related transaction dealings, which range from the short-term flexible use of electronic markets and the regular use of existing electronic markets or purchasing platforms to the electronic connection to an electronic market such as Covisint (see e.g. Krauch 2002). The latter instance ultimately involves a type of tight networking. However, the design of business process networks as the third step in business process optimization is not limited to design options that address whether transactions are external or internal or precisely how the form of processing deemed to be practical is to be organizationally designed; the design can also include location considerations, not least on the basis of the available information and communication support alternatives. This will ultimately contribute to further optimization of the business process. If information exchange and communication at several different locations is as unproblematic and free of media transitions as it is at a fixed location, it will be possible to include specialist resources and expertise existing at different locations in task implementation on a task-by-task basis, and internal modules and company units can be outsourced to external locations. An example of this is found in the company called '24translate' which provides high quality technical translations at unrivalled speed in 27 languages and 32 specialist areas. The idea behind this enterprise is that companies should concentrate on their core competencies and leave to 24translate the

Business process management in real-time companies

39

translation of documents into all required languages and for all target markets. Translation is coordinated by only 12 employees in the head office who oversee 900 editors and freelancer technical translators around the world. Each of these editors or technical translators can be characterized as a module in the sense described above. Each module is specialized in specific languages or technical sectors, and can carry out the required translation completely and independently. In order to shorten translations delivery times, efforts are constantly invested in new XML-based technical transmission solutions. This is complemented by the deployment of a Web Services module that provides the complete service portfolio from translation to linguistics, editorial work and cultural adaptation, and that can be integrated in every Content Management system, every Intranet, every platform and all applications. A particularly explicit interpretation of the principle of flexible, realtime oriented networking of value added structures is exhibited by the Internet sector. Companies such as Ebay, Yahoo!, Lycos, MSN, Freenet, Amazon and many more offer their customers a barely manageable quantity of information, purchasing and service offers that can be individually selected and combined on the company's own portal. These can normally be accessed online and in “real-time”. They are connected with the portal provider on the basis of Internet technology, standardized integrated processes for information, contracting, payment and transaction, and bilateral contracts with supply, service and incentive agreements. In this way, these companies provide the technical organizational basis in specific fields for triggering and implementing ad-hoc value added processes in real-time – while being supported by a large network of external specialists. These examples make it clear that customer-oriented solutions can be realized in the shortest time through the flexible networking of globally acting expertise and specialists. At the same time, however, they also show that the concept can only succeed if specific preconditions are fulfilled. In a technical sense, this includes the requirements placed on the information and communication support infrastructures as the basis of a format-matched exchange of data and documents, as has already been addressed in the previous section. In addition to intra-company and inter-company process optimization as was described in the Karstadt and EDI examples in point 3, there is another important aspect that must be considered: the flexible integration of different companies between whom there may not be any long-term business relationships as in SCM (Supply Chain Management). This quickly pushes the limits of classical concepts such as EDI/EDIFACT or EDI/XML; the underlying complexity of EDI requires bilateral agreements, thereby

40

Arnold Picot, Thomas Hess

making flexible forms of application more difficult to achieve. A very promising solution may well be offered by web services – as is shown in the example of 24translate (see e.g. Picot/Breidler 2002, Löwer/Picot 2002 and Alt/Österle 2004). This involves software building blocks that provide an open communication interface that is used to call up and use specific information or services, independent of platform, program and location. If the use of web services in cross-company applications is successfully supported in the medium-term to long-term, web services will be able to provide the required flexible integration of different application systems. Web services will then be able to contribute significantly to the emergence and propagation of business process networks and real-time companies. Recently, visions of a “semantic web” have come under more intensive discussion. As a further development of the World Wide Web (WWW), a semantic web would substantially ease communication between different data stocks and systems, raising the real-time company to a much higher stage of realization (see e.g. Neumaier/Löwer/Picot 2003). It is important to note that in all cases IT investments are not insubstantial and must be endorsed by all participating companies – with their frequently quite different interests – (see Wohlgemuth/Hess 2003). In addition to the technical requirements, there is another important point that must be taken into account: The role of (process) management. The principle of linking different organizational design options which underlies business process networks primarily places new demands on management (see e.g. Picot/Reichwald/Wigand 2003, Hess 2002, Hess/Wittenberg 2003). Above all, these requirements affect: – the flexible design of business and subprocesses on the basis of underlying customer wishes and tasks to be fulfilled; – the flexible configuration of partners in the execution of the different subprocesses; – the cross-company and cross-location control of these partners. From a process management point of view, management must first grapple with the underlying business processes. Primarily this means that processes must be constantly realigned and reoriented according to customer wishes and underlying tasks on the one hand and changing and available skills and resources on the other. If, for example, individual subprocesses change as a result of new customer wishes or changing business models, suitable resources must be found and flexibly integrated via existing information and communication technology infrastructures. For exam-

Business process management in real-time companies

41

ple, if a customer of translate24 commissions a translation in a new technical field, the appropriate specialist for this technical field must be found as quickly as possible and flexibly integrated. Process management must therefore keep an open mind towards and be able to deal quickly with changed subprocesses, new forms of networking and configuration as well as changing partners. This does not only apply to virtual companies such as 24translate or the Internet portals. Thus, a connection initially limited to electronic markets such as Covisint for the procurement of standard parts can eventually develop into a longterm far-reaching cooperation, such as in the area of construction, ultimately leading to changes in existing processes and new types of partnerships.

900 specialist editors and Configuration and control by translators around the world 12 employees in the head office Internet / XML / Web Service module

Specialist translation

Virtual companies – flexible, customer-oriented and task-related configuration of processes

Customer

Figure 8: translate24 as a business process network (schematic diagram)

The continuously changing nature of processes requires more than just a flexible configuration of partners. New demands on process management also arise from cross-location and cross-company control of these partners, a task that requires decentralization, delegation and results-oriented management methods. It is here that the boundaries of information and communication technology support become apparent: while gaining confidence is essential for steering business process networks, this confidence is difficult to achieve in information and communication technology supported relationships. Hence, the significance of reputation and face-to-face communication should not be underestimated for establishing confidence in business process networks. In reference to the example of 24translate

42

Arnold Picot, Thomas Hess

described above, this means that assignment delegation and results-oriented management alone will not bring about high-speed translations. Confidence that the editors and translators will deliver the necessary quality at the required time is also required. This confidence cannot be established via the Internet or other electronic communication routes alone.

Conclusion and outlook We have attempted to describe the evolution of business process management in a real-time company in three steps. Table 3 presents the most important characteristics of all three steps in summary. Table 3: Evolution of business process management Step

Beginning

Focus

Driver

New management problems

1

Beginning of the 90s

Internal company processes

Fully integrated systems, particularly ERP systems

Cross-department perception

2

End of the 90s

Inter-company business processes

Mostly proprietary interfaces between companies, occasionally also comprehensive integrated data stock

Process design as negotiation process

3

Current

Business process networks

Standardized interfaces, flexible electronic integration of internal and external partners, web services

Flexible design and control of processes with different internal and external company partners

The creation of real-time companies requires techniques that enable fast, prompt data transmission that is free of media transitions. This article illustrates that, in addition to this, the processes within a company and between companies must be designed to ensure that the technological potential actually comes to bear and that processes can take place in real time. The use of mature technologies promises little success if, for example, market information transferred in “real time” cannot be further processed as a result of internal hierarchical structures, or if customer-related or market-oriented

Business process management in real-time companies

43

requirements cannot be implemented quickly enough due to internal coordination and authorization processes. Hence, the creation of real-time companies not only requires the use of appropriate technologies, but to an even greater extent demands that customer-oriented processes be flexibly designed in the sense of a differentiation into subprocesses, that there is an allocation to suitable internal or external company modules and that these processes be controlled. This is not, however, easy to realize if structural and procedural organizational systems already exist; on the contrary, it requires combined project and change management. This includes employees, customers and partners, and starts with individual projects that, for example, can bring about changes in the two first stages – the optimization of intra-company and inter-company processes – in order to get the company moving in the direction of a real-time company. Nevertheless, individual projects such as the optimization of internal processes through a data warehouse, the improvement of inter-company processes through EDI or web services or the introduction of supply chain management should always bear in mind the design of the entire business process (as it was described in point 4).

Bibliography Alt, R. / Österle, H. 2003: Real-time Business, Lösungen, Bausteine und Potenziale des Business Networking, Berlin. Buxmann, P. 2001: Informationsmanagement in vernetzten Unternehmen: Wirtschaftlichkeit, Organisationsänderungen und der Erfolgsfaktor Zeit, Wiesbaden. Franz, A. 2003: Management von Business Webs – Das Beispiel von Technologieplattformen für mobile Dienste, Wiesbaden. Hammer, M. 1990: Reengineering Work: Don’t Automate, Obliterate, in Harvard Business Review, Jg. 68, 4/1990, p. 104–112. Hess, T. 2002: Netzwerkcontrolling, Wiesbaden. Hess, T./ Wittenberg, S. 2003: Interne Märkte in Dienstleistungsnetzwerken, in: Bruhn, M./ Stauss, B. (eds.): Dienstleistungsnetzwerke – Dienstleistungsmanagement Jahrbuch 2003, Wiesbaden, p. 161–184. Holten, R. 2003: Integration von Informationssystemen, in: Wirtschaftsinformatik, Jg. 45, 1/2003, p. 41–52. Kilian, W. / Picot. A. / Neuburger, R. /Niggl. J. / Scholtes, K.-L. / Seiler, W. 1994: Electronic Data Interchange (EDI): Aus ökonomischer und juristischer Sicht, Baden-Baden. Koch, D. / Hess, T. 2003: Business Process Redesign als nachhaltiger Trend? Eine empirische Studie zu Aktualität, Inhalten und Gestaltung in deutschen Großun-

44

Arnold Picot, Thomas Hess

ternehmen, Arbeitspapiere des Instituts für Wirtschaftsinformatik und Neue Medien, Nr. 6/2003, München, 2003. Krauch, M. 2002: Covisint – Der elektronische Marktplatz in der Automobilindustrie, in: Picot, A. / Breidler, J. (eds.): Web-Services – Bausteine für e-business, Heidelberg. Löwer, U. / Picot, A. 2002: Web Services – Technologie-Hype oder Strategie-Faktor?, in: Information Management & Consulting, 17. Jg., 2002, vol 3, p. 20–25. Neumeier, F. / Löwer, U. / Picot, A. 2003: Das Semantic Web – Neue Perspektiven für die verteilte Wertschöpfung?, in: Information, Management & Consulting, 18. Jg., 2003, vol 3, p. 79–84. Picot, A. / Breidler, J. (eds.) 2002: Web-Services – Bausteine für e-business, Heidelberg. Picot, A. / Dietl, H. / Franck, E. 2002: Organisation: eine ökonomische Perspektive, 3rd edition, Stuttgart. Picot, A. / Franck, E. 1995: Prozeßorganisation – Eine Bewertung der neuen Ansätze aus Sicht der Organisationslehre, in Nippa, M. / Picot, A. (eds.): Management prozeßorientierter Unternehmen – Ansätze, Methoden und Fallstudien, 2nd edition, Frankfurt am Main, 1995, p. 13–38. Picot, A. / Reichwald, R. / Wigand, R. 2003: Die grenzenlose Unternehmung – Information, Organisation und Management, 5th edition, Wiesbaden. Scheer, A.-W. / Abolhassan, F. / Bosch, W. 2003: Real-Time Enterprise, Mit beschleunigten Managementprozessen Zeit und Kosten sparen, Berlin. Trebilcock, B. 2003: The real-time supply chain, in: Modern Materials Handling, Jg 58, 9/2003, p. 58–59. Wohlgemuth, O. / Hess. T. 2003: Strategische Projekte als Objekte kollektiver Investitionsentscheidungen in Unternehmensnetzwerken, in: Schreyögg, G./ Sydow, J. (eds.): Strategische Prozesse und Pfade – Managementforschung Band 13, Wiesbaden, p. 195–223.

Michael Meyer

Real value in real time

When communications technologies, applications and IT systems work together savings of between 30 and 50% can be realized. Barriers are removed via the seamless integration of real-time telephony with IT systems and business processes. The speed at which business processes take place has increased dramatically in recent years. This trend applies to all industries and to companies of all sizes. Even medium-sized organizations must be present at a multiregional or even international level and they must be able to respond to customer requests anywhere in the world. However, research conducted by Accenture and Siemens (“Real-time Communications: Scenarios for Implementing Innovative Technologies”) indicates that two thirds of the companies surveyed were unable to use the data and analysis capabilities generated by their enterprise systems to support their daily operations. In addition, less than half had implemented an information exchange system with the customers and suppliers. This means that there is a significant gap – almost a chasm – between IT and telecommunications systems and the problem is exacerbated by the speed at which automated business processes take place. When things go wrong, as they do in the real world, decisions concerning issues and events have to be taken in short time frames. A delay of a few hours can result in a significant loss of revenue, e.g. €50K if a container ship is held up for a day. Important customers can also be lost if sales and marketing issues are not resolved quickly and callers to contact centers expect immediate answers. Computer systems can automate business processes but they cannot make decisions involving issues and events: for that you need people. In recent years processes have been engineered into a series of discrete steps. Responsibility for the overall process is therefore divided around various specialists and/or specialist groups. Thus, more people need to be involved than in the past; moreover, the right people must be identified, checks made on their availability, and communications established as soon as possible.

46

Michael Meyer

And the 'right people' may involve ecosystem partners as well as teleworkers and employees who are mobile. The need to reach more people as a result of business process division is illustrated in figure 9. In addition, there is often a need to link these parties with the information that will enable an informed decision. A parts shortage, for example, might involve accessing a database so that an alternative supplier can be contacted. It may also be something more basic, like a RFID code in a logistics process that can't be read. And, to round out the challenge, mobile parties may be employing different communications devices that are operating on different types of network. Business processes under pressure

Non-localized / globalizes resources Communication/ intervention

Communication/ intervention

Business process

Continuous work division

Communication/ intervention

Demand for speed

Figure 9: These three mega-trends represent a generic challenge that is putting business processes under pressure

Real value in real time

47

In an ideal scenario, one that resembles the way we used to work, all employees and process participants would be located in one large room. They would all use the same information systems and be able to access a single, integrated, up-to-the-minute database. This situation would allow all relevant participants to assign the highest priority to the current transaction, i.e. they would not be disrupted by other tasks. The result would be a highly efficient and effective set of business processes. There are no delays in this dream-on scenario since all communications take place in real time. As soon as a new task occurs in say a sales process all responsible partners and employees are engaged and there is no need to search for specialists or wait for authorization. This is the environment that real-time communications technology seeks to replicate. However, in the real world people work in different locations and time zones and many are mobile most of the time. The challenge is therefore formidable, it might even appear to be insurmountable, but most of the requisite technology is available and the remainder is under development. In fact, the implementation of solutions that allow real-time communications functionality to become an integral part of real-time IT applications such as CRM and ERP has started. Even better is the fact that these solutions are based on industry standards. The business case for tackling formidable challenges has to be more than compelling – it should be overwhelming. There has to be a tangible return on investment; the cost and complexity of implementing new technologies must exceed the benefits. A seemingly obvious statement but the ICT industry has been long on promise and short on delivery in recent years. The real-time communications case is overwhelming because it is predicated on the fact that the IT side of mainstream business processes is close to optimization and that the only way to make significant improvements in customer service, productivity and profitability is to close the gap between people and computer systems. Adding bigger and supposedly better IT systems will not increase operational efficiency significantly. That can only come by realizing the full potential of the systems that are already in place, i.e. by bringing people into the business processes that run in real time. Until recently real-time communications and IT systems functioned in different environments; environments so different that they were almost alien. Interoperability was possible, but computer telephony integration technology was complex and therefore somewhat expensive and in most cases the integration cost outweighed the benefits. That is the historic explanation for the gap referred to earlier.

48

Michael Meyer

IT involves the processing of information (data) by computer systems. The C in ICT involves the transfer of information, usually automated, from one system to another. Thus, ICT has been a machine-centric concept. People may request the information and input content, but to date they have not been an integral part of automated business processes. When issues arise they are resolved via off-line discussions. Voice-data convergence facilitates closure, i.e. when voice is packetswitched it can share the same environment as data traffic. Convergence is therefore an important and necessary development, but network sharing should be seen as a first-generation IP (1gIP) development. There are tangible benefits, but they do not add up to an overwhelming business case. Key research findings Potential productivity gains as a result of using real-time communications are: Up to 50% for field force enablement Up to 30% for sales Up to 40% for supply chain management. These figures are the result of Accenture's own analysis, pilot projects and surveys with experts. These results also indicate that significant gains can be realized for other time-critical processes. The same Accenture-Siemens research project indicated that the key technologies and applications needed for next-generation (2gIP) developments such as the Real-Time Enterprise were presence awareness, unifiedand instant messaging, interactive voice response and mobile portals. (Presence and managed availability are covered in detail in the article titled “Delivering the promise. Making it pervasive.”) In a 2gIP environment communications and information are unified at the application and service level. Unified Communications and Information indicates functionality over and above that of convergence at the network transportation level. 2gIP is therefore the development that allows people to become an integral part of business processes. When issues arise they can now be resolved on-line, in real time, using a much richer communications medium than email or IM (instant messaging). In the 2gIP model that Siemens developed and is currently marketing (HiPath OpenScape) there is a presence-aware software suite that enables smarter ways of communicating and collaboration as well as seamless integration with mainstream IT platforms and business processes. Figure 10

Real value in real time

49

illustrates the concept. The italicization of 'as well as' indicates that this software provides a holistic solution – one that addresses productivity, profitability and customer relationship issues as well as the RTE. Web Services Real-time communications Voice Video Mobile communications Presence awareness-based, multi-modal, real-time communications Business applications CRM ERP Workforce management Supply chain management

Messaging E-mail Appointment calendar Texting/instant messaging Short message service

Cpllaboration support Conference calls Whiteboarding Document sharing

Figure 10: When real-time communications is seamlessly and deeply integrated with business processes and IT applications, information such as alerts can be sent to any client device over any IP network. [Change Web services to Application Convergence, e.g. using Web Services]

The software functions as a presence-aware module that contains everybody's contact information as well as their availability. In an ideal scenario they would be identified via a knowledge base (e.g. a skills-based directory), thereby allowing recurring events to trigger conference calls automatically. This is a groundbreaking concept since a single instance of the presenceaware module can: • Serve all employees in all locations • Boost individual and workgroup productivity • Ameliorate generic issues such as telephone tag and information overload • Leverage resources that are currently underutilized, e.g. CRM and groupware systems • And last but not least, deliver the promise of the real-time enterprise. The RTE was placed at the end of the benefit list because Siemens believes that this commendable objective should be derived from a holistic

50

Michael Meyer

solution. In other words, the RTE is not a stand-alone goal that is enabled via yet another point solution. Events that are triggered by interruptions in the business process and that need to be addressed by consensus are not different to other day-to-day issues that are best handled in real time. The hypothetical ERP communications and collaboration process of figure 11 could also represent CRM or SCM. Thus, the same software tools can be employed; to do otherwise does not make sense. "Interruptions" in partially automated processes Business process Process part

Process part

Process part

Process part

Decision regarding solution Consensus/decision requirement

Figure 11: This is a generic business process. In the case of CRM calls come in and if the agent cannot answer the query an event is triggered. The relevant parties are then contacted and the required information given to the agent, who then passes it along to the prospect

The RTE concept is important, but there is a tendency to examine time-critical business operations in isolation. Delays often occur when employees work from different locations or when they are mobile. As a result, key parties cannot be contacted or responses to email are delayed. Discontinuities in the communications flow, e.g. between different media or between people and machines are another reason for slowdowns in business processes. Thus, the solution must address these issues if it is going to be holistic. It's hard to imagine a communications landscape more fragmented than the one we currently employ. Businesses have TDM systems, wiring and phones, but are migrating to converged voice-data infrastructures employing IP systems and IP phones. LANs are getting wireless extensions and the

Real value in real time

51

same technology is being deployed in public hot spots. This allows software phones to be used in these locations. We have three wireless networks: 2G, 2.5G and 3G (four if you count hot spots). Each network has its own mobile phone; some phones are multimodal as are some PDAs. And we have wireline and DECT cordless phones in the home and some offices. This is a chaotic way of working; it's something we take for granted even though there are too many discontinuities for the RTE. Ideally all media and end-user devices that are needed for data exchange with mainstream IT systems are provided and enabled. In addition, all employees and process participants use the most suitable medium (wireless or wireline phone, IM, email, SMS) in order to ensure the highest possible level of availability. Another issue is the fact that much time is spent messaging and not enough communicating. Email inboxes fill up, making it hard to recognize information that is important, and it is easy to add attachments and to send URLs to colleagues. The overall result is information overload, which makes it harder to have the right data to hand when decisions are needed, e.g. when in a meeting. Information workers now receive 64 times more information than in 1970 and growth continues: the curve is exponential. Email has become the primary communications medium in most businesses, particularly those that operate across time zones. However, this only enables asynchronous communication and it is not suitable for time-sensitive operations. This is one example of the way that preferred communications tools do not necessarily contribute to increases in productivity. Information overload as a result of the ease with which emails can be distributed also introduces the danger that important information may not reach decision-makers on time or that the message will remain unread in the mailbox. Thus, the technology that is supposed to make us more productive can end up having the opposite effect. Instant Messaging (IM) provides a partial solution: partial because the medium is not practical when difficult decisions have to be taken and this is particularly true when several parties are involved. In these cases the phone should replace the PC. Voice is a rich medium that can express emotions and nuances, making it a much better way to clear up misunderstandings and reach decisions. Moreover, the easiest way to reach mobile employees is via the mobile phone. Voice-data convergence has allowed telephony to become data type, thereby enabling presence (free to talk or not) to be displayed via on-screen icons. In addition, application software such as HiPath OpenScape lets

52

Michael Meyer

users manage their availability via customized communications profiles. Telephony is therefore returning to its former role as the preferred way of communicating and collaborating and it is clearly the optimum medium for the RTE. There is also an important role for IM: when an event occurs the process can trigger a message and this can be used to set up a conference call automatically. A message can also be generated and sent to parties who are talking informing them that an urgent issue has arisen. So far so good, but bringing mobile workers into the equation is a challenge given the fact that different devices are employed and communications can be enabled over different wireless networks. HiPath OpenScape addresses this issue in two ways. One, from the user perspective all he/she needs to do is indicate the times when the mobile is the preferred device. This is done via the availability pull-down window. And two, middleware is used to create unified user experiences and unified communications domains. In a nutshell, this means dissolving the barriers that separate different devices and different networks. Do that and the communications landscape is no longer fragmented despite the fact that the same devices and networks are employed i.e. there is no disruption. In addition, the functionality of mobile office devices becomes nearly identical to those of wireline desktops. That has been an ICT objective for a decade or more, but now the hype is over and it is really happening. It's important to see developments such as the RTE in an enterprise-wide context. Enterprises currently have three principal domains: real-time communications (telephony); information (databases); and IT (network and other services). There is a need to merge these domains into a unified framework and this development is well advanced on the information front. This need is visualized in figure 12.

Real value in real time

Mobile Phone

PDA

PC

53

Phone

Real-time Communication Clients & Portals End-user Experiences Unified Domain

Media

Access Mode

Presence

Unified Domain

Information & Communication Resources Real-time Communications Middleware Business Applications

IT Services

Communication & Collaboration

Figure 12: This architecture elevates the functionality of existing applications and also facilitates the ability to communications-enable mainstream business processes such as CRM, ERP and SCM

Enterprise Application Integration (EAI) is taking place and enabling transactions to flow from one system to another. However, there is an equally clear need for telephony to transition in the same way. The Gartner Group has coined the term Integration Broker to describe a new generation of middleware that can help bridge the current disparate flows of data and realtime communications. Siemens employs a slightly different term – Communication Broker. Siemens' middleware has the same focus, but in addition to addressing the infrastructure of the back office it also focuses on the needs of individuals and workgroups. In addition, it features device mediation, which is used to provide the appropriate communications features to the user's device, i.e. the system recognizes the resources such as screen size. Media translation includes text to speech to speech to text. HiPath OpenScape works in conjunction with mainstream IT platforms such as Microsoft's Live Communications Server (LCS) or IBM Lotus Notes. The combination transforms a groupware application into a communications hub that can transmit all message types: email, voice, fax, IM and SMS. This allows a non-intrusive SMS to be sent to parties who would otherwise be unavailable. This facility is particularly important in the case of the RTE since it allows specialists and managers to be reached even if they are in a meeting. The system handles all callers on a top-down priority basis, which allows all employees to have their daily flow of communications filtered smartly. All they have to do is keep their calendar updated.

54

Michael Meyer

Every day companies and organizations around the world use business applications such as email, calendar software, browser-based intranet applications and specialized software such as ERP and CRM. Customers and partners often interact with companies using Web-based applications such as ordering systems. Real-time communications can make all these applications much more productive. Communications hubs that allow the medium to match individual availability, location, device and network are an added bonus. A range of new applications that integrate communications will follow this development. Bringing collaboration related functions into traditional business software is something that Siemens is pioneering and promoting. For example, the Communications Broker incorporates APIs so that third parties can develop applications, e.g. create workgroup productivity solutions for vertical markets and special requirements. Delays often occur in communications-intensive processes when employees work from different locations and/or when they are mobile. If the relevant people cannot be contacted because they do not respond to emails or respond too late then issues tend to escalate. Discontinuities in the communications flow, e.g. different media and/or mobile devices being used can also result in longer delay times. Over half the average workforce is mobile and this figure is set to rise to two-thirds by 2006 according to an IDC report. Couple this statistic with the need to access, manage and exchange information and it is clear that a mobile replication of the wireline office is a pressing issue. Previous attempts were not successful. Wireless voice and wireline data occupied disparate domains and integration was complex and therefore costly. But the new 2.5G networks and Wi-Fi hot spots employ the same communications protocol (IP) as wireline networks and this has enabled the wireless paradigm to be realized.

Real value in real time

55

Access via any device

Figure 13: Communications portals enable anytime, anywhere access to corporate data, personal information and collaborative tools

The increase in mobility, driven initially by the spectacular success of cellular telephony, has resulted in the rapid growth of Wireless LANs, i.e. extensions to the wireline infrastructure. These and other developments indicate that wireless will become the normal way of communicating and working, i.e. mobile devices and air interfaces will be increasingly used to access the resources of wireline corporate infrastructures. Sales and service employees stand to gain the biggest benefits from wireless access to corporate applications. The type of mobile portal shown in figure 13 shortens response times and boosts productivity. Access to messages and information, both personal and corporate, is enabled via the browser interface, which can be customized by users in order to match their individual requirements. For example, a salesperson would typically go directly to the product and customer information that he/she handles. Customization therefore meets the critical need to find the right information at the right time and in the case of mobile workers that need can become mission-critical. Conclusions. The real-time enterprise is predicated on the need for people to decisions in real time. Currently there is a gap between the speed of automated business processes and the ability of the relevant participants to react to events. Voice-data convergence (1gIP) facilitates closure but that is

56

Michael Meyer

all: it is a necessary first step. In 2gIP environments communications and information are unified at the application and service layer. This allows people to become an integral part of business processes. Siemens' 2gIP real-time software also enables smarter, more productive ways of communicating and collaborating as well as seamless integration with IT platforms and business processes. Thus, the RTE is not a stand-alone goal that is enabled via yet another point solution. The ability to take decisions in real time is itself predicated on the need to know who is available, i.e. presence needs to be displayed to all authorized parties. Presence also needs to be managed and at times be overridden by a non-intrusive message. This facility is particularly important in the case of the RTE since it allows specialists and managers to be reached even if they are in a meeting. The system handles all callers on a top-down priority basis, which allows all employees to have their daily flow of communications filtered smartly. All they have to do is keep their calendar updated. And last but by no means least, the mobile employees must be given the same functionality as their wireline colleagues and have access to the same corporate resources.

Tom Davies, Barry Gilbert, Jeff Swartz

Competitive response: a new lens for evaluating company performance

The global business environment companies compete in today is more complex, turbulent, and dynamic than it has been in decades. Not since the early part of the 20th century have businesses operated with so much uncertainty, volatility, and risk as they struggle to grow and compete for advantage in a global marketplace. Top executives are being exhorted at every turn to be faster, more nimble, more resilient, and more adaptive. They are recreating their global enterprises, in some cases from the ground up. Transformation is the new watchword and speed the new imperative. This new business environment is forcing top executives to re-think how they compete, and what capabilities they need to do so. Most are in the business of playing catch up. After spending a good part of the last decade focused primarily on internal priorities, businesses are once again shifting their attention outward. As companies shift their focus, it is becoming increasingly apparent that speed is only one of many new competitive mandates. Adding to the challenge, as pointed out in a recent Harvard Business Review article, is that speed, like many of the capabilities companies need – such as leadership and talent – is an intangible asset.1 It is also becoming increasingly clear that company executives need new frameworks for evaluating their competitive capabilities and measuring performance against them. In the absence of an organizing framework, companies run the risk of becoming overly reactive, responding too quickly to the competitive imperative of the moment and risking a disappointing overall business performance. In our work with leading global technology companies, we have found it useful to view corporate business performance through a new lens that we Dave Ulrich and Norm Smallwood, “Capitalizing on Capabilities,” Harvard Business Review (June 2004: 119-127.

1

58

Tom Davies, Barry Gilbert, Jeff Swartz

call competitive responsiveness. The term is defined as the overall capability of a company to respond to changes in its external environment – especially opportunities and threats – in a way that leads to measurable improvement in business performance. The framework enables a company to manage its response to turbulent changes in its external environment – competitors' moves, changes in economic and political conditions, disruptive technology, sudden regulatory changes, and the unanticipated realignment of supply chain partners – rather than feeling that external events are beyond corporate control. While still in its formative stages, we believe the framework is a useful construct for understanding and evaluating changes in a company's competitive performance over time, as well as against its direct competitors. The competitive responsiveness framework also serves as a valuable tool for determining how to best respond to competitive pressures and rapidly unfolding external events. The framework can also be used by top executives to identify the root causes of business performance problems in order to address areas where future competitive responsiveness can be improved. We firmly believe that the relevance and importance of competitive responsiveness is not limited to the private sector. As the events of 9/11 demonstrated, responsiveness to changes in global, political, economic, and military conditions is of critical concern to countries as well companies. The independent investigations of the events leading up to, and following, the terrorist attacks on the United States have shown that improving competitive responsiveness is one of the greatest challenges facing all global enterprises, public and private alike.

Competitive responsiveness Competitive responsiveness is the measure of a company’s capability to respond to changes in external conditions and events. The overall goal is to improve the company’s business performance by creating an advantage over its competitors in its ability to react to its external environment. The emphasis and point of departure of the framework is its focus on understanding how the business responds to changes that originate outside the boundaries of the company. Quite simply, competitive responsiveness requires an externally-driven view of a business and its capabilities. Of particular interest is how companies respond to external changes that create new opportunities or in some way threaten the company. Competitive responsiveness is a multi-dimensional lens. We have identi-

Competitive response

59

fied three recurring dimensions that companies should use to measure competitive responsiveness: SPEED – How quickly does the business respond to external opportunities and threats? Companies compete on how fast they respond to events and developments that occur every hour of every day in their markets. Whether it is breaking news, acquisitions, or changes in economic and demographic conditions, head-to-head competitors compete in a shared environment, and thus compete on the pace at which they respond to events that change their business environment. CONSISTENCY – How consistently across the enterprise does the company respond to external opportunities and threats? Ralph Waldo Emerson once said, “A foolish consistency is the hobgoblin of little minds.” Of course, Emerson was never a customer of a large global enterprise. If he had been, Emerson would have given consistency greater respect. Consistency in response to a market opportunity or threat can be the difference between execution success and failure. In a global enterprise, much of the energy and organization is now devoted to “getting the company’s act together.” The goal is to present the market with a unified look and feel, rather than having different divisions acting as individual silos under the corporate banner. Companies are struggling to keep all parts of their operations in synch. It’s not unusual to find that the messaging going out to customers via their contact centers is out of synch with the messages the direct sales force is promoting, and even more out of synch with their channel partnersresellers and distributors. EFFECTIVENESS – How effectively does the business answer the question, “Why should a customer choose my company?” If being first were all it took to win, businesses would be optimized for speed. But businesses are not, and one reason they are not is that winning is about creating new customer relationships, and being first is only part of that equation. While being first to respond to a customer is a competitive advantage, the challenge of convincing the customer to choose to do business with a particular company remains.

60

Tom Davies, Barry Gilbert, Jeff Swartz

We are all familiar with companies that pride themselves on being the first to respond to a market opportunity, only to discover that they lost the sale to a competitor that arrived on the scene much later. We see this frequently in markets with intense competitive rivalry and multiple customer touch points. Carefully prepared marketing plans and promotional launches often go astray, for instance, as soon as the customers begin to pepper the contact center with a laundry list of questions about the company’s products and why they are better than the competitors. Contact center agents are typically unprepared to answer difficult questions regarding differentiation that often require direct company comparisons. The three dimensions of competitive responsiveness- speed, consistency, and effectiveness- aren’t equally important, however. While speed is often elevated to a high priority, competitive responsiveness is by no means limited in scope to this one dimension. The relative importance of various dimensions of competitive responsiveness depends on the overall competitive strategy of the company and its competitive positioning. For example, a telecommunications company recently rated effectiveness as the most important dimension of its competitive responsiveness. This company’s competitive strategy was to be a “fast follower,” and, as a result, the telecom business rated speed as a lower priority. It is also important to note that speed, consistency, and effectiveness are not the only dimensions of competitive responsiveness. For example, the congruence of the response with other company capabilities and resources is an extremely important consideration. Companies that act without having the necessary resources and systems in place are likely to stumble. It is important to note that responsiveness does not mean simply being reactive. Responsiveness encompasses both an ability to anticipate, and proactively seek out, as well as an ability to react. A competitively-responsive company is one that reacts in a way that leads to improvements in business performance, vis-a-vis its competition, by being both proactive and reactive, depending on the situation. A number of factors contribute, either directly or indirectly, to a company’s competitive response capabilities. A company’s capabilities have their origin in the business’ business processes, systems, skills, structures, and culture. For example, a company that rewards inward-looking values and behaviors is not likely to place a priority on building capabilities that enable it to be competitively responsive to external events.

Competitive response

61

Business Performance

High

Vulnerable Coasters

Responsive Performers

Laggards

Underachievers High

Low Competitive Responsiveness

Figure 14: Competitive Responsiveness Quadrant

The quadrant in Figure 14 depicts what we see as the overall relationship between competitive responsiveness and business performance. The four quadrants show the relative scores of companies on competitive responsiveness (X axis) in relation to their relative scores on business performance (Y axis). Companies fall into one of four groups: Responsive Performers (Upper Right Quadrant): These companies are highly responsive to changes in the external market with deep capabilities to respond quickly, consistently, and effectively to rapidly changing external circumstances. The resulting inherent competitive advantage translates into an excellent business performance for the company. Underachievers (Lower Right Quadrant): These companies are highly responsive but have been unable to convert their advantage in competitive responsiveness into improved business performance. They are likely to have other significant barriers in the way that are impeding results and are inhibiting the company from assimilating the external changes. Vulnerable Coasters (Upper Left Quadrant): These companies suffer from inadequate competitive response capabilities, but demonstrate superior business performance. This is an unsustainable position. These companies are highly vulnerable to shifts in their environment, which will likely result in a major company crisis. They may have reached this position by being competitively responsive in the past, but are now suffering from inertia or an inward focus. Technology companies, which experience rapid changes in technology life cycles, may be especially vulnerable in this area. 2 Michael L. Tushman and Charles A. O’Reilly III, “Winning Through Innovation,” Harvard Business School Press (2002, Boston, Massachusetts).

62

Tom Davies, Barry Gilbert, Jeff Swartz

Laggards (Lower Left Quadrant): These are companies that have weak competitive response capabilities and whose business performance is lagging. There is little reason to expect these companies can significantly improve their position without a fundamental transformation of their business and their competitive response capabilities. They are probably overwhelmed by the rate and complexity of change occurring in the market.

Competitive response model In working with our clients to measure and improve their competitive responsiveness, we have identified a group of best practices that we believe all companies – especially those facing highly dynamic and fast-changing market conditions – need to implement. We have found that the best way to understand these competitive response practices is to place them in the context of a life cycle of competitive response. We refer to this as the Competitive Response Life Cycle (see diagram). The model reflects the lifecycle of activities that an enterprise must execute in order to “sense and respond” to changes in its external environment. The life cycle model unfolds in five phases. Life cycle phases

Sense and Capture: At this stage of the life cycle, the business is continuously scanning its environment, and having sensed that a change has occurred, the business has captured that change in its systems. Interpret and Create Awareness: Once an action or event has occurred, its importance and meaning to the company is determined, and the enterprise is made aware of what has happened in the market. Analyze and Inform: Competitive analysis places the change in context and helps determine how the change impacts the business. Deliberate and Decide: Alternative responses to the change are discussed and a command decision is made about how the enterprise will respond. Resources are allocated to ensure a fast, consistent, and effective response. Respond and Engage: Action is taken, directives are issued, resources are allocated, and the organization is engaged. Measure and Correct: The speed, consistency and effectiveness of the response are determined and appropriate corrective action is taken where needed.

Competitive response

63

Each phase of the life cycle is dependent on the prior phases. Consequently a breakdown in performance during any one phase will cause competitive performance issues in subsequent phases. Most obviously, a failure to sense an important change in the market will lead to huge gaps in competitive responsiveness. But breakdowns can and do occur at every phase, often going undetected for long periods of time without any apparent drop off in overall performance. Sense & Capture Interpret & Create Awareness Measure & Correct

Competitive Response Lifecycle Analyze & Inform Respond & Engage Deliberate & Decide

Figure 15: The competitive response life cycle phases

Competitive response best practices

One way to prevent degradations in competitive response performance is to put a series of best practices in place. Best practices can be associated with each phase of the life cycle. We have developed six best practices from our experience in working over the past six years with dozens of well-known, global information technology and telecommunications companies. Best practice: intelligence repositories

Intelligence repositories are collections of data files that serve as a repository of historical and real-time tactical intelligence. These repositories are often event driven and are grounded in typologies of different types of events that are likely to affect the performance of the business. One example of a repository is an increasingly rich intelligence file of

64

Tom Davies, Barry Gilbert, Jeff Swartz

changes in competitors' pricing and promotions collected on a daily basis in highly volatile markets, such as the wireless services market, which includes wireless service plans and handsets, as well as desktop systems, which includes PCs, notebook computers, PDAs, and peripherals. There are many sub-practices associated with these repositories such as: • Enforcing measurable update standards, such as updating the repository within 24 business hours of any change to an important market condition or the occurrence of a market event. • Making the intelligence available in several formats so the data can be imported into internal systems and desktop databases. • Dedicating teams of trained research associates to track critical market conditions, such as monitoring competitor print advertising and Web sites. • Using standardized data definitions and business rules.

Lifecycle Phase Sense & Capture Interpret & Create Awareness Analyze & Inform

Best Practice Intelligence Repositories Alerts & Broad Access Reporting based on Pattern Recognition

Deliberate & Decide

Escalation

Respond & Engage

Scripting

Measure & Correct

Scorecards

Figure 16: Best practices associated with competitive response life cycle phases Best practice: alerts and broad access

Intelligence about changes in external conditions and competitors' actions needs to be disseminated throughout the extended enterprise to include channel partners, and in some cases, even customers. To optimize productivity and effectiveness, intelligence must be made available to authorized individuals, at any time, from any location, and repurposed to support their business function and job responsibilities.

Competitive response

65

One of the most promising areas for realizing significant “return-onintelligence” is the effort of a business to make intelligence available on a broad basis to all customer-facing personnel. For example, customer service personnel need access to competitive intelligence that allows them to make side-by-side comparisons of different offers. Alerts need to be personalized to highlight issues of concern to specific individuals in specific job functions. A product manager for a new offering may need to be alerted to changes in multiple areas – such as pricing, promotion, competitor moves – whereas a pricing analyst may only be interested in fluctuations in pricing. Access to external intelligence to promote competitive response is facilitated by: • Building capabilities to both “push” intelligence to users and to enable users to “pull” intelligence from it. • Linking repositories and files such as competitive pricing and promotions to reveal underlying relationships and provide context. • Preparing intelligence assessments that answer core tactical questions such as “what happened,” “so what,” and “now what” • Giving employees, customers, and key members of the supply chain access to: – Industry news – Sales, channel, and partnership news – Region-specific variations – Pricing and promotions – Product specifications – Independent third-party analysis and reports • Providing inquiry support to help interpret and make sense of the external changes that have occurred. Best practice: reporting based on pattern recognition

Summary analytical reports are needed to provide answers to structured competitive assessment queries. The most useful ones employ pattern recognition in order to reveal subtleties that are not apparent at first glance. The reports need to be designed in collaboration with the functional areas of the organization they are intended to support. For example, highly targeted analytical reports can be customized for executives, field sales, pricing analysts, and product managers.

66

Tom Davies, Barry Gilbert, Jeff Swartz

While reporting norms are often very company-specific, there are some proven practices: • Use standardized formats to facilitate a shared understanding of the external landscape. • Embed aggregate reports with the appropriate links back to the details contained in the intelligence repository. • Use pattern recognition tools, such as geographic information systems, to identify “hot spots” of increased competitor and customer activity. • Use repeatable queries to structure the reports in ways that are most meaningful to the intended audience. • Allow for personalization and customization. Best practice: escalation

Escalation practices are standard operating procedures that guide an organization's performance and response to external events. These serve to direct company actions and often become the way business is done. Escalation practices provide clear description of who is responsible for doing what, and within what time frames, in response to critical competitive events such as the introduction of a new product by a key competitor. One company we've worked with has clear standards for responding to potentially threatening competitor moves. Their escalation practices unfold according to strict time frames, milestones, and clearly defined roles and responsibilities. For example, an alert is immediately distributed to functional managers, and executives notifying them of the external development. Within hours a second alert is sent out containing an evaluation of the details of the competitor move and a determination of its threat. Managers representing the different functional areas of the business then meet to examine options and prepare a recommended response. This is followed by a meeting of the senior executives overseeing the business who deliberate and decide what actions to take. While escalation practices tend to evolve over time, they should contain: • • • • •

Pre-defined milestones and prescriptive behavioral actions. Clear definition of roles and responsibilities. Measurable standards of performance. Gradations in response depending upon the severity of the market event. Triggers for other business processes impacted by the external event

Competitive response

67

Best practice: scripting

All enterprises struggle with the challenge of presenting a unified face to the market. Scripts are one tool for achieving this ever so elusive goal. In their simplest form, these are pre-scripted competitive response messages for customer-facing personnel such as direct sales, customer service, and contact centers. Used effectively, they enable an enterprise to stay “on message.” Most often the scripts are created by experts in competitive response and marketing communications. Script development is overseen by a peer review process, and these scripts can be bounded by legal constraints if they make claims about a competitor's product. The most effective scripts are: • Written in response to very specific tactical external events. • Provide very clear answers to questions such as “Why do business with us?” • Grounded in fact-based or evidence based research. • Refreshed continuously as market circumstances change. Here is an example of a script that was used by a company in response to a competitor introducing a new threatening product. This script was fed directly into the company's contact centers, retail outlets, and direct sales force within 48 hours of the competitive event occurring. Why Choose ABC Wireless Corp. Over Competitor XYZ Inc.? • Competitor XYZ uses technology that only allows its users to access their service while in a XYZ Nationwide Network city. • A contract with ABC can be terminated for only $175, while XYZ charges $200. • ABC Wireless offers always-on connectivity, new faster data transmission rates, access to wireless Internet sites, e-mail access, and cross-carrier shared messaging

68

Tom Davies, Barry Gilbert, Jeff Swartz

Scorecards

Competitive response scorecards enable benchmarking against agreed upon milestones and standards of performance for competitive responsiveness. These scorecards contain readily understood measures of actual competitive response performance across all dimensions of speed, consistency, and effectiveness. And scorecards also show which dimensions of competitive responsiveness are strengths for the company and which areas need improvement. Competitive response scorecards need to be tailored to the business function. For example, the marketing team may want to determine how competitively responsive the operational contact centers are in executing a new product roll out. Scorecards based on calls into the contact centers allow marketing to evaluate the competitive response performance of the contact centers. These “market pulses” allow for measurement and tracking of the competitive responsiveness of contact centers over time and across different locations. Corporate level strategists may want a more strategic look based on ratings of companies and an assessment of how their market momentum is changing. Or strategists may require near-term risk assessments of the volatility and complexity of the business environment. Scorecards are decision-making aids that produce their greatest value when they do the following: • Follow a predictable schedule, as in quarterly or monthly, as the business requires. • Track individual competitor's actions over the recent past and likely actions in the upcoming period. • Provide both forward looking and retrospective views of the company's competitive responsiveness • Contain measures of speed, consistency, and effectiveness of direct competitors • Include an independent third-party perspective to maintain objectivity. Work remaining

The competitive response framework is a work in progress. One promising area of development is measurement tools. Companies need to see how they are responding competitively to opportunities and threats compared to their competitors. The work we've done in this area to date has relied primarily on subjective evaluations and perspectives of key executives. While

Competitive response

69

informative, this approach clearly has its limitations. It's been our experience that companies have “notational” ideas about how responsive they need to be in different situations. But it is rare, however, to find any company that has developed agreed-upon measurable performance goals around competitive responsiveness (for example, how quickly it needs to respond to specific situations or events, or how much consistency in responding to a competitor threat). A second area for future development is the design and deployment of enterprise solutions which provide companies with the competitive response capabilities they need to be successful. Comprehensive solutions that address a company's organizational structure, systems, information management practices, and business processes are needed. End-to-end competitive response solutions will grow in importance, and may possibly rival investments that have been made in customer relationship management (CRM). A third area of future development is to deploy the competitive response model in industries other than information technology and telecommunications. Financial services and pharmaceuticals, for example, are equally dynamic and fast-changing industries requiring companies to rapidly adjust to shifting circumstances on a continuous basis. Companies are just now beginning to invest in developing the competitive response capabilities needed to respond competitively to market opportunities and threats in this new era of globalization. As made painfully clear by the events of 9/11, it is now unacceptable, even for a brief moment, to take your eyes off what is happening outside the company walls. Companies are beginning to make up for their past preoccupation with what is going on inside their enterprises. As they shift their focus externally they will need to become as proficient in responding competitively to turbulent and risky environments as they are at managing critical resources such as human resources and finance. Competitive response is the core capability that will determine the fate of many in the years ahead.

II. Typical examples from industries

There is no single path to the real-time enterprise, but there is the right path for each company and it is paved with information and communication technology. Concise industry examples describe approaches that have been tested in practice and indicate how the technology was successful deployed. Thus, the real-time enterprise is taking shape.

Susan J. Unger

Building the Real-Time Enterprise at DaimlerChrysler

The concept of the Real-Time Enterprise has emerged gradually at DaimlerChrysler over the past several years. It has developed in parallel with continuing advances in communication technologies and the changing role they play in business. Communications have evolved in recent years from more or less peripheral support systems to become a critical component of real-time operations. Today, real-time communications must not only be integrated in core business processes, it must be fully embedded: real-time is core process.

A global operation Worldwide, DaimlerChrysler employs 362,000 people, operates 104 plants in 37 countries, maintains 13,000 sales outlets in 200 countries, and partners with some 14,000 suppliers. The DaimlerChrysler IT organization supports 240,000 internal users, 150,000 PC's and 63,000 PDA's. The past few years have seen enormous growth in the use of cell phones, PDAs, laptops, email and video conferencing. Instant messaging, for example, is used by 18% of our internal users today, compared to 8% just two years ago, and Chrysler Group alone saved $14 million through increased use of video conferencing last year. Today, average daily volumes at DaimlerChrysler have reached 110 video conferences, 1 million voice calls, 2 million internal emails, 400,000 Internet emails, 96 million hits on public web sites and 4 million hits on Intranet sites. Communications, however, is not just about email or the number of PDAs and PCs and voice calls or video conferences that take place every day. Real-time is a much broader concept, and more fundamental issues are at stake. At DaimlerChrysler, we are convinced that the company’s ability to compete and flourish depends on the extent to which business communications is integrated with core business processes. The objective, in fact,

74

Susan J. Unger

is not just to integrate, but to fully embed real-time communications in all core processes. Looking at core processes in the automobile industry, we can illustrate the evolution toward a real-time communications environment with examples from our efforts at DaimlerChrysler. The common starting point of all real-time initiatives lies in the struggle to identify and separate useful information from an ‘overload’ environment, and the common objective is to ensure that the right information can be delivered to the right people at the right time. At DaimlerChrysler, the Engineering effort spans the globe, requiring communication and collaboration among design groups working at different locations, in different time zones with different systems, software and languages. Our response to this fundamental business challenge was first formulated as a question: how can we make it possible for engineers on different continents, across different divisions to communicate and collaborate—around the world and around the clock? The pragmatic response on the part of the IT organization has been to develop the DaimlerChrysler eEngineeringPortal. The portal provides web access to engineering product data across the enterprise, enabling design groups around the world to access and work from the same set of data. Operating 24 hours a day, seven days a week, the portal provides a platform for global vehicle development. It offers powerful tools for viewing 2D and 3D engineering drawings, provides easy access to complex engineering data stored in different EDM (Electronic Document Management) systems, and enables the exchange of CATIA (collaborative design software from Dassault Systèmes AG) drawings and data for comment, mark-up and further processing by members of the various vehicle development teams. In some cases, design groups working on different continents are able to “hand off ” work at the end of a day to another group at the beginning of theirs; at the end of the day, the second group in turn hands its work over to a third group. In the space of 24 hours, work has rolled from Asia to Europe to the USA and back to Asia. By eliminating delays and unproductive downtime, this rolling 24-hour development process has contributed to a marked reduction in the time required to move a new vehicle from concept to production. More than 20,000 people at Mercedes Car Group, Commercial Vehicle Division, Chrysler Group and Mitsubishi Motors Corporation have access to the Engineering portal. At the time of writing, log-ins are running at 5,000 per week from more than 40 plants around the world. Usage of the Engineering Portal increased by 300% during 2003, and continues to grow.

Building the Real-Time Enterprise at DaimlerChrysler

75

Limited, secure access has also been provided to certain qualified suppliers, further extending the reach and power of this portal.

Figure 17: DaimlerChrysler eEngineering Portal

Bridging the gap between core processes Engineering has long been autonomous, but never entirely isolated from other core processes. In the past, when Engineering had completed designs for a new vehicle, the specifications and drawings were delivered to advanced manufacturing experts for comment and corrective feedback. At DaimlerChrysler, new 3D visualization tools have made it possible to obtain much more rapid feedback from manufacturing—during vehicle development and design phase, not following it. Working with Dassault Systèmes AG and the DELMIA software, we have developed a “virtual factory”, a digital manufacturing environment that we fully expect will revolutionize the way the automobile industry designs and manufactures vehicles. A further development of visualization tools that have been used in the design environment for a decade now, the DELMIA virtual manufacturing environment allows manufacturing experts to observe a vehicle as it moves through the assembly process, enabling them to foresee and eliminate

76

Susan J. Unger

potential problems and to feed this information back to the developers. A simple example: using the DELMIA virtual reality tools, manufacturing personnel were able to spot a wiring harness problem that would have required an assembly worker to crawl up into the trunk of the vehicle, reach up under an extrusion and make a “blind” installation. The engineering team member responsible for that design spec was informed of the problem, “saw” the error of his ways, and was able to make a simple design change to correct the problem. In the past, such visualization tools did not exist for the manufacturing environment. There were no imaging systems to help translate engineering data into visual reality. Today, the digital manufacturing environment makes possible a much richer exchange of information—instantaneously. The design teams are given immediate corrective feedback from the virtual factory floor regarding the impact of any given design on the manufacturing cycle. The result is a new capacity to pre-emptively eliminate design errors that in the past would have required costly, after-the-fact adjustments to the assembly line machinery. The virtual factory is an extremely powerful tool and has had an enormous impact precisely because it enables real-time collaboration and communication between manufacturing and engineering. The success of this collaborative visualization effort has led to its extension to the technical service division, whose experts are now able to use the digital manufacturing environment’s real-time visualization and mark-up tools to provide analogous feedback to engineering, allowing design teams to optimize a vehicle in terms of its serviceability as well. The result of these efforts and related initiatives has been a dramatic reduction in the product development cycle. Only a few years ago, it took from five to eight years to bring a new vehicle to market. Since then, progress in eliminating delays, shortening feedback cycles and bridging gaps between different processes, along with the introduction and progressive integration of real-time communications, has brought significant improvement. From concept to launch, development of the Chrysler Crossfire took only 18 months.

Real-time communications in manufacturing The need to operate a manufacturing plant 24 hours day, 7 days a week in order to meet customer demand for certain vehicles has put DaimlerChrysler’s manufacturing processes under enormous pressure. To help meet this challenge, extensive use is made of wireless communication technolo-

Building the Real-Time Enterprise at DaimlerChrysler

77

gies. Embedded in the machinery, wireless is used, for example, to monitor operations, allowing the identification of potential failure points and the replacement of failing components before they bring the production line to a halt. It has also enable a more effective distribution of the work load across the team responsible for maintaining the line, as wireless frees them to roam the entire line rather than wait for something to break and be poised to rush in and repair it. Wireless is essential in managing quality, as well. An Inline Quality Management (IQM) system is in use in all German plants and in East London, South Africa. IQM establishes automatic communication between quality inspectors, robots, reworkers and all production departments. Wireless handhelds are used to scan product numbers into the Production Control System (PCS), and quality inspectors are provided with a list of components and inspection guidance for each vehicle. If a fault is found, the handheld device transmits the relevant information via the PCS for subsequent rework downstream. Robots connected to the PCS receive a separate set of data for each vehicle and transmit real-time data to the PCS regarding torque and angle, welding and gap dimensions. The data feeds downstream body, paint, and final assembly operations. Wireless is also used in logistics to manage and control material flow and vehicle distribution. Every hour, some 1500 different carriers deliver components to DaimlerChrysler assembly plants, a total of more than 11,000 parts per day. Delivery is managed using “just-in-time” methodologies and is refined further using “sequence-in-line” wherever possible. In this case, wireless technology enables the DaimlerChrysler factory to track a supplier truck bringing in tires, for example, from the point at which the truck leaves the supplier location heading for the plant. The sequence in which vehicles are being built on the factory line dictates the time and precise location at which the tires will be required. This information is provided in real time to the supplier truck while it’s on the road, i.e. the driver is informed as to just when and at what point on the line the tires must be delivered. Between sequence in line and just in time, approximately 95% of deliveries to DaimlerChrysler plants are handled in this manner. Radio Frequency Identification (RFID) technology is used for specific applications such as in-plant tracking of returnable component containers, and on a limited basis to communicate vehicle-specific data from the vehicle to the body, paint and assembly control systems. As costs for this technology come down, RFID will find much broader application. The use of wireless technologies in manufacturing operations represents an example of fully embedded real-time communications. In the Sindelfin-

78

Susan J. Unger

gen (Germany) plant alone, DaimlerChrysler builds more than 500,000 vehicles a year, no two of which are identical on any given day. Embedded, real-time wireless communications makes it possible to manage that diversity.

Sales and marketing The diversity of vehicles manufactured in Sindelfingen mirrors and is a response to the changing requirements of a mass market for customized goods. Changes in the market have been accompanied by changes in the way the company communicates with its customers, as well, as elements of real-time communications are integrated into the core sales and marketing process. The real-time response capabilities of web sites represent a separate front in the continuing effort to reduce and eventually eliminate delays and down-time. By providing a broad variety of information and services to customers and prospects, the Mercedes-Benz web site, for example, serves to shorten the sales cycle. The site offers a wealth of information: the history and traditions associated with the world’s mostly highly prized brand, Mercedes lifestyle news, sports, and entertainment. Prospective customers are able to arm themselves with detailed product information, locate a dealer and take advantage of financial services on offer from DaimlerChrysler Bank, all on the site. Further support for the sales and marketing process is currently under development in the form of a dedicated “DealerConnect” portal. Designed to support the sales process on-site at dealerships around the world, the new dealer portal will provide all of the tools, applications and access to realtime data necessary for dealers to do their jobs. Both the “DealerConnect” portal and the various public web sites are examples of the continuing effort to integrate real-time communications into the core sales and marketing process.

Real-time employees At the risk of stating the obvious, it’s clear that employees are a company’s most valuable asset, and no core process is more important than the employee process at DaimlerChrysler. The primary business challenge is to enhance the productivity of all employees, which is more of a challenge than might first appear. Despite the remarkable growth in the use of com-

Building the Real-Time Enterprise at DaimlerChrysler

79

munications devices and systems, the reality that lies behind the figures cited earlier is both complex and ambiguous. According to a recent survey published by the American Management Association, the average user today spends 1 hour and 47 minutes a day on email. What the survey does not indicate is the extent to which that time is well spent, or what it costs companies in terms of unproductive time—the time it takes to access different accounts on different systems, to recover lost messages from backup systems, to ship archives and address books to new systems when people move on to new jobs, and so on. Six years ago, before the merger that led to the formation of DaimlerChrysler, there were 17 (!) different email systems in use. Today, there is one. As part of the company’s broad-based effort to enhance productivity, the IT organization at DaimlerChrysler has promoted standardization across communication device types and extensive integration of disparate communications systems and networks. With respect to employees, recent efforts have been concentrated on the employee portal. The portal represents one facet of the company’s long-term effort to recruit and retain the most talented people on the labor market. It is designed to deliver the best working tools available and to enrich the day-to-day work experience of DaimlerChrysler employees.

Figure 18: DaimlerChrysler Employee Portal

80

Susan J. Unger

Accessible by more than 170,000 employees in Germany and Spain, the employee portal averages more than 80,000 logins per day. Sixty percent of all employees with PC access make use of the portal, with usage increasing as features are enhanced and availability extended to reach more employees. Designed to help reduce information overload and eliminate downtime, the portal aggregates job-related and private employee information under a single interface. Employees need sign on only once to access information and services ranging from job boards to time logs and travel management, on-line work processes and job-specific knowledge bases and web sites. Real-time communications tools are being progressively integrated into the portal as it evolves; plans foresee the integration in a single interface of real-time services including voice and video-conferencing, instant messaging and presence information, online collaboration using white boarding and application sharing, as well as telephony, email, calendar and access via PDAs to web-based data and applications.

Future developments: expanding the Real-Time Enterprise The portals, projects and real-time communications initiatives discussed in this article illustrate the determined effort underway at DaimlerChrysler to transform the company into a Real-Time Enterprise. Ideally, real-time communications eliminates unnecessary delays and unproductive downtime, and this remains the overall objective. Progress continues on a variety of fronts across the enterprise as real-time communications tools are integrated step by step into evolving core processes. In certain areas, such as the use of wireless communications in manufacturing, we can legitimately speak of fully integrated, “embedded” real time communications. The development and implementation of engineering and employee portals represents an extension of the integration process, as does the progressive expansion of portal access to customers, dealers, suppliers and business partners of DaimlerChrysler. From an embedded communications core, real-time communications continues to expand, progressively implementing the DaimlerChrysler vision of a Real-Time Enterprise, an enterprise in tune with the real-time operation of the global economy. Real-time is core process.

Bernd Voigt

Real-time-oriented service architectures in the airline sector

The schism in electrical telecommunications – the splitting of data and voice communication into separate channels – is coming to an end after more than 150 years. Voice over IP (VoIP) enables the convergence of data and voice traffic. And this, in turn, makes it possible to design real-time-oriented service architectures. In an ideal scenario, all media and terminals are available for exchanging data with company IT applications. All employees participating in the business process communicate data and voice using the medium best suited for the task. In addition to the communication requirements within a single company, new demands arise from the cooperation between different companies. For example, the complex processes that prevail in the airline sector require real-time access to information from alliance partners. Further applications are found in the process chain from passenger check-in to arrival at the destination, communication during flight preparation or telemedical queries. Real-time-based service architectures ensure that these specific and highly sensitive communication structures and process flows run smoothly. The integration of telephony (VoIP) makes it a key component of the basic IT infrastructure. A real-time-oriented service architecture must be professionally operated in one or more computer centers with fixed service level agreements (SLA). After a short digression into the beginnings of electrical telecommunications, this article will present a number of case studies from the Lufthansa environment (FlyNet, Integrated Ground Cockpit Communication, Message Switching and StarNet). Basic features of a real-time-oriented service architecture will be described in conclusion.

82

Bernd Voigt

Introduction or: ‘On the ancient Greeks, Mack the Knife and a horse that does not eat cucumber salad’

Marathon

September 490 BC: On the coastal plains west of Athens near Marathon, the Persian army faced the clearly outnumbered armed forces of Athens. The ensuing events are well known: The soldiers of the Attic democracy emerged from the battle victorious and Greek general Miltiades sent a messenger to Athens 40 kilometers away to bring the city news of the triumph. The exhausted messenger was barely able to deliver the news before he collapsed and died. We expect more from real-time architecture. Reliability and sturdiness are obvious requirements. Spectacular yet unrepeatable peaks in performance are suitable for the Guiness Book of World Records but rarely form the basis of a successful company. Brecht’s mounted messenger from the queen

Bertolt Brecht’s “Three Penny Opera” also features a messenger, who plays an important role toward the end of the play: When Macheath, better known as Mack the Knife, is being led to the gallows, a messenger appears with a pardon from the queen, an event that is commented on by another protagonist of the play with the words: “And so, in the end, everything turns out well. How easy and peaceful would be our lives if the queen’s mounted messengers always came." The queen’s mounted messenger arrives “just in time,” as we would say today. And so we find in Brecht a key element of real-time architecture: Information is accessible whenever and wherever it is required. Yet it does not weigh us down, just as we are not robbed of suspense in the Brecht play. Real-time architecture is the resolute realization of the “anywhere, anytime, any device” principle, complemented by “as much as necessary and whenever required.” Telegraphy

More than 2,000 years lie between the battle of Marathon and the first attempts at electrical telecommunications with encrypted signals by

Real-time-oriented service architectures in the airline sector

83

Samuel Soemmering in 1809. The discovery of electricity in the 17th and 18th century led to the development of electrical telegraphy and the telephone one century later, laying the foundation of today’s communication systems. A milestone on the path to modern telegraphy was the telegraph, developed in 1837 by Samuel Morse. A more advanced device based on this principle was put into operation between Washington and Baltimore in 1844. Even the “old Europe” recognized it as a sign of the times. In 1848, a company founded one year earlier by Werner Siemens and Johann Georg Halske, the “Telegraphen-Bauanstalt Siemens & Halske,” was commissioned by the Prussian State with setting up a reliable telecommunications connection between Berlin and Frankfurt/Main, thereby making it possible for the king and the government to follow the decisions of the German National Assembly in the Paulskirche in Frankfurt in real time. Telephony and the schism in electrical telecommunications

The second half of the 19th century saw the beginnings of the development of the telephone (derived from Ancient Greek, “to sound from afar”). The Parisian telegraph officer Charles Bourseul, who in 1854 wrote a paper on the techniques of electrical voice transmission, was still ridiculed as a “harmless lunatic.” Only seven years later, the German Johann Philipp Reis presented his telephone to the public in a lecture at the Physical Institute in Frankfurt/Main. “A horse does not eat cucumber salad,” was the first transmitted and verified sentence. This essentially silly phrase itself is an indication that serious applications of this new technology went unrecognized at first. In addition, telegraph companies that had become well-established by then tended to view the new invention as a threat rather than a promising new business opportunity. A convincing business case was missing, and it is here that the schism in electrical telecommunications had its origins. Wars, too, played no small role in advancing the development of telecommunications technology. The first transatlantic cable for telephone traffic was finally put into operation in 1955. The natural drive for more flexibility and faster business processes proved to be stronger than any unease established companies may have felt at the cannibalization of their business models. The fast will conquer the slow. And herein lies the core of real-time architecture: acceleration of business processes through the optimal utilization of advanced communication technology.

84

Bernd Voigt

Integrated communication processes using Lufthansa as an example

FlyNet

Lufthansa flight number LH 418 from Frankfurt to Washington: After the meal, a passenger suddenly breaks out in a sweat. As his condition worsens, he calls a flight attendant. The passenger collapses in front of the flight attendant and loses consciousness. Flight attendants are equipped for emergencies: Even if there is no medical doctor on board, the crew can use the emergency case to quickly take an ECG and measure heart rate, blood pressure, blood oxygen saturation and body temperature. For this purpose, the patient is connected by cables to a computer inside the emergency case. For example, the oxygen measurement sensor is connected to the passenger’s finger and the blood pressure meter is attached to the arm. ECG electrodes are also applied. The computer records and forwards the data. The patient data is encrypted and transmitted in real time via the onboard wireless LAN to an antenna on the airplane, then to a satellite and from there to a ground station and finally via the Internet to a 24-hour call center. There, an emergency doctor analyzes the data. In reverse order, he sends treatment recommendations back to the airplane. In this case, he offers reassurance: “No heart attack, no stroke, only temporary circulatory collapse. Alcoholic beverages are strictly contraindicated for this patient.” The crew notifies the pilot: Flight may continue, stopover not required. A vision of the future? Not entirely! The situation described here is a test scenario: In real life, the “patient” is an employee of the German Center for Air and Space Travel (DLR) in Cologne. He was “treated” by the director of medical services of Deutsche Lufthansa. The expert on the ground was the director of the Institute for Air and Space Travel Medicine at the DLR. This trial examined the potential for communication based on the new broadband Internet connection on board an airplane of the Lufthansa longrange fleet and novel medical equipment. The communication system is based on FlyNet, the broadband Internet connection used on the Lufthansa fleet. The airline will gradually equip the entire intercontinental fleet with Flynet by the summer of 2006. By means of an antenna, the airplane maintains steady contact with a satellite that, in turn, remains in contact with the ground station. From there, the information reaches the Internet. On board

Real-time-oriented service architectures in the airline sector

85

the airplane, the system is accessed via wireless LAN. The feasibility of installing a broadband connection was what prompted Lufthansa Medical Service to test whether onboard telemedicine can be practiced effectively. The devices tested by the partner organization DLR included prototypes of onboard emergency care devices. Flight attendants are faced with medical emergencies on a daily basis. In 2002, there were 1,600 medical incidents in the air for a total of approximately 44 million Lufthansa passengers. Of these, roughly 900 were due to circulatory collapse. In addition to the transmission of measured patient data, it is also possible to establish direct voice contact between the emergency doctor on the ground and the flight attendants responsible for the patient. Changes in the patient’s condition can be communicated directly. If the verbal description is insufficient, images of the patient can also be transmitted. On the ground, Lufthansa works together with International SOS, an organization with access to medical expertise and a database showing which medical services are offered at what location in the event of an emergency. When flying over Siberia, Nepal or Greenland, this is not a trivial task. It is of little use to a patient, if he is dropped in a place without medication or specialists, when he could have been admitted to a modern hospital two hours later. The services provided to flight attendants by FlyNet are not limited to medical emergencies. By means of the Internet connection on board the airplane, flight attendants can connect to the Lufthansa network via an encrypted connection, retrieve briefing information on the next flight, coordinate their subsequent assignments and exchange information with other flight attendants. For the first time, a large number of flying personnel can be integrated into a direct communication system. This technology also offers added convenience for passengers: They can read news, send e-mail meesages or simply surf the Internet, thereby keeping up to date throughout the flight. Integrated Ground Cockpit Communication (IGCC)

Even before departure, sophisticated communication methods are required during flight preparation to ensure that communication is rapid, secure and efficient. Different employees must communicate with one another to keep airplane ground time to a minimum. One system that has been in use for about six years is Integrated Ground Cockpit Communication (IGCC). This system supports communication processes during flight preparation using a function-based connection between individual officers (flight manager,

86

Bernd Voigt

ramp agent, cockpit, etc.). Communication is established via the respective flight number by entering the function without having actual knowledge of the type of terminal being addressed or its call number. The IGCC system offers interfaces for a great variety of communication systems. Regardless of whether flight preparation participants communicate via landline telephone (e.g. employees at the check-in counter), mobile phone (e.g. crew members) or trunked radio set (e.g. ramp agents), they all log into the IGCC system once. The system automatically recognizes the type of communication connection and stores this information together with data on the flight number and function of the logged-in participant. A caller then establishes a connection with another party using attributes (flight number = LH0123, function = pilot). The system identifies the person called and routes the call to the appropriate terminal (radio set, trunked radio set, landline or mobile phone). The advantage of this functionbased system is that it is not necessary for the caller to know the name of the person to be reached. Message Switching (MES)

In the airline sector, a global exchange of information and automated messages between applications has long been a necessity. A prime example of this is the so-called Message Switching system (MES), which was developed as the “New Lufthansa” was being formed in 1955 and emerged from the telecommunications/telex services field of the time. Communication first took place using Lufthansa’s own network and was performed manually (punched tapes were used for switching). Dedicated lines existed between the exchange in Frankfurt and the telex offices in New York City and Johannesburg. They permitted unobstructed communication to the United States and South Africa. In addition, there were connections to other telecommunication organizations whose networks were used for a fee. Deutsche Lufthansa had the most modern punched tape switching system of the time. Beginning in 1972, communication took place on a UNIVAC 494 computer system, the Lufthansa network and the SITA network, and on direct connections to selected airlines (e.g. Air France and British Airways). The MES application offers access to services such as telex, telefax and e-mail and acts as an exchange between onboard and ground communication. Any organizational unit can be reached anywhere in the world. MES is also used as a service by other airlines and airline-related companies,

Real-time-oriented service architectures in the airline sector

87

such as handling and forwarding agents, and it offers each customer access to real-time global communications. Under application of IATA standards, MES is the medium used by Lufthansa and participating airlines to handle worldwide intercompany communications between all air carriers and to guarantee that messages are delivered. In the airline industry, MES is a central component of corporate communications. As architectures are reorganized, it will be supplemented and replaced by IP-based services, thereby becoming an integral part of real-time-oriented service architecture. StarNet

With reference to this and as an example of a worldwide intercompany communications architecture, the StarNet product used by Star Alliance will be described. Since its formation in 1997, Star Alliance has become the largest and most developed alliance in international aviation, with approx. 433 billion seat miles sold annually, roughly 369 million passengers transported every year and a fleet of 2,477 airplanes. Star Alliance currently consists of 16 partner air carriers and continues to grow. New members are scheduled to join this year and next. StarNet is based on a shared network infrastructure of Star Alliance (Star Alliance StarNet Data Network). Among the objectives of StarNet are the integration of airline-specific applications and, above all, the harmonization of business processes. This includes, for example, transparent data access for all cooperation partners, rapid market access for the Star Alliance (application) services and easy integration of new business requirements and Star Alliance partners. At the starting point are completely different architectures, systems and protocols, and mostly mainframe systems that have been continuously developed and adapted over the years to match the processes specific to each particular airline. The solution is an architecture that is clearly structured into four layers. These four layers are connected with each other by an intelligent middleware. Figure 19 shows a schematic diagram. With this type of architecture, it is possible to define and realize shared business processes even with the most varied initial situations. The systems of the partner airlines are each connected once to StarNet via clearly defined interfaces. This allows a reduction of complexity. As a result, different companies can act as a virtual real-time enterprise. The system’s success is demonstrated by the applications that have already been implemented: Employees of an airline now have real-time access to current information on partner flights, allowing them to provide customers with immediate

88

Bernd Voigt

up-to-date information (e.g. flight scheduling changes) on flights of the cooperation partners. It is also possible to access flight booking data of the different partner airlines. Customers of frequent flyer programs (e.g. ‘miles and more’) can be given immediate information on the option of booking a flight in the corresponding booking class of a Star Alliance partner. The “messaging layer” supports the following main communication types: “query and response” (e.g. for flight booking queries) and “publish and subscribe” (e.g. for flight information, similar to the MES system described above). Detailed criteria for automatic information can be entered using a web-based subscription tool. What is StarNet ? An architecture consisting of ... Standard Middleware Services

Business Control Layer harmonizing data and processes

Messaging Layer supporting airlines standard traffic types

Translation Layer enabling applications to talk to each other

Connectivity Layer accessing airlines legacy systems

Figure 19: Schematic representation of StarNet architecture The messaging layer supports main traffic types ... Real-Time processing within 2–3 seconds query

Transactional Traffic (Type A) response

Store & Forward within 10 seconds publish

Messaging (Type B)

subscribe

■ Matching users expectations and experience on response time ■ Handling of widely used telex communication ■ Customer defines scope of information required

Figure 20: “Publish & subscribe” procedure

Real-time-oriented service architectures in the airline sector

89

StarNet connects applications of the various carriers via StarNet gateways in Chicago, Frankfurt/Main, Singapore and Tokyo. This enables the automatic adaptation of differing network protocols and data formats. The StarNet gateways can also be used to carry out “business rules,” such as the implementation of airline-specific data elements in IATA standards or data validation. An improvement in quality in the process chain, which is brought about by faster processing speeds and proactive customer service, affords Star Alliance considerable advantages.

And now?

The end of the schism

In all of the examples described above, real-time process integration is of central importance. One example deals with pure voice integration (IGCC), the others with data and system integration (MES and StarNet). FlyNet goes one step further: voice and data are integrated in a single shared process, a concept that goes far beyond what is commonly understood by “Voice over IP” (VoIP). This brings to an end the schism in electrical telecommunications that has lasted for some 150 years. In light of the reassuring news that bandwidth will definitely not lead to bottlenecks in the foreseeable future, this marks the birth of real-time enterprise. Integration – but how?

The examples presented above describe process states as they exist at present. The challenge now is to interconnect the various applications and functions of a company within an overall architecture. This is where the wheat is separated from the chaff. The goal is to provide a real-time-oriented service architecture that can be flexibly adapted to changes in business processes, that meets the relevant safety requirements and (last but not least) that is reliable and economical to operate. Admittedly, these are ambitious requirements, but they can be met. Below, we provide information on the key features of a real-time-oriented service architecture. One comment before I begin: All the talk about “digitization” misses the

90

Bernd Voigt

point. The fact that we “digitize” everything today is merely an indication of our technical limitations. We shouldn’t let myriads of zeros and ones block our view of reality. Reality also consists of other states, and it is only a matter of time before word gets (back) out to the IT branch. Humans: Are we still needed?

Real-time-oriented service architecture enables and supports cooperation of humans with systems, systems with humans, and humans with humans. Of primary importance is the cooperation between employees, whether within one company or in separate companies. To underscore this, we at Lufthansa also use the term “collaboration infrastructure.” Consequently, telephony based on “Voice over IP” (VoIP) is one of the cornerstones of real-time-oriented service architecture. Voice and data are integrated into one and the same architecture. With that, the age of separate telephone systems comes to an end. Telephony is now simply another server-based application that is hosted in a computer center, as are the others. Telephony has become an integral part of the basic IT infrastructure. Middleware

Figure 19 describes a layer-oriented architecture. The basic structure for connecting these layers is provided by middleware. As a result, a companywide architecture requires a company-wide middleware concept. It ensures that long-serving mainframe and client-server applications can “talk” to one another. IP-based voice telephony simply represents an additional interface that needs to be taken into account. Therefore, it is no coincidence that the most significant projects being implemented in many companies today deal with the introduction of a company-wide middleware architecture. Portals

Middleware ensures that IT systems are able to communicate with one another. Of course, it would also be beneficial if employees were able to seamlessly access the systems they require directly from their workplace. Possible solutions are provided by portals that combine all relevant information in a single user interface and integrate all applications, including e-mail, office applications and telephony. The user interface and menu

Real-time-oriented service architectures in the airline sector

91

guidance always follow the same logic and information can be controlled for specific target groups. Portals do more than simply open the door to the world of browser-based applications. They also make it possible to set up function-based workplaces and suitable workflows. Finally, an exchange of knowledge between participants can be promoted using filing systems and discussion forums. From his ground-based workplace, a doctor who is “treating” a patient on board an airplane can access the same systems he needs to do his work as the pilot in the cockpit who wishes to communicate with the ramp agent or the check-in agent who requires access to flight booking data. Each participant can access the applications necessary for his particular function. It goes without saying that a company-wide real-time-oriented service architecture, which connects a multitude of terminals and communication channels, requires a company-wide security policy. Of course, such a system must actually be implemented and enforced, and portals are ideally suited for this task as well. General access authorizations, definitions of functions and privileges, mechanisms for authentication, and a guarantee of confidentiality through certificate, signature and encryption are ensured by a company-wide security policy that is provided in the portal. Unified messaging systems (UMS)

Listening to and processing messages takes time. The mobile phone and the office landline telephone have separate voice mailboxes. SMS and e-mail messages also have to be processed. All of this can be greatly simplified by using unified messaging systems (UMS). In a UMS, a wide variety of different telecommunication channels (voice, e-mail, SMS, MMS, fax, etc.) are collected and stored in the mail system. The messages can then be accessed and processed from a single system. Having all messages accessible on a standardized interface greatly improves user efficiency. It is no longer necessary to dial into different systems and messages can easily be distributed for further processing or put into intermediate storage. It is irrelevant whether this is done on a PC or notebook, with Thin Clients or a Personal Digital Assistant (PDA). Personal Digital Assistants

Personal Digital Assistants are particularly suitable for mobile access to the unified mailbox. In modern system architectures, the PDA can access mailbox data through encrypted communication channels via any Internet

92

Bernd Voigt

access type, be it Wireless LAN, Bluetooth, GPRS or, in the future, UMTS. This is either done as a synchronization procedure initiated by the user or using an automatic push operation in which data is kept current in the central mailbox and the PDA without any further user intervention. The user can adjust the scope of transmitted data in accordance with his information requirements and the available bandwidth. For example, if push operation provides him only with the first 100 lines of an e-mail message without attachments, he can decide whether he wishes to receive the entire e-mail message with attachments during the next synchronization run or whether, based on the information already available to him, he wants to forward it to a colleague for further processing. In addition to mailbox data, push operation can also be used to synchronize other data such as price lists for the sales department or CRM data on all connected terminals. These types of systems ensure that the user always has the latest information available. Text-to-speech and speech-to-text

Through enhancements such as text-to-speech, text-based messages can also be accessed directly on telephone systems. Text-based messages such as e-mail, SMS and even faxes can be converted to voice messages and read to the user over the telephone. Control of the system is also voice-based. The user navigates through the e-mail system with voice commands to send, delete, forward and even respond to e-mail messages. E-mail messages can be answered by creating a voice file and sending it along with the e-mail message. Alternatively, so-called “speech-to-text” extensions can be used to convert spoken messages directly into the format of a text-based email message for sending. Availability management

We once believed that the liberation of information from corporeality (i.e. the much-invoked digitization) would take us directly into the ”land of milk and honey” of unlimited transparency and accessibility. “Always on” was one of the much-used buzzwords. Today we know that we don’t actually wish to be “always on.” Nor is the value of a manager measured by the number of daily e-mail messages or the number of mobile phone calls during meetings. To reiterate what has already been said above: Real-time architecture is the resolute realization of the “anywhere, anytime, any device” principle, complemented by “as much as necessary and whenever required.” This is supported by so-called “availability management,” which

Real-time-oriented service architectures in the airline sector

93

in actual practice tends to be “non-availability management.” Rules define how a telephone contact is to be established in each particular situation. For example, while going to or leaving a meeting, the user can have all calls diverted to the mobile phone but have them routed to the unified mailbox during the meeting itself. This can be defined as a general rule and would therefore be applied to all appointments entered in the schedule, including schedule items such as “I do not wish to be disturbed.” Exceptions can be made for the secretary or “supervisor needs info.” The communication paths available in each case can be displayed online, i.e. in the portal, and let the user trying to make contact select the most efficient method available. This enhances a person’s virtual availability while at the same time generating space for creativity. Running a real-time-oriented service architecture

Implementation of the “anywhere, anytime, any device” principle demands that the entire real-time-oriented service architecture be professionally operated in one or more computer centers with fixed service level agreements (SLA). The requirement for high-availability telephones (“always works”) now applies to the entire LAN. A central data storage and backup concept must be devised and implemented. A fact often overlooked is that e-mail messages require an archival and retrieval strategy as well, especially to comply with data safekeeping legislation. The good news is that solutions such as the SAN (storage area network) concept are available today. Business processes are subject to continuous flux in the markets, and the IT infrastructure and supporting applications must be able to keep pace. A combination of customer-specific concepts, development and high-performance operation of the IT infrastructure ensures a high degree of variability of business processes. Consequently, real-time-oriented service architectures must be provided that can be flexibly adapted to changes in business processes. This is the guiding principle of Lufthansa Systems, one of the leading providers of IT services in the airline and aviation industry.

Michael-W. Hartmann

Strategic perspectives for the hotel industry

More than any other business sector, today’s global hotel market is characterized by take-overs, consolidations and expansions. As well as this intensive competition, the establishments face the challenge of keeping pace with the growing requirements of demanding business and holiday travelers and, at the same time, to take steps against declining brand loyalty. The international hotel chains strive, therefore, to position and establish themselves globally. This is indispensable for achieving a high brand value. At the same time, it is becoming ever more apparent that there is a gap between the business strategy being pursued and the corresponding strategy in the Information and Communication technology (IaC) sector. In particular, the telecommunication solutions employed in hotels are often regarded purely as a necessity and not as a strategic means of improving services for the guests and increasing the efficiency of the hotel operation. Thus, above all, many establishments in the 4 and 5 star category offer antiquated and limited communication services that no longer satisfy today’s requirements of travelers and that do not meet the requirements of consistent brand positioning. As a result of the progress in hotel chain consolidation, companies are frequently confronted with a heterogeneous technological infrastructure, a basis which often only allows arduous implementation of a consistent level of services and working processes. The hotel industry is also faced with essential increases in efficiency, integration of new technologies in current infrastructures and optimization of guest services at a time when budgets are decreasing.

Growing requirements of guest service Today’s hotel guests have ever higher expectations of service levels. The means of communication which are used at home and at the workplace should also be available with the same quality when they are en route and

114

Michael-W. Hartmann

traveling. Standardized services are therefore expected as a matter of course at the different hotel locations. High-tech facilities in the hotel room quickly becomes a basic requirement which, in itself, does not result in any noticeable competitive advantage. Rather, this is achieved by excellent services and their efficient administration by corresponding technologies. Centralization and integration of the workflow for the reduction of costs and real-time management are the greatest requirements here.

Increasing the brand loyalty of hotel guests To counteract the decline in the brand loyalty of guests, many experts regard the consequent use of Customer Relationship Management – also known as Customer Retention Management – as a key concept for the hotel industry in the 21st century. The major challenge now is to define suitable brand-specific and segment-specific service levels. This must take account of the growing demands of hotel guests whilst, however, also providing the same quality worldwide. This business strategy should determine the required IaC strategy – and not the other way round.

Efficient infrastructure as a basic requirement The IaC infrastructure in a hotel must support the multitude of services which an establishment offers its guests. And, indeed, in such a way that costs are optimized and all services can be made available seamlessly in a process oriented way. For this reason, timely and professional planning of the infrastructure constitutes a considerable factor for success. The integration of different systems and the creation of standard user interfaces is simplified if the hotel chain collaborates with a provider who can provide complete turnkey solutions, because such a partnership reduces costs and facilitates coordination at a global level. This also allows easier outsourcing of these activities which do not belong to the core competences of a hotel company. Hoteliers must constantly find the optimum balance between a high level of customer satisfaction and the profitability of their hotel. This means they require an IaC infrastructure which is flexible, efficient, secure and reliable. Costs saving solutions, which have proved successful in other industries, should also be implemented in the hotel industry with its extremely mobile

Strategic perspectives for the hotel industry

115

‘round the clock’ operation and with hotels at different locations around the globe.

Modified investment model Currently, the hotel industry is changing its IaC investment model from CAPEX (Capital Expenditures) to OPEX (Operational Expenditures). Instead of fixed capital costs, the focus is placed on flexible operating costs. The OPEX model is primarily concentrated on centralized, combined service centers for voice and data hosting. This new approach significantly changes the value chain for hotels and providers of IaC solutions. That means the following points will gain considerable significance: • The organization model and the integration of the supply chain (centrally controlled processes, central monitoring and controlling, outsourcing and out-tasking, and redefinition of Key Performance Indicators (KPI)); • New business intelligence instruments which monitor, analyze and represent the defined KPI; • Technical infrastructure (interfaces, data security, availability, network redundancy, service quality, bandwidth availability, managed services); • Partner selection (Application Service Provision (ASP) models, global profile, financial strength, strategic alliances); • Business model (increased turnover, cost savings, internal service charges).

Trend towards real-time communication A major trend which will shortly affect the entire hotel industry is real-time communication. The convergence of voice and data communication on the basis of Internet Protocol (IP) is the core element of modern applications and contributes considerably to value added through the optimization and differentiation of business processes. This is due to the fact that a standard communication framework can be achieved by connecting different systems at different locations without any time delay. The introduction of RealTime Communications (RTC) will make the processes in hotel operation more efficient, more rapid and more secure within the foreseeable future. The focus of the first generation of IP communication (1gIP) is cost reduction, for example by more efficient utilization of the current network

116

Michael-W. Hartmann

infrastructure and joint network management. The use of IP convergence technologies means that advantages such as the reduction of operating costs and improved asset management can already be realized in this stage. What is interesting about the second generation of IP communication (2gIP) is that its main focus lies in the optimization and differentiation of the processes inside the company. The anticipated benefit of real-time communication for hotels is the reduction of investment, operating and management costs by at least 20 percent. This, however, can only be achieved by the implementation of a whole range of measures, e.g. economic consolidation of network operator and service contracts, managed services, network consolidation, and the introduction of convergence technology. Clear cost reductions can be made if voice services are managed in the same way as data applications. Also, voice communication via Internet protocol (Voice over IP) of the first generation still adheres to the traditional telephone system model. This, however, leads to some disadvantages in the current form of voice networks within companies. Location-related decentralization, proprietary protocols and terminals with reduced compatibility between the products of different manufacturers have made it impossible to establish proficient solutions with simple handling. IP systems of the second generation change the traditional communication model in every respect and consistently implement the existing advantages of IP in the networks of the next generation. Communications services thereby become data applications. Cost savings in operation and maintenance as well as the availability of new services and innovative hospitality applications rank among the first noticeable results which can be achieved by means of hosted real-time communication solutions. Siemens has developed a clearly defined strategy in order to provide cost effective models which ensure a transition to extremely scalable IP communication with tangible consequences on the business results. Moreover, Siemens Hospitality Services provide a vertically specialized line of business within Siemens Information and Communication Networks (ICN). Jointly with Siemens Building Technology (SBT), Siemens Hospitality Services – as a system integrator and turnkey solution provider – is responsible for the planning, implementation and the management of all IT, communication and building management systems within the hotel environment. “Vision 200X”, which was developed in workshops together with experts from the hotel industry, already incorporates a series of concrete ideas for future hotel solutions – i.e. in the guestroom, lobby, conference

Strategic perspectives for the hotel industry

117

rooms and back office areas. They are presented here.

Advanced hotel solutions for guestrooms Before long logically-linked “workpoints” will enable the connection of IP telephones with PC-based systems and displays. The following easy-to-use applications could soon become reality and lead to considerable increases in productivity: • Mobile “Voice over IP” telephones that work in a WLAN environment and make unified messaging services available; • Integration of multimedia devices in the room for IP workflows (unified messaging, managed telephone conferences, Internet access, content management); • Multiple output of information (telephone, PCs, PDAs, plasma displays, etc.); • “At your Service” workflow that provides and controls all guest-related services; • IP-based, centralized monitoring of the room via integrated touch screens; • Centrally hosted interactive multimedia contents which can be externally uploaded and transferred (“interactive TV”).

Advanced hotel solutions for public areas Hotels are increasingly becoming hotspots for WLAN roaming. This allows guests to establish an Internet connection to their network providers at different locations around the world. The new generation of mobile phones will be able to establish a connection to UMTS networks and also to the WLAN network on the hotel premises. In public areas such as the lobby or the spa, the increased security for data networks gained through new encryption methods plays an important role. New communication services will soon become available because the transmission bandwidth has increased tenfold in recent times as a result of new technology standards. In order to regenerate increased brand loyalty, profiled guest portals with different Internet access rates that depend on a specific member status (e.g. Platinum member) are of vital importance for retaining customers.

118

Michael-W. Hartmann

Advanced hotel solutions for conference areas In the conference area of the hotel, booking systems for communication services including multimedia conferences are progressively becoming a basic requirement for the organization of events and meetings. Also, temporary Virtual Private Networks (VPNs) reliably offer presentations and other event material after the conference. Participants can log into the hotel network from their own rooms and download the files they require. A software-based conference solution enables the reservation of communication and data services for planned conferences or meetings. All voice and data services can be made available at any time without the need for technical personnel. Both last minute changes and long term reservations are possible. Unlimited configuration options are available: analog or ISDN connections, high-speed Internet connections, LAN and WAN structures as well as various bandwidths and fixed IP addresses.

Advanced hotel solutions for back office/administration areas In recent years communication systems have developed from traditional TDM models (Time Division Multiplexing) to “Converged Switches” for voice and data. Subsequently, secondary IP softswitches for local company networks based on VoIP appeared, which have now been superseded by central IP softswitches of the second generation via public networks with central application hosting. Like other vertical industries, the hotel industry today is changing from analog telephone technologies directly to IP telephony based on the Session Initation Protocol (SIP). The HiPath Hospitality Service Center was developed by Siemens as a completely new software application for the industry. This client/server application optimizes the telephony functions on a workstation in the hotel and provides comprehensive integration options to exploit the benefits of central data resources. This guest service center represents a conceptual breakthrough – it permits direct contact with guests and supports the intelligent monitoring of guest requirements from the moment they are conveyed up to their fulfillment. It represents a momentous step for the future in the integration of communication services and the management and monitoring of guest services. For that purpose, the HiPath Hospitality Service Center connects a guest

Strategic perspectives for the hotel industry

119

contact center with an intelligent module for service tracking and a data mining system. This is used for productivity control, employee disposition and creation of management scorecards. The service tracking module instantly informs the responsible hotel employees of guests’ wishes via DECT telephones, PDAs, PCs, etc. so they can immediately go about dealing with them. The guest and service data generated in the system are statistically evaluated by the data mining software and made available to the responsible parties. These data are not only accessible by the local hotel management in the Management Information System, but are also immediately available at the head office of the hotel chain as a real-time management tool, the so-called ‘Executive Information System’ (EIS). Instruments for the quality measurement of services rendered as well as a large selection of information and statistics on guest data and services will prove very helpful in the future to position brand value-related indices and meaningful key performance indicators – e.g. on guest satisfaction and the migration risk or the more efficient deployment of hotel personnel on the basis of peak period and productivity statistics. Its open architecture also lets the HiPath Hospitality Service Center support solutions from third-party providers for charge accounting, property management, terminals and third-party switches.

Hotel Palafitte: the future has arrived Lake dwellings have a long tradition in the 3 lakes region of Switzerland between Biel and Yverdon. Even in approximately 3000 BC, there was a series of farming communities in houses on the water of the Neuenburgersee. About 5000 years later this ancient tradition was revived again in the Swiss regional exhibition, Expo.02: 40 pavilions with 120 beds serve as comfortable guestrooms in the newly constructed 5 star Hotel Palafitte. 24 of these are on stakes driven into the bottom of the lake, thus appearing to float directly on the surface of the water. In contrast, the reception area, bar, restaurant and conference hall, kitchens and offices are located in the main building on the shore which is encircled by the remaining pavilions. The top class hotel, opened in April 2002, is a joint project of Siemens with the Sandoz foundation which invested approximately 20 million Francs in the project. As the prime contractor, Siemens equipped the “hotel of the future” in Monruz-Neuchâtel on Neuenburgersee with the most modern technology in the areas of information and communication, building

120

Michael-W. Hartmann

management and security and lighting. The trigger for this commitment was the challenge of demonstrating the complete solution approach incorporating technically pioneering achievements and participating in a hotel concept of the next generation. In accordance with the guidelines of Expo.02, particular value was placed on ecological construction during the planning of the five star hotel and the project was designed with the use of regenerative energy sources such as geothermal power and solar radiation. Each pavilion has its own ventilation system with warm water regeneration so that each guest can adjust the temperature according to their own requirements and can create a personal “feeling of well-being”. New paths were also forged in the lighting system of the Palafitte. Novel light directing technology, the use of light diodes behind a floating sail in the restaurant and intelligent lights coupled with movement sensors ensure economical power consumption and balanced illumination. A biometric access control system contributes greatly to the security of hotel guests. When the guest checks in, a fingerprint is briefly taken and stored in a computer. Instead of opening the door of his pavilion with a key or card, the registered guest now merely holds his finger under a reading device to gain entry. Unauthorized intruders have no opportunity of gaining access to the room with a skeleton key or similar. An interactive fire alarm system also contributes to safety. Its configurable algorithms are precisely matched to the specific fire risks in each area of the hotel complex which incorporates a great deal of wood. The communication technology backbone of the complete infrastructure in the “hotel of the future” is a Gigabit Ethernet which transfers data with the aid of the Internet protocol. The individual pavilions are connected to this fiber optic backbone with a bandwidth of 10 Mbit/s. The local network (LAN) allows guests to use the high-speed Internet access available in each room on a PC equipped with a color printer. Guests also have access to games, the hotel’s own service and infotainment portal as well as other applications. In addition, the functional scope of the web-based portal includes the control functions of the facility management such as lighting, heating or blinds. Clients who do not wish to operate the applications directly from the PC with the keyboard and mouse can use the SIMpad (Siemens Multimedia). This wireless operating device is approximately the same weight and size as a book and is available to guests in each pavilion of the Hotel Palafitte. It has an 8.4” color display and input is made using a virtual keyboard or by means of character recognition directly on the touch-sensitive touch

Strategic perspectives for the hotel industry

121

screen. All functions of the hotel technology can be controlled via the Web portal. The top view of the respective pavilion is shown on the display and the guest can intuitively operate individual elements simply by touching the display. The SIMpad is connected wirelessly with the local network via a Wireless LAN access point in each pavilion which, consequently, can also be used outside of the room. Thanks to the broadband Internet access, supplying individual guest apartments with video or music-on-demand from the hotel’s own network is technically no problem. A Dolby 5.1 Surround Sound system, a 48” plasma TV screen and a DVD player in each pavilion ensure acceptable replay of the media. All rooms in the Hotel Palafitte are equipped with a modern ISDN fixed network telephone and a cordless telephone. Next to the bed and in the bathroom is a desktop cradle for the cordless phone which works with the DECT standard (Digital Enhanced Cordless Telecommunications). For example, while the guest with the cordless telephone on the balcony two meters above the lake orders a drink from the hotel staff, a completely “normal” external telephone call can be conducted on the other phone. Telecommunication in the Hotel Palafitte is based on the HiPath 3000 real-time platform that uses the existing data network for voice transmission. Connection to the hotel software, which provides a range of features, is made via a standardized communication interface: These include the saving and transfer of call data for further charging, switching authorization of the guest telephone at check-in/checkout, the entry and deletion of the name of the guest, activation of the message waiting lamp on the guest telephone, the keeping of an alarm list as the basis for the automatic alarm service, the registration and carrying out of “Please do not disturb” wishes (hold all calls), registration of the room status, personal registration and deregistration, input of special services and the administration of the various room connections. When guests check in, their mother tongue is noted and transferred to the communication system. This enables all devices, operating systems and means of communication to be changed automatically to the respective language. As well as the alarm function, this includes operating instructions for the telephone and the personal voice mailbox. If telephone calls arrive for guests during their absence, this is recorded by the Voicemail system and signaled when they enter their rooms. The complete solution also integrates the digital alarm and communication server (HiPath DAKS) that enables alarm, conference or announcement systems to be set up without additional cost. Among other things, the

122

Michael-W. Hartmann

DAKS ensures that, when a call is made to the room, both telephones always ring simultaneously. The occupant can then select the telephone on which he wishes to take the call. Furthermore, the DAKS allows alarms to be issued to targeted sections of the hotel in cases of emergency. If there is a potential fire alarm for example, it is possible to simultaneously inform all pavilions and request the occupants to leave the rooms. If the alarm call is particularly urgent, it will also interrupt any ongoing communication. If the subscriber does not answer, the procedure will be logged and repeated until a reaction is obtained or the hotel personnel themselves can intervene. Naturally, technological innovations are not just available for guests in the pavilions. Even on entering the building, the interactive SIVIT system (Siemens Virtual Touchscreen) provides an overview of the installations in the hotel and their elements at the reception. The Gestik computer functions without monitor, mouse or keyboard. All control elements and displays are available exclusively as a virtual projection on a flat surface. The hotel information system is operated without contact simply by moving a finger within the light cone of this projection surface. A small infrared camera captures the movements and a computer in the backdrop converts this into a command. Consequently, the guest can quickly and simply call up the necessary information regarding the hotel and its environment. This principle of user-friendly man-machine interfaces has been implemented as widely as possible throughout the entire hotel. A integrative service concept, specially adapted to the requirements of the guests, is realized in all applications. “Simplicity and comfort” is the maxim. This approach meets with great acceptance by visitors. Not without reason can the hotel management celebrate the above average utilization of this exceptional residence – even after the end of Expo.02.

III. Technological networking

Real-time communication is the foundation on which real-time enterprises, more efficient productive business processes and personalized communication are being built. The Internet protocol is the enabling technology and its potential for future innovation is far from exhausted.

Bernd Kuhlin

Designing more productive business processes with convergent networks

Optimizing the integration of real-time communication into existing business processes is one of the biggest challenges of the next few years. This step brings together today’s disparate communication media using Second Generation IP (2gIP) technology in order to provide new solutions. This will enable users to exploit the full diversity of their communication channels, i.e. to employ the feature-rich functionality of innovative solutions via standard, but customizable open platforms. The abolition of the artificial borders between operators, enterprises and home networks results in a unified communication experience that is independent of physical locations and devices used. The resulting activation of untapped potential for increased productivity generates the homogeneous communication landscape that will ultimately lead to a dramatic reduction in the expenditure of time and money as well as an appreciable increase in quality of life.

Introduction Whether viewed locally within the company or globally, the business world has changed radically in recent years. Information and communication technologies have played a significant and increased role: data processing and new media have enabled innovative business processes, which places increased demands on the basic technologies. This was intensified by the development of networked data and communication installations, which opened the path to a globally distributed economic system. On the communication side, companies began in the mid 80’s to network telephone systems between the different company locations. Corporate networks that supported voice features throughout the entire network were created and, as a result, users had the perception that there was a single communications system.

146

Bernd Kuhlin

The QSIG protocol standard specified by the ECMA (European Computer Manufacturer Ass.) enabled the networking of systems from different manufacturers with full feature consistency, e.g. name display, call forwarding, call back. The early connections between computers and communication systems were made using CTI functions (Computer Telephony Integration). Various CTI applications improved company procedures. They included simple dialing aids as well as so-called ‘screen-pops’, which displayed a range of information on the caller before the call was accepted. For telemarketing activities in Call Centers, the processing of call lists was steered by the server and, for a defined course of interviews, simultaneously linked with the respective interview masks. In the mid 90’s mobile radio technology revolutionized the communication world, making it possible to work outside the office and still have access to company information. This development was the start of today’s mobility paradigm. At that time different networks and communications protocols were employed and this limited the usage of services between networks. Voice traffic was transmitted over circuit switched networks, which used a number of different protocols. Data, in contrast, was transferred as small packets over packet-switched corporate and public networks. With the emergence of the ATM networks (Asynchronous Transfer Mode), an attempt was made to overcome this separation, but ATM was not able to prevail for various reasons. At the end of the 90’s the first communication systems that transmitted voice over packet-oriented networks using the Internet Protocol (IP) appeared. This development was driven by the success of the Internet and Internet technologies.

Designing more productive business processes with convergent networks Today it is clear that the convergence of voice and data networks has laid the basis for innovative and improved communication solutions. Robust, feature-rich solutions are being marketed and implemented, and further groundbreaking applications for the optimization of processes in the “RealTime Enterprise” are being developed and several are currently available. They include: networking terminals, services and applications; integrating voice and video communication with IM (Instant Messaging); the embedding real-time communication in company portals; and, finally, the integra-

Designing more productive business processes with convergent networks

147

tion of communication options in business processes and workflows. The central elements of this new development are presented below.

Modern communication networks

Converged campus networks

First-generation IP technology (1gIP) enabled new communications topologies. The concept is simple: telephone signals are digitized and converted into the IP format for transmission as data packets. The resulting Voice over IP (VoIP) packets are then converted back into voice signals at the other end. Today’s voice communication systems can be “IP enabled” by adding VoIP gateways. This allows the IP-based terminals to be connected directly to the communication server via a Local Area Network/ Wide Area Network (LAN/WAN). In addition, complete systems can be networked over an IP-based infrastructure. The migration to native IP-based communication systems is an essential precondition for incorporating new applications on these networks. Seamless embedding real-time communications functionality in business processes results in an increase of productivity within the company (details follow). Nevertheless, implementing IP technology already provides clear advantages. VoIP allows companies to consolidate voice and data lines, resulting in savings in connection costs as well as in other areas, e.g. network management. This applies especially to the relocation of employees (moves and changes). Previously companies used leased lines to establish voice communication between different locations via Time Division Multiplex (TDM) channels. This architecture was very reliable, but inflexible with regard to bandwidth. For example, connecting new locations often required additional lines or complicated changes to the network topology. IP connections allow a Virtual Private Network (VPN) to be set up and operated, the key benefits being flexibility and efficiency. Voice and data traffic no longer requires separate connections and networks; both media types can be transported on the same IP-based infrastructure. Small branch offices or field staff can be connected directly to corporate resources via the WAN – without the need for their own local communication system. Thus, remote users can employ the same features and applications as the headquarters. In the case of bottlenecks or faults in the IP connection, intelligent routing management allows calls to be rerouted via the PSTN (Pub-

148

Bernd Kuhlin

lic Switched Telephone Network). The use of IP technology brings a number of additional benefits: communication systems, once monolithic blocks having proprietary protocols, become modular and are based on industry standards. The logical functions of call control and gateway function for steering media streams (e.g. voice, video) can be realized as independent network components. At the same time, systems are becoming more open. The use of standard protocols (e.g. H.323, H.248, SIP – Session Initiation Protocol) and open interfaces means that the systems of different manufacturers can be interconnected. A clear trend is evident: communication systems are increasingly using open platforms and data-centric technology. This expands user choice: components can be selected from different vendors, integrated and expanded over time. Consequently, the quality of a communication solution lies (as far as the use and value added are concerned) in the software and in the services. The hardware and the operating system are based on de jure or de facto standards such as Linux or Windows. The open application environment then allows applications from selected manufacturers to be quickly implemented in the company’s business processes. Communication networks can also comprise a combination of circuitswitched, converged TDM/IP and native IP systems. Traditional terminals continue to be connected to circuit switched interfaces, while IP telephones access the communication server via LAN switches. This guarantees full interoperability between the different terminals as well as the seamless transfer between telephones and software-based telephony applications. Virtual Private Networks (VPNs)

Voice communication in the public sector is changing at an accelerating rate: instead of dedicated (own or leased) lines that can only be used by a single company, VPN technology allows several companies to ‘share’ the same infrastructure. To guarantee security and quality of service (QoS) special measures must be taken. VPNs are predicated on the concept of being able to use the common public infrastructure as if it were a private network. At the same time, the same functionality should be available as for private dedicated lines, but at a considerably lower price and with the same level of security. The applications of IP VPNs initially gave priority to the connection of small branch offices via WAN connections. The different types of data and communication devices connect to the head office via the VPN, and Service Level Agreements (SLAs) with the relevant network operator ensure

Designing more productive business processes with convergent networks

149

the necessary bandwidth as well as security and QoS. VPNs allow the inclusion of home and mobile employees: they connect via xDSL or wireless access interfaces (WiFi at locations with high traffic volume or 3G). In this way, if there is sufficient bandwidth, end users can employ the same services independent of the access technology. Today many companies want to use VPNs for long-distance intranets and extranets, i.e. replace traditional access methods such as Frame Relay or ATM. Security requirements involve being able to guarantee the authenticity, integrity and confidentiality of the information. This task is undertaken by tunnel protocols and various encryption algorithms. The main tunnel protocols are IPSec, L2TP and PPTP. The IETF standard, IPSec (Internet Engineering Task Force), comprises a collection of protocols that also encompass security mechanisms. But there are often special adaptations required in the routers to prevent a reduction of performance and to support address translation (NAT, Network Address Translation) for network transfers. The encryption of the information depends on the use of algorithms that are based on international standards that can only be cracked – if at all – with considerable resources. With a key length of 256 bits, the Advanced Encryption Standard (AES) is one of today’s most secure processes and also offers a considerably higher level of performance than its predecessor 3DES (Data Encryption Standard). In addition, the inclusion of a firewall in VPN solutions is important in order to control access and to indicate attempts to gain unauthorized intrusion. Not only can VPNs realize private data connections, they can also carry out the simultaneous high quality transfer of voice or video traffic (i.e. realtime communication). Many different types of data and communication devices can be connected with one another via a LAN/Wireless LAN. The necessary QoS and security functions are provided by IP edge devices having elements for VPN, routing and firewall. The VPN tunnel makes all services in the data processing center visible and accessible to the user. Wireless LAN

Wireless LANs (WLANs) were originally introduced to facilitate data communications in departments and workgroups as well as specific locations such as warehouses. In 1999, the IEEE (Institute of Electrical and Electronics Engineers) defined the first 802.11 (WiFi) standards for secure information transfer. Once a standard was established many organizations recognized the new application possibilities, e.g. flexible and fast file exchange as well as

150

Bernd Kuhlin

mobile access to email, Internet or intranet information. The cost of Access Points and PC cards was relatively low and the installations simple and later on WiFi access became a standard feature of notebook PCs. Employees recognized the benefits that came from wireless access and it is no exaggeration to say that WLANs are (and continue to be) a user-driven development. WLANs are also used in the home and in so-called ‘Hotspots’. Hotspots allow mobile professionals to access the Internet in locations such as airports, conference centers, hotels and even cafés (Starbucks). Wireless cellular networks (2.5G/3G) incorporate IP technology and packet switching. This allows users to stay online while mobile, i.e. it is no longer restricted to the office environment. In addition, users now require a seamless transition between different locations, e.g. WLAN to cellular to hot spot to home and vice versa. This indicates that wireless access is something we are starting to take for granted: it’s the new norm. However, IT managers and CIOs have some valid concerns about end-user installations as well the use of hot spots. Security is the key issue. Although standards that address security and other issues are available, those concerns have not gone away, but at the same time users are demanding higher access speeds and seamless mobility, i.e. the data equivalent of GSM’s “anywhere, anytime” service. WEP (Wireless Equivalent Privacy) was the initial security standard, but it was off by default and relatively easy to decode when enabled. Products that support the IEEE 802.11i standard are now being marketed. In addition, there is the WPA (Wired Protected Access) standard that was specified by the Wi-Fi alliance. This contains substantial parts of 802.11i and provides very robust data encryption and it has been widely used as an intermediate solution. The expansion of WLANs over corporate and other campuses has developed in a similar way to that of regular wireline. During the 80s the high data volume between different nodes (e.g. PCs) became a problem. Although networks were subsequently partitioned by switches, the size of Layer 2 sub networks nevertheless remained restricted to around 200 connections. This limitation was only removed through the introduction of routers and Virtual LANs (VLANs), which led to a further growth of the LANs (often unstructured). Once again, this reflects the pragmatic advancement of the computer industry: Problems are quickly solved through proprietary technologies, standards are established later and are then implemented retroactively. The development aimed at the widespread provision of WLANs is proceeding in the same way as the setup of networks within the mobile

Designing more productive business processes with convergent networks

151

telecommunication industry. Each Access Point (AP) forms a wireless cell environment and the APs that are spread across the campus form a wide area wireless network. To provide interruption-free operation, so-called roaming and handover functions are required as well as an adequate number of APs. Also, the individual cells in the network must overlap to guarantee roaming, i.e. provide uniform accessibility in different cells. A user can, for example, download a large file in the cell environment of his/her office and walk into a conference room that is covered by another cell and the download will continue. There is a brief handover interruption, typically 400–500 ms, but this is not a problem for data transmission (unlike voice). The connection is not lost during this handover process as long as the user stays in the same sub network. New IP addresses are normally required at the boundary to other subnets and different techniques can be employed to retain the connection when moving between sub nets. The IETF “Mobile IP” standard provides a process whereby the endpoints always retain their IP address during the transition to another network domain. However, to date this layer 3 protocol has not been widely used. The set-up of “home agents” and “third party agents”, which work as proxies in the respective cells, is complex. Another technology is based on implementing a layer 2 overlay network over the different sub nets and then to address all endpoints from a logical point of view as if they were in a single sub network. This process is considerably simpler and also has the advantage that the delay from the conversion of IP addresses and re-authentication does not arise, i.e. there is no conversion. This process, which is known as “Layer 2 tunneling”, enables voice over WLAN networks. Dedicated switches are now available which enable interruption-free roaming for voice communication via IP. This brief overview shows that WLANs have different engineering requirements to regular LANs and this is particularly important when used for real-time communication. WLANs continue to be a growth market in both the consumer and the business sector. The first Access Points were designed as stand-alone products, so the natural development of propagation and associated higher scaling requires specific network management functions. In certain industries, however, there is such a strong interest in this technology that there is a willingness to adopt intermediate solutions. However, companies will first have to engineer their infrastructure, i.e. prepare the data network for Voice over IP (VoIP). Networks require the requisite bandwidth, the desired QoS, and voice traffic must be prioritized. Once in place,

152

Bernd Kuhlin

the implementation of a WLAN infrastructure can proceed. The following figure shows a highly scalable WLAN switch connecting the Access Points on the basis of a Layer 2 network. Today’s scalability is normally sufficient for the support of up to 30,000 users and 4,000 Access Points with no additional software required by the client devices. Interruption-free cell transition is even possible at speeds up to 100 km/h. The administration of the individual Access Points is carried out using VLAN technology, whereby the entire WLAN network is addressed as a single, flat Layer 2 network.

DHCP Server

Ethernet Switch

Access Control Access Points

Ethernet Switch

WLAN Switch Router

Internet

NAT/ Firewall

Ethernet Switch

LAN

Location Server

Communications System

Figure 30: Set-up of a WLAN network

Mobility is a key element for the improvement of process flows and a resulting increase in productivity. The transmission rates of WLANs are at least an order of magnitude higher than those of cellular networks. WLAN technology is therefore developing increasingly into an integral part of corporate infrastructures and by the end of 2005 telephony will even become an important driver for the installation of WLANs. Hospitals represent a typical scenario since GSM is prohibited. In this case the convergence of data and telephony networks creates an innovative real-time application: not only does voice run over the same (WLAN) network as the internal data administration, but doctors and nursing staff are

Designing more productive business processes with convergent networks

153

also able to access this data directly and immediately using a Tablet PC or PDA at the patient’s bedside. Thus, admittance data or laboratory values need not be collected manually, but can be entered directly into the system where they are taken and requested where they are needed. This leads to an enormous increase of productivity. In this example, the importance of the term real-time operation is clear: access to information or communication can be carried out at any location and at any time.

Communication networks of the future In their real-time communication strategy, many companies pursue the declared objective of reducing CAPEX and OPEX by at least 20%. This can be achieved, but only if a number of different measures are taken, e.g. consolidation of network operator and service contracts, implementing global helpdesks, network consolidation and the introduction of VoIP. However, it is difficult to achieve with first generation IP technology (1gIP). Second generation IP systems (2gIP)

Second generation IP communication systems will change the existing model in every way and also exploit IP’s intrinsic functionality in next generation networks. The implementation of communication services tracks the data model, i.e. communication services become data applications. Thus, real-time communication is integrated in data domains. The new model can be characterized by three concise expressions: centralized operation, conformity to norms and integration in workflows. Centralized operation

As soon as company-wide IT networks have been engineered for converged services, all IT applications and communication services can be located at any point in the network. This means that the capital, operation and administrative costs can be reduced by an even greater extent than with 1gIP systems. Consequently, the incentive is enormous for large companies to develop strategies for the centralization of communication services, including telephony, and to integrate them in their data centers.

154

Bernd Kuhlin

Norm conformity

Companies and Service Providers who adopt centralized IP communication services need standardized solutions for the servers and operating systems. Systems must be scalable, capable of handling the needs of up to 100,000 users, and they must interoperate with the products of other suppliers at both the equipment and application level. Compliance with open standards is a precondition. (Session Initiation Protocol (SIP) has become a particularly important IP communications standard. The SIP standard is currently being defined by the Internet Engineering Task Force (IETF) and it follows the Web paradigm. SIP looks set to become the foremost communication protocol that is accepted by both network operators and the business community. Its strength lies in the simultaneous support of telephony, video and data services such as Instant Messaging and Collaboration. SIP comes from the IT operator world; it is lightweight and employs relatively few commands when compared to the H.323 protocol, which comes from the telecommunication sector. This means that SIP is much faster and more flexible. Integration in the workflow

Large companies already recognize the benefits of the SIP protocol. Its integrated functionality means that it will impact on the way we will communicate with one another in future. New applications will accelerate the transition to a new communication paradigm. Data and communication services will not only be combined, they will also be seamlessly integrated within the user’s workflow. In a future that is almost here, multimedia services will be used via Web interfaces, thereby becoming as easy to access as the telephone service. In addition, mobile users and home workers will be tightly integrated and able to access the same applications as their colleagues in the office. If a customer calls a company, for example, he/she quite rightly expects the query to be answered quickly and competently. But nowadays long waiting times or follow-up calls can lead to high costs and dissatisfaction. On the other hand, in a central call center, all available information on customers and the right dialog partner for answering the query can be provided. Real-time communications software allows the availability of the relevant individual to be displayed on the agent’s workstation and enables calls to be forwarded immediately. In addition, all relevant information on the caller and the query can be sent at the same time. It is clear that this process

Designing more productive business processes with convergent networks

155

will improve agency productivity, but more important, the ability to deliver a better service to customers and prospects gives the company a significant competitive advantage. Implementation models

There are two different business models for the implementation of these services: 1) Companies operate all data and communication services in their own data processing centers 2) The service is outsourced, completely or partially, to a network operator or service provider. There is one aspect that is critically important for both models: the centralized provision of data and communication services in a single network. This is the only way that the cost reduction mentioned in the introduction to this section can be achieved. Increases in productivity are achieved in the long term using the new, presence-based applications. So, what is the impact on installed IP PBX systems? The vendors of these systems clearly need to protect their customers. If a business wishes to proceed step by step, it can start by implementing softswitches in the regional data processing centers and in that way centralize the administration of all users, applications and gateways in a region. And some of the installed IP PBX systems can continue to be used as gateways to the PSTN. The number and positioning of these gateways depends primarily on the respective network technology, the degree of network availability and the effects on the dial plan (direct dialing numbers). The main focus of highly scalable softswitches comes from innovative services. Presence and Instant Messaging are baseline services. In addition, various multimedia and multimodal applications can be created based on standard APIs. Automatic voice detection technologies allows multimodal applications to combine voice and data inputs from different sources (wireline/wireless telephones, PCs) in order to access corporate data. This means that users can work transparently via a service/devices mix, e.g. data requested by voice can be transferred to the Web browser of a PDA. The current technology trend for Web services and standards such as XML, voice XML, SALT and XML CSTA, must be reflected in these solutions. Only through open and standardized interfaces can the optimum integration in business applications be realized.

156

Bernd Kuhlin

Mobile employee Data center

Data center IP

Applications

Data center

Applications

IP Applications

Data center Branch offices Communication system Survivable media gateways Standard voiceenabled router

Figure 31: Real-time IP systems integrated in data centers provide centralized telephony features and applications thus lowering capital, operational and management expenses Collaborative applications

Innovative software tools that permit a common view of the different communication and collaboration systems in the company can be hosted on 2gIP communication platforms. End users have a personal portal for access to email, calendar, and IM as well as presence-based telephony and voice/video conferencing. As described in the previous section, the SIP protocol is a key component of these applications. Presence-aware communication comes from Instant Messaging: IM transmits short text messages in de facto real time, although technically it’s an asynchronous medium. Unlike email or SMS communication, IM lets authorized parties see who is online and therefore available. Software tools allow presence to be managed. Users can select IM-type messages to indicate their status (e.g. “Busy”, “Be Right Back”, “Away”). Since telephony presence (phone on/off hook) will be displayed as an icon there is no need to select “On the Phone”. They can also see the presence status of selected colleagues. ‘Buddy lists’, similar to those of IM programs

Designing more productive business processes with convergent networks

157

such as MSN Messenger, show who can be reached via which communications medium. The important point here is that the presence information of individuals and workgroups is not linked to a terminal type or an application. The portal contains click-to-call icons that establish telephony or video connections, IM and email. Clicking on the relevant names or the workgroup icon sets up the call. Another key benefit is the ability to have a single ‘virtual’ number. Callers to this number will be directed automatically to the relevant terminal, e.g. if the called party has set his/her status to “Away” then the system will ring the mobile phone. Given the number of communication alternatives it would seem that technology has given us unlimited accessibility, but this is not the case. Email, for example, is asynchronous (not real time) while telephony is a real-time (synchronous) medium. Because of the difficulty of reaching parties by phone, email is often used as an alternative to a real-time communication. This medium may also be used if the caller is concerned about disturbing the called party. On the other hand, if the call is urgent the calling party may ring the mobile after leaving voicemail and then switch to another medium. These problems don’t arise if one can see the presence status of the other party, e.g. if the status is “Be Right Back” then the call can be delayed unless it is urgent. Presence-aware applications are therefore enabling a new communications model, one that aggregates when and via which media a person can be reached. The application also gives users control over who can call and be put through and who will be diverted to voicemail. Moreover, to protect personal privacy, users can specify to whom their presence should be transparent. Video

Video communication is a medium that will be employed more and more in future. Consider the advantages of telephony: in addition to the actual message, the pitch, volume, speed and variation of the human voice conveys subtle information such as emotion and mood. This feature-rich medium is ideal for discussing difficult and controversial subjects. In recent years, however, email has become the principal communication medium in the business world, primarily due to its cost, efficiency and effectiveness. Email is ideal for communicating over different time zones and for transferring files (attachments). But email is not feature rich; emotions cannot be conveyed and all too often messages are misunderstood.

158

Bernd Kuhlin

That’s why the combination of a “human” real-time medium with electronic enhancements is required. As a real-time communication medium, IM complements both email and telephony. For example, a communication may start with IM and then switch to voice simply by clicking on the adjacent telephony icon. The addition of the visual element is an additional and important complement to the communication process: it’s the electronic equivalent of a face-to-face discussion. Email is ideal for routine work and for processes where decisions can be made without discussion. Today, if text-based communication of the task is no longer adequate, we reach for the phone.

Figure 32: Video over IP is the logical extension of real-time communications over IP

Video is now putting in an appearance on PC desktops. The concept is not new and various technologies have been employed in the past without significant success. However, just as voice becomes a data type in IP communication so does video. That’s why video is the next “quantum leap” in personal communications and this medium will be taken for granted in a few years. For this reason “Video over IP” is the next logical step for many companies. It uses the same LAN/WAN infrastructure, improves the ROI, and does not require a major upgrade to the current workstation. The informa-

Designing more productive business processes with convergent networks

159

tion and communication network must, however, be engineered in order to handle the additional data traffic. If this is done in the early planning phase – ideally before the changeover to IP telephony – the subsequent introduction of video communication will proceed smoothly. There are numerous application scenarios. From face-to-face communication and entire conferences in special video rooms to real-time emergency assistance during operations or catastrophes. Video allows experts to provide assistance via video links over long distances and to examine the situation in detail on their communication device. These and other scenarios are not only possible, they have already been tried and tested.

Unified user domains IP technology enables the centralization of services as well as the optimum integration of real-time media in data applications. The architecture of modern communication systems of the future in the “Next Generation Network” (NGN) must, however, go one step further. The ultimate objective is the unification of home, company and carrier networks, both wireline and wireless. Through the integration of home workplaces, small offices, branch offices, regional branches up to the company headquarters, this approach enables the creation of unified domains in both private and public networks. This also allows the network borders that were created by the different technologies to dissolve. The result is a unified end-user experience, whatever the network being accessed or the device being used. This experience means that people can perform their work more productively, even from home. At the same time, network unification gives carriers a strategic differentiator. There are a number of on-going initiatives that will facilitate this process, e.g. the Siemens Communications LifeWorks concept. In-depth research into the needs of companies and their customers indicates that there is an overwhelming need for communication applications that have a common, secure, seamless and customizable user interface. Similarly, a convenient and cost efficient access method is required for both personal and business applications and services. In recent years many companies have focused their resources on core competences and mainstream activities such as manufacturing have been outsourced. This has resulted in the creation of ecosystems, i.e. complex webs of inter-company transactions between customers, suppliers and other authorized third parties. Enabling secure, seamless communications

160

Bernd Kuhlin

between these parties is one of the goals of the LifeWorks concept. It also recognizes that we have become a mobile society. Many employees only work in an office some of the time: they may also work from home or on the move. Analysts estimate that over half the average workforce is mobile and the percentage is rising. In response to the increasing need for mobility, different vendors have developed a plethora of devices, applications and services that allow us to stay in day-to-day contact. On the device front we have mobile phones, office and home phones, PCs, PDAs, etc. It’s a long list. These devices meet our basic communication needs and their usage reflects the blurring of the boundaries that used to separate the business and personal sides of our lives. However, most of today’s communication services and applications are island solutions. We employ different technologies, tools and applications for different communication activities and locations, e.g. one or more company offices, a wireless environment while mobile, and a private communication environment at home. The new communication landscape must dissolve these borders. A typical example of the difficulties of business communication is the number of attempts which are often required to reach an employee for an urgent meeting or an important call. This is directly linked to mobility and the use of different devices on different networks. By projecting the costs associated with this process onto a whole company, and by adding the vast quantity of information that each individual has to process daily, the obvious question to ask is how can processes be improved in order to boost individual and workgroup productivity? There is a clear starting point for the optimization of processes and the associated increase in productivity, i.e. the ability to display and manage presence, which was covered in an earlier section. Once this functionality has been implemented the productivity benefits can be extended to mainstream applications and business processes. In order to achieve a significant improvement the company’s fundamental communication paradigms have to be re-evaluated. If we see a company as a collection of individuals who cross network boundaries and work, as described earlier, in different environments, then we begin to recognize what is needed. We need concepts like LifeWorks in order to eliminate the dilemma that comes from the current need to synchronize, integrate and manage the different communications tasks. Instead, there should be communications processes that are not dependent on the media, devices, applications and locations being used.

Designing more productive business processes with convergent networks

Standard domain

Communication system at the operator

161

Communication system in the company

Applications Operator network

Mobile employee

Home work place

Company network

Branch office

Company headquarters

Standard communication experience Integrated operator-conform quality and reliability

Figure 33: A unified network and applications domain allows users to communicate uniformly and end-to-end across different devices and media and, in particular, tears down the barriers between private and pubic networks (fixed and mobile networks)

The impact for the different target groups: • End users: Obstacles to communication and access to information in the office comes from the fact that the required task is not optimized for the location. However, LifeWorks controls the communication flow on the basis of the accessibility status of end users, thereby increasing their productivity at the workplace, when mobile, and in the home. • Companies: Profitability is the primary objective of companies. Achieving this objective is facilitated by the integration of personalized services such as conferencing and collaboration as well as the integration of real-time communication with IT applications. A key objective is to boost productivity by making tasks easier to perform and as a result mimimizing stress at the workplace. • Network operators: LifeWorks has a convincing and unique value proposition for network operators. The concept recognizes the need to provide VoIP and VPN as an integrated service to companies, selfemployed individuals, mobile employees and home workers. This indicates that Lifeworks gives carriers a distinctive competitive advantage and even more important, it allows them to offer profitable new services in short timeframes.

162

Bernd Kuhlin

When the concept is implemented as a system it allows 2gIP applications to scale to carrier levels. For an inter-company working team, the same presence and availability functions that are employed in the core enterprize can be extended to authorized partners. This means that a standard domain and a standard means of communication apply to all participants, whether they are on a private or public network. When the communication and collaboration paradigms are expanded in this way, business processes become more efficient . In addition, a noticeable reduction in transaction costs is attained via the efficient distribution of tasks to individual subcontractors. Outsourcing manufactured goods allows prices to be reduced, but a distributed process requires many more transactions than in-house production. This means transactions costs become a significant overhead: 50 % of the cost of the finished product is typical . Consequently, savings in this area have a much greater influence on prices than savings in production and distribution where only minor improvements can be realized. LifeWorks’ aim is to create a homogenous communication landscape with new technology and to enable a secure, seamless, personalized user interface. Regardless of when and where communication takes place, whether in a personal or business environment, it should always be a simple and efficient process. The process is based on modular communications software and middleware, which enables integration with the IT domain. Virtually nothing changes for users, except an incremental amount of new functionality that is added to legacy applications, e.g. Outlook. The use of an established and familiar interface and seamless access to information in real-time unlocks the intrinsic added value of the 2gIP for users and at the same time it boosts the company’s productivity and profitability.

Summary From the end of the 90s, telecommunications experienced an unprecedented expansion of voice and data services. Mobility, 24-hour availability, new convergent services based on a single infrastructure were the buzzwords. However, these innovations were conceived as island solutions and this resulted in a fragmented communication experience. Thus, the challenge for the next few years is considerable: on the one hand, users should be able to utilize the full range of communication functionality via innovative solutions. On the other hand, they should be given full control over the various

Designing more productive business processes with convergent networks

163

communication processes as well as the required access to information via a standard-based, personalized user experience. When IP telephony and real-time applications are integrated into business processes the proven result is a significant cost reduction. Savings in operation and maintenance costs are the first tangible results of hosted, real-time communication solutions. With its strategy of cost effective transition strategies, Siemens is able to offer highly scalable, 2gIP communication that makes a significant and positive impact on productivity and profitability.

Jörg Eberspächer, Josef Glasmann

QoS Architectures and Resource Management in the Intranet

The Internet has developed at an astonishing rate over the last few years. For a long time now, it has stopped being merely a network for pure data applications (eMail, FTP, WWW). New real-time applications that place increased demands on the IP infrastructure such as IP Telephony, Audio & Video Streaming, Video Conferencing and Network Games have become widespread in recent years. Collaboration, eCommerce and Tele-Education systems have also become prominent. These complex systems frequently include both data and time critical services where most real-time critical services (telephony, video conferencing, etc.) are still transacted via the conventional networks (Public Switched Telephone Networks, PSTN). In future, not only data traffic, but also real-time critical traffic will be transmitted via IP networks (Internet, Intranet). New demands on IP networks

Interactive communication services in IP networks, however, only gain customer acceptance if they offer quality (Quality of Service, QoS) that is at least comparable to that of the PSTN. For example, a PSTN user would rather have his call attempt blocked by the network than tolerate fluctuating connection quality or, even worse, have the connection break while in use. In particular, users are particularly sensitive to momentary fluctuations of connection quality during the transmission of voice data (e.g. IP telephony). High jitter values and packet loss have a direct effect on the listener as they reduce the fidelity of the call and the ability to understand what is being said. Similarly, if the overall delay of the voice transmission is too high, this will lead to the degradation of call quality. This places new demands on the IP infrastructure. In order to provide time-critical traffic with hard QoS guarantees and to thereby ensure con-

166

Jörg Eberspächer, Josef Glasmann

sistent connection quality, new mechanisms must be introduced to manage available network resources in such a way that the required end-to-end quality is attained. Pioneering role of the Intranet

Intranets are based on Internet technology, so they play a pioneering role for the Internet since the infrastructure required for the introduction of new network services is easier to accomplish in a managed environment. Normally an Intranet comprises an environment confined to a single administrative domain that uses a limited number of network technologies, network services and applications. The questions of migration, scalability and interoperator aspects are not as critical as for the operator of a public network. Moreover, the company can determine which network services and applications are made available to (internal) users. This means new services can be introduced more quickly and the resulting benefit for the company is therefore easier to calculate. Even today, most companies still have two separate networks for their telephone service and data applications. Both networks are made up of terminals, network nodes, servers, software and lines, all of which have to be installed, administrated, maintained and renewed or expanded when necessary. In particular, network and service maintenance is associated with high running costs. These costs can be considerably reduced by managing all services via a single network. A single network also allows easier and quicker introduction of services such as end-to-end QoS solutions and telecommunication and data services can be integrated more rapidly in the Intranet than in the “public” Internet. Also, because network transmission quality when used for communication services often plays a greater role in business processes than in the private sector, there is greater market pressure to ensure a high QoS in company networks. Today's rule : Best Effort – no QoS

Apart from adequate provisioning of the networks, the key QoS issue concerns the method used in the network nodes to forward the packets. If packets from different sources of a network node are conveyed to the same output at the same time (traffic aggregation), this usually causes congestion on the output line. The way that network nodes deal with these conflicts has a significant effect on the QoS of a network. To avoid losses, the packets are

QoS Architectures and Resource Management in the Intranet

167

temporarily stored in buffers and then read out after a delay in accordance with a specific schedule (scheduling). Traditionally, IP routers have only one buffer per output from which the packets are read out using the FIFO principle (first in, first out). Consequently, today's Internet offers only one service class, the so-called “Best Effort” service. This is based on the equal treatment of all packets. As a result, all active data streams (TCP/UDP) on a link interact and the achievable quality of service depends on the number of currently active subscribers and the traffic they generate. In addition, most applications transmit with variable packet rates (burst-mode traffic). Thus, even in adequately dimensioned networks, the probability that application requirements will briefly exceed network capacity grows with the number of active subscribers. This can result in lengthy packet delays and even losses. The aggregation of many packet streams having variable packet rates can also result in strongly fluctuating delays (jitter). Simply enlarging network capacities (overprovisioning) while still maintaining a reasonably economic framework will, in itself, not be sufficient to overcome the problem of jitter in the network. As mentioned earlier, time critical applications are particular sensitive to fluctuating delays or losses, so additional measures are required to ensure that IP networks offer a fast response. The network operator

Network operators normally balance a number of objectives. On the one hand, they need to operate the network as economically as possible; on the other they want to offer the user specific QoS guarantees. In principle, there are a number of measures available to a network operator in order to achieve these objectives. As well as replanning the entire network, traffic engineering and resource management can also be employed. The basic requirement for QoS is an adequately dimensioned network. Replanning and a corresponding network expansion are normally associated with higher costs. In order to defer these high investments for as long as possible, the available resources (in particular the transmission capacities) must be leveraged. Resource management attempts to prevent momentary overload situations (duration < 10ms) from occurring in the network nodes can be introduced. For example, by implementing measures for traffic checking and traffic control (Classifying, Admission Control, Policing), traffic engineering techniques can be employed that allows the network to operate at full capacity and as evenly as possible. Traffic engineering (QoS

168

Jörg Eberspächer, Josef Glasmann

Routing, Load Sharing, Multi Protocol Label Switching, MPLS, etc.) can be used to dynamically adapt the routes in the network to changing load ratios. Consequently, overload situations occurring in the medium term (typical duration < 1 day) can be avoided and the date for expanding the capacity delayed. Nevertheless, with steadily increasing network traffic, the load limit of the network will eventually be reached and it will not be possible to exploit new capacity reserves through traffic engineering. Redimensioning or expansion planning will then become inevitable. The Intranet

Intranets transport the “internal” and “external” IP traffic of the employee. “Internal” traffic describes the traffic within a company network, i.e. between the workplaces of the employees of a company. “External” traffic describes the outgoing traffic, i.e. in a public network. In small companies, the spatial distribution of the employees is frequently limited to one or more adjacent buildings. The entire internal traffic can therefore be carried out via a local company network, which can be operated by the company. . Larger companies may have several national or international locations. They cannot normally be connected with one another via their own network, but must be connected to public networks. Depending on the amount of traffic, these can be switched connections, fixed rented connections (Leased Lines) or virtual private networks (VPN). A VPN is the virtual network of a user group. It is created by virtually distributing the common resources of a public network to multiple user groups. If the network is a switching network, this distribution is achieved by allocating and implementing public and private addresses. IN technology (Intelligent Network, IN) is required for address implementation. For example, in IP networks a VPN can be realized using MPLS technology. MPLS is used for setting up permanent end-to-end paths (Label Switched Path, LSP) and for embedding (Encapsulation) private user data. With the aid of MPLS, public Internet traffic and private Intranet traffic can be processed via a single network. In order to connect local company networks to the VPN, service contracts must be negotiated with a public IP network operator. The terms of the contract, for example, can be based on the volume of traffic to be exchanged between the locations under fixed conditions. This necessitates precise specifications of the various types of traffic (voice, video, data) with regard to their traffic characteristics and the QoS requirements.

QoS Architectures and Resource Management in the Intranet

169

The operator of a VPN provides an access node for each company location. All company-internal location-spanning connections then run via this VPN. The underlying transport network structure of the VPN is usually transparent for the company. In the case of external connections in a deregulated market, there are no fixed contractual obligations to a single public network operator. A different public network operator can be selected each time a connection is set up.

Public IP/MPLS network

VPN

Access nodes

Local area network

Dedicated lines

PSTN / IN

Dial connections

Figure 34: Example scenario of an Intranet

Figure 34 shows a typical large company network. It consists of several locations, each with a local company network. In some cases, the local location networks of the company are connected by switched connections and dedicated lines via a public PSTN operator. In other cases they are connected directly to a public IP network operator that has set up a VPN for the company. Figure 35 is an example of a local location network that connects a number of buildings via an intra-site IP backbone network. The interoffice networking of workplace computers is carried out as a star-shaped configuration via a Switched Ethernet LAN (Local Area Network). A Switched Ethernet LAN is a network (see IEEE 802.3, 2002) where every subscriber is connected to one port of a switch (ISO Layer 2 network nodes) via its own dedicated line (Twisted Pair, Full-Duplex). This means each subscriber has a connection rate of 100 Mbit/s or 1 Gbit/s in both directions. The individual lines are combined into several levels: e.g. working group, department, floor and building. The principle of traffic classes is used in both the backbone network and the access network. The logical separation of real-time critical and non-critical data streams, which starts with

170

Jörg Eberspächer, Josef Glasmann

the subscriber lines, is a prerequisite, i.e. it is needed in order to offer service quality guarantees for data transport. At the subscriber access point, Layer 2 packets are classified as delineated in the IEEE 802.1 D standard and marked accordingly in the packet head. The building switch works on OSI Layer 3 and classifies and marks the IP packets according to the DiffServ standard. The external traffic of the location is mapped on MPLS paths at the access node to the VPN according to the DiffServ (DS) service classes (Codepoints) and the target location (VPN, Internet).

Internet, VPN IETF MPLS: “Classifying” “Marking”

Building D

Building F Building E

Building C IP core network

Building B

Building A

Building G Building H

10/100 Switched Ethernet

IETF-DiffServ: “Classifying” “Marking” IEEE 802.1 p: “Classifying” “Marking”

Figure 35: Scenario “Local location network”

Network operators that offer real-time services must be able to control the traffic in their network. A possible structure for this type of traffic control follows. First, an overview of the different QoS mechanisms is given in Section 2. The most important QoS architectures based on these mechanisms are then introduced in Section 3.

QoS Architectures and Resource Management in the Intranet

171

QoS mechanisms Subscriber take the quality of an “end-to-end” communication service seriously. This includes all stages involved in processing the original signal, which is done in accordance with the OSI layer model: User Plane. To maintain the QoS parameters, various QoS mechanisms are required in the network nodes and terminals. They are coordinated by a resource control system. Network architectures have a resource control system (OSI: Control Plane) and normally provide monitoring of the transmission quality actually achieved (OSI: Management Plane). This chapter starts with a short overview of the different aspects surrounding the topic of “QoS in IP networks”. The main focus “lies in new approaches for realizing end-to-end QoS”. The technologies in each of the individual areas provide concrete technical solutions, which are referred to as QoS mechanisms. These run on the terminals or in the network nodes. A single QoS mechanism can be used in isolation or in connection with others and constitutes a building brick of a QoS architecture (see Chapter 3). Terminals

This section describes QoS mechanisms that are used in terminals. These include the areas of user data processing (OSI: User Plane) and control (OSI: Control Plane). In the context of terminals, the QoS aspects are concentrated into the following main topics: • • • • •

Coding Task scheduling Synchronization Transport mechanisms Subscriber signaling

The coding process is concerned with techniques for data reduction (silence suppression, audio coding, video coding) that also allow multiple quality levels (layered coding). They can be flexibly adapted to the loading of the network and are not susceptible to transmission errors (adaptive delay and error control). In addition, consideration is given to scheduling processes. They form a part of the operating system and are therefore responsible for distributing the central resources (e.g. processor time) of a multimedia terminal to several application processes with different performance requirements. The

172

Jörg Eberspächer, Josef Glasmann

field of synchronization of time-critical media streams is closely associated with the coding of media streams on one hand and real-time-capable operating systems on the other. Synchronization in a distributed system must be carried out by the receiver of one or more media streams. It ensures that the coordination of packets within a media stream or in several independent media streams is maintained or restored. System and storage models, protocols and complete software architectures were developed for this purpose. The RTP protocol of the IETF has prevailed as a transport mechanism for the transmission of real-time traffic in IP networks. Besides user data, it contains information on user data content (e.g. media type, coding process), information for transport control (packet sequence number, time stamp) and a feedback mechanism (RTCP). Subscriber signaling in IP networks is composed of two processes decoupled from one another – service signaling (e.g. ITU H.323 Standard, IETF SIP Standard) and resource reservation (e.g. IETF RSVP Standard). Service signaling distributes the media streams and their QoS requirements among the terminals. Afterwards, resource reservation is used to allocate in the network those resources that are necessary for the transmission of the media streams. The RSVP (Resource Reservation Protocol) reservation protocol, for example, runs on both the subscriber network interface and in the network between the network nodes. IP networks

Control and suitable steering of network traffic are pivotal tasks for guaranteeing QoS. Not only do they comprise methods of traffic modeling, traffic steering and traffic control, but they are also mechanisms of the internal network signaling and traffic management. The developed processes run in the network nodes and concern the areas of packet processing (OSI: Data Plane), control (OSI: Control Plane) and network management (OSI: Management Plane). The main topics are: • Traffic modeling • Traffic steering • Traffic control Traffic Conditioning (Classifying, Policing, Marking, Shaping) Scheduling (service disciplines) Buffer management

QoS Architectures and Resource Management in the Intranet

173

• Access control • Reservation protocols • Configuration management Traffic modeling characterizes the behavior of voice sources (ON/OFF, MMPP), video sources (MMPP, ARMA, TES), data sources and the longterm dependent behavior of aggregated Ethernet traffic. An overview on the most frequently used traffic models is given by: Adas, A. 1997 and Rueda, A., Kinsner, W. 1996. The models are used to determine the buffer performance and bandwidth requirements of the traffic and provide the basis for the development of scheduling, access control and load-dependent traffic steering processes (Adaptive Routing) as well as of methods for planning and dimensioning networks. Novel traffic steering processes were developed in the traffic engineering sector in order to optimize traffic behavior, even during regular network operation. This enables greater utilization of network capacity and better control of network congestion. Based on utilization monitoring, the routing of data through the network is dynamically adapted. For example, the underlying mechanisms depend on routing that is based not merely on the target but also on the source and target, on an optimization of routes that are constrained (Constraint-Based Routing), and finally on the adjustment of routes via metric adaptation or route fixing (Route Pinning: e.g. MPLS). These processes are not described in detail. They do improve the load situation and the QoS, but they cannot give subscribers a QoS guarantee. The traffic control process takes in all QoS mechanisms that an IP packet can encounter on its way through a network node (Forwarding Process). They are normally implemented on the outputs of a network node and contain methods of Traffic Conditioning, Scheduling and Buffer Management. The term “Traffic Conditioning” was coined within the framework of the IETF standardization of IntServ and DiffServ QoS architectures. It comprises the processes of classifying and marking a packet as belonging to a specific IP flow, monitoring (Policing) of an IP flow regarding a preconfigured traffic profile and shaping of an IP flow to a specific traffic profile. These mechanisms are familiar from the standardization of the ATM and are now also used for IP networks. For a definition of the terminology: see Blake, S. et al. 1998. Scheduling processes were developed for broadband networks (ATM) at the beginning of the 90s. They determine how the transmission bandwidth of a network node output line is distributed across packets that arrive simultaneously. They also specify the sequence in which the buffered packets are

174

Jörg Eberspächer, Josef Glasmann

processed, thereby forming the basis of every QoS architecture. Scheduling processes normally manage a number of (mostly virtual) buffers and enable a prorated allocation of the available storage capacity and operating rate per buffer. They differ in how they distribute the operating time of the outgoing link on the individual buffers . To do so, they either dynamically adapt the operating rate of a traffic class to the current load situation (rate based) or they allocate time-dependent priorities (delay based). In addition, the dynamic behavior of a scheduler is distinguished by its handling of unused resources. Either they are prorated to the other traffic classes (work conserving) or they remain unused (non-work conserving). The latter certainly has disadvantages with regard to the achievable network utilization, but has advantages in the transmission of jitter-sensitive real-time traffic. For an overview of the best-known processes: see Zhang, H. 1995. Buffer management processes were specially developed for TCP traffic to control the buffer level in the network nodes. They ensure that packets are discarded in good time if there is congestion, i.e. even before the buffer overflows. The buffer level can only be controlled if the applications concerned detect the losses in the terminals and reduce their sending rate. Transport protocols for real-time traffic (RTP) and for data traffic (TCP) support this mechanism. Examples include: Early Packet Discard (EPD), Random Early Discard (RED) and Fair Random Early Discard (FRED). The access control process differentiates between measurement-based and parameter-based processes. In order to enable robust QoS guarantees, an explicit reservation of network resources in the network nodes is necessary. This requires corresponding protocols and access control processes. A receiver-oriented reservation process (RSVP) was standardized for the Internet. An aggregation of individual reservation messages is recommended in RFC 3175. There are numerous other approaches that reduce the signaling traffic and are therefore easier to scale. In order to reduce the status information in a multidomain environment, a BGRP reservation process based on DiffServ networks and BGP (Border Gateway Protocol) was developed in: (Pan, P. et al. 2000). It aggregates the reservation states with regard to common target domains (Sink-Tree-Based). In the NSIS (Next Steps in Signaling) working group of the IETF, efforts are already being made to produce a unified transport mechanism for different reservation processes. For this purpose, the “Transport” and “Message Processing” functions are separated (Signaling Layer, Transport Layer). The introduction of QoS mechanisms in IP networks also gave rise to the requirement for dynamically steering the configuration of the network ele-

QoS Architectures and Resource Management in the Intranet

175

ments. This means that the allocation of resources (e.g. setting up the traffic classes) is not a one-off process, but can be dynamically adapted to the current requirements of the subscribers. Furthermore, subscriber access to these resources must be steered and monitored through the configuration of classifying and filter units. Also, an ever-increasing diversity of different device types which support different QoS mechanisms necessitates a new management approach. The IETF has selected two approaches that address this complexity issue and the associated high management cost . One draft was designed in the SNMPCONF working group and is based on the SNMP concept. The second approach was devised in the RAP (Resource Allocation Protocol) working group and is based on the COPS (Common Open Policy Service) protocol. Both are based on the model of a central server and several clients. The configuration of the network is described in a standardized form (Policy = set of rules) by the administration and is assigned to the server. For the specification of such policies, a corresponding SPPI (Structure of Policy Provisioning Information) syntax was defined based on SMI (Structure Management Information), and a PIB (Policy Information Base) data structure was defined in compliance with the MIB (Management Information Base). The server distributes the network-wide valid PIB objects across the clients. The clients map them onto their locally available QoS mechanisms and attend to the implementation of the policy rules in the network nodes.

QoS Architectures The control of complex multimedia services as well as the processing of voice and video signals in real time places significant requirements on the terminals and the network. In order to meet these requirements several service architectures, e.g. H.323/SIP, ICEBERG and TINA, were developed for controlling services, and several network architectures, e.g. ATM and Internet (IntServ, DiffServ), were developed for controlling networks. For the sake of simplicity, it will be assumed that the tasks are allocated to the service architecture and network architecture as follows: • Service architectures: These architectures are used for negotiating service features between terminals. They determine the number and type of connections (medium, coding process) and the minimum level of transmission quality (QoS parameter) from the point of view of the sub-

176

Jörg Eberspächer, Josef Glasmann

scriber. • Network architectures: These architectures administrate network resources and make sure that the transmission quality required by the subscribers is guaranteed for the duration of the connection. Therefore, they are also referred to as QoS architectures. The preceding chapter gave a brief overview of the different QoS mechanisms in the terminals and in the network. In the case of a QoS architecture several of these QoS mechanisms are embedded in a unified higher order system context and coordinated by a system control. In the recent past, various QoS architectures was drafted in order to improve the quality compared to classical IP networks (Best Effort). The QoS architectures can be subdivided into two classes having different aims: • Class A: These approaches primarily attempt to realize a transmission service which, similar to a connection-oriented network (ATM, TDM), enables robust QoS guarantees such as the IntServ (Section 3.1), RRA (Section 3.3) and SCORE (both in Section 3.4) architectures. • Class B: Other approaches attempt mainly to follow the philosophy of the Internet and to realize the best possible transmission service on the basis of a simple and robust architecture. A typical example is the DiffServ architecture (Section 3.2), but it cannot warrant a QoS guarantee. Other approaches such as the Bandwidth Broker (Section 3.3) and AQUILA, Egress AC and Endpoint AC (all Section 3.4) implement additional control functions and can warrant (relatively weak) QoS guarantees. Integrated Services Architecture (IntServ)

IntServ is a connection-oriented process that enables the reservation of network resources end-to-end for each communication connection (IP data stream). Reservation requires the implementation of protocol finite state machines in the routers and terminals. The IETF recommends the RSVP (Resource Reservation Protocol) as a reservation protocol. It is designed for sessions with many subscribers and is based on soft states. This latter property requires that reservation states in the router be updated by periodically sending so-called refresh messages. The routers must process incoming user data packets as well as signaling messages. The required resources must be calculated and provided for

QoS Architectures and Resource Management in the Intranet

177

each reservation process. In contrast to conventional routers, additional mechanisms such as Classifying (per packet), Policing (per data stream) and Shaping (per data stream) are run through during packet processing. Note that buffer management and scheduling are considerably more complex because a separate buffer must be administrated and operated for each data stream, as specified in the reservation. The multitude of processing steps per packet, and the fact that more than one million IP data streams can run simultaneously over a router, especially in larger backbone networks, places high demands on router performance. In addition to the complexity of buffer management when using RSVP, the quantity of signaling messages also leads to scaling problems. A message flood can result form the fact that for each individual media stream of a multimedia session, a separate RSVP signaling channel is required. In addition, periodic refresh messages must be sent. In Baker, F. et al. 2001, a mechanism was standardized for the aggregation of reservation states (State Reduction) at the domain borders. App. App.

Source

Drain

IntServ-Router (classifying, policing, shaping of individual flows)

Signalization and data route

Figure 36: Example network of an IntServ architecture Differentiated Services Architecture (DiffServ)

DiffServ was developed in 1998 by the IETF in order to create a robust, simple and scalable QoS architecture for IP backbones. The architecture is based on the concept of service classes and a relative interpretation of service quality as opposed to absolute quality guarantees. DiffServ provides a set of concepts and QoS mechanisms that enable the realization of different

178

Jörg Eberspächer, Josef Glasmann

transmission services. The standard only contains the description of the behavior of an individual node (PHB: Per Hop Behavior) and no specification of the end-to-end achievable service quality (PDB: Per Domain Behavior). Blake, S. et al. 1998, introduced the architecture concept and the mechanisms for traffic control. The architecture is based on a domain concept that provides for static traffic and access control at the domain borders and that works completely statelessly within these borders. The following mechanisms are provided for the realization of the service classes on all network nodes of a domain: • Classifying: Differentiation of different application data streams into classes. • Scheduling: Allocation of resources to a class. This largely determines the PHB. Additional mechanisms can be employed in the border nodes: • Marking: Marking of packets as belonging to a class (DiffServ Codepoints, DSCP). • Monitoring of incoming (Policing) and outgoing (Shaping) traffic according to the previous agreement reached, e.g. with a neighboring domain (Traffic Conditioning Agreement, TCA). Before setting up a service class, the following definitions are made at the domain borders: rules for the classification of the data streams, a unique marking (DSCP), a specific behavior (PHB) of the network nodes when processing packets (Forwarding) and the QoS mechanisms to be employed. To identify a data stream, several fields of the MAC, IP, and TCP/UDP headers can be used (Multi Field Classifying). A data stream is marked by setting specific bits in the packet head (DS field). Which bits are used for this purpose in the IPv4 and IPv6 are defined in Nichols, K. et al. 1998. During network operation, the nodes classify a packet using the DSCP. This maps a PHB and assigns the packet to the reserved resources of the service class. For the DiffServ architecture, two PHBs, e.g. Expedited Forwarding and Assured Forwarding (AF), have been standardized. The Expedited Forwarding (EF) PHB is intended for real-time critical applications; it defines a DSCP and qualitatively describes the behavior of a network node with respect to the delays and losses that can be expected for packet processing. It is largely determined by scheduling and network utilization. The require-

QoS Architectures and Resource Management in the Intranet

179

ments that are placed on the technical implementation are derived from the PHB. The EF PHB, for example, is defined as follows: “The operating rate of an output must exceed the incoming rate for short and long time periods, so that only minor delays, jitter and losses occur”. How this requirement is to be ensured is not defined. In contrast to the EF PHB, the AF PHB defines several DSCP and describes which QoS mechanisms (traffic control, buffer management) are provided at which position in the network in order to implement these service classes. As with the EF PHB, no absolute QoS parameter values are defined ; only qualitative data are specified. The standard only contains relative statements between different AF classes. Within the network the operator can always select the mechanisms provided within the framework of the standard and can implement any service. Difficulties may arise if a service quality that is predictable for the subscriber must be realized across several network operators. To achieve this, a network operator must know the end-to-end behavior of his network (PDB, Per Domain Behavior). Operators can then define an appropriate service (SLS, Service Level Specification) and make agreements at the network border with other network operators (SLA, Service Level Agreement). An SLA represents a contract between two network operators. Among other things, it contains information on traffic type, traffic volume (TCA) and service quality (SLS). The original DiffServ approach is based on an agreement that is of a static nature and is set up manually by the network administration.

App.

App.

Domain 1 Domain 2

Source

Core routers

Leaf Router (police, mark flows)

Egress Edge Router (shape aggregates)

Drain

Core routers

Ingress Edge Router (classify, police, mark aggregates)

Figure 37: Example of a DiffServ architecture with two domains

Data route

180

Jörg Eberspächer, Josef Glasmann

The following two sections will now consider more extensive approaches. They are based on DiffServ architecture, but also introduce an access control function. DiffServ-based approaches with central access control

This section discusses two architectures that implement a dynamic resource management function in the form of a central server in the network. Bandwidth Broker

The IETF has standardized an implementation example for the DiffServ architecture, namely “A Two Differentiated Services Architecture for the Internet” (see Nichols, K. et al. 1999) (class B). The architecture provides two service classes: a Premium (PS) and an Assured Service (AS). They are realized by a simple classifier (2 bits: P, A) and two buffers which are operated by a Priority Scheduler. A peak bit rate for PS traffic and a medium rate for AS traffic as well as a maximum burst size are agreed as the traffic profile. Traffic is monitored at the network access point by filters. In the case of PS traffic, the filter has a maximum burst length of only one packet (peak rate allocation). What is notable in this example is the provision of access control at the connection level and the dynamic allocation of resources. Resource management agents, so-called “Bandwidth Brokers” (BB), and a signalization per IP flow between subscriber and BB as well as between adjacent BBs are implemented for this purpose. The task of a BB is to dynamically steer subscriber access to the resources of a domain. To do so, the BB carries out the following operations: • Reservation message processing, • Access control and • Access node configuration (Classifier, Policer, Marker, Shaper). A reservation message contains data on the required service class (PS or AF), the transmission rate, the maximum burst size, and the time and duration. The access control function first checks the identity and subscriber profile of the sender. It then checks whether the available bandwidth (SLS) on the link to the adjacent domain is sufficient. If so, the available bandwidth is first reduced by the required amount and then the connection data (Flow Spec) are saved. If not, it can request additional resources (capaci-

QoS Architectures and Resource Management in the Intranet

181

ties) from a BB in the adjacent domain. After successful reservation, the BB configures the access node (Classifier, Policer, Marker) with the connection parameters of the authorized connection. Bandwidth Brokers BB (access control, configuration of the traffic classes, on the Leaf or Edge Router)

App.

BB

BB Domain 1

Source

Domain 2

Core routers

Leaf Router (police, mark flows)

Egress Edge Router (shape aggregates)

App. Drain

Core routers

Ingress Edge Router (classify, police, mark aggregates)

Signaling route Data route

Figure 38: Example network with two DiffServ domains Resource Reservation Agents (RRA)

The RRA concept was published by O. Schelen and S. Pink. In every DiffServ domain, it implements a central server (RRA) that carries out an access control function on all links of the user data route. The access control process is parameter-based and supports advance reservations. For this purpose, the RRA requires information on the network topology and configuration. It participates passively in the routing process and can extract the required topology data by eavesdropping on the routing messages (e.g. Link State Advertisement). It obtains the link configurations (capacity) through targeted interrogation of the router using network management protocols (e.g. SNMP). It only considers Layer 3 topologies. A very simple protocol for establishing a reservation over several domains is provided between instances of the RRA. The messages only contain a source and target address pair and a rate parameter. The acceptance decision per link is made on the basis of the rate parameter, the sum of all rates of the reservations already accepted on a link and the link capacity. The resource requirements are determined by the source alone. The concept provides merely one service class (Priority). A service quality

182

Jörg Eberspächer, Josef Glasmann

specification and proof of the achievable QoS are not included in this concept. Resource management architecture (RM)

An RM architecture based on the RRA approach was developed in Glasmann, J., Müller, H. 2002. The focus of this work lies on the easy importability of the RM system into an existing system environment and on the ability to prove the QoS. The architecture is characterized by the functional separation of access control (RM) and topology exploration (TM). Thus, both independence from the underlying network technology and an additional degree of freedom is achieved in the configuration and network operation. This brings about independence from the underlying network technology and an additional degree of freedom in the configuration and network operation is achieved. Furthermore, the RM architecture is connected to a service control unit (H.323 Gatekeeper, SIP Proxy); thus, there is not need for any direct interface to the subscriber. The approach provides a distribution into RM domains in order to also guarantee good scalability for larger networks. A process for this is proposed in Riedl, A., Glasmann, J. 2002. DiffServ-based approaches with distributed access control

End-to-End reservation (SCORE)

Stoica et al. have developed an architecture called SCORE (Scalable Core) for DiffServ networks with the objective of simulating the transmission service of a state-based network architecture with a core network that is stateless (as far as possible). A technique with the name DPS (Dynamic Packet State) was implemented in the core routers for this purpose. DPS is a process in which ingress routers code the state information of an IP flow in the IP header (TOS, IP Fragment Field). The state information is transmitted together with the user data and evaluated by network nodes along the data route. Thus, state information for each IP flow is only saved in the ingress router and not in the core router. Subscribers signal their reservation requirements using RSVP. The RSVP messages are exchanged transparently between the edge routers. If

QoS Architectures and Resource Management in the Intranet

183

the reservation message of the recipient arrives at the ingress router again, the router triggers a reservation within the domain. This uses a proprietary signaling process which prompts all routers along the data route to carry out an access control function. The access control function is very simple and is carried out at a rate of rnew that must be reserved. Each router requires only aggregated knowledge of all the reservations that are already active (Σrold + rnew < C). After successful end-to-end reservation, the ingress router is configured with the connection parameters that are required for processing the user data packets in the core routers. They are written into the IP header of each packet by the ingress router. Stoica, I., Zhang, H. 1999, introduced the Core-Jitter-Virtual-Clock scheduling process which is controlled by the header information and permits an end-to-end guarantee with regard to the delay and the bandwidth. Access control at the input (AQUILA)

The European IST AQUILA project defined and developed a QoS architecture based on DiffServ that implements a subscriber network interface in the access region. In ACA instances (Admission Control Agent), there is an access control at the inputs which works on a parameter and/or measurement basis. A central control instance RCA (Resource Control Agent) has measured values available to it at different locations within the network and these are used to restrict the maximum traffic inflow per input and traffic class (budget). The budgets are administrated by the ACA and dynamically adapted by the RCA to the traffic conditions in the core network. The reservation protocol runs between the terminals and the edge routers. In Bak, A. 2001, a scheduling process is defined which can be used to realize only one service class for time critical traffic. This service class is independent of the packet arrivals of other traffic classes. Access control at the output (Egress AC)

The process defined by Knightly et al. is based on a similar architecture as the preceding access control function. It does not use states within the domain and only provides access control at the egress router. To signal reservation requests, it uses a modified RSVP whose messages are, however, only processed by the egress routers. Ingress routers add an identification of the traffic class, an identifier of the ingress router, a packet sequence number and a time stamp to the header of the IP packet (TOS, IP Fragment Field). With this information, the egress router can identify packet delays

184

Jörg Eberspächer, Josef Glasmann

and determine losses. For this purpose, the system times of the ingress and egress routers are synchronized via NTP (Network Time Protocol). From each transferred packet of a class and a path (ingress and egress pair), the egress router knows when it arrived in the domain and when it left the domain again. Access control in the terminal (EP-AC)

In addition to the access control function in the network (Bandwidth Broker, egress router), there are also architectures that provide this function in the terminal. They are measurement-based and use test packets which are sent before the user data are transmitted. They check the packet loss of a traffic class and only permit transmission services with a lower QoS guarantee like the Controlled Load Service of IntServ. Normally, the terminals are host computers check the load situation in the network by sending out test packets at the required rate and with the corresponding marking. This is achieved by determining the current loss probability of the test packets. Using a preset threshold value, they can then decide whether the new connection is allowed. The processes are distinguished from previous versions inasmuch as the test packets are sent in a traffic class with a lower priority (out band) than the permitted user data packets. Thus, the access decisions of this process are more defensive, resulting in a safer QoS prediction than in those described earier.

Discussion In conclusion, the various QoS architectures will now be compared. Their QoS guarantees can be generally grouped into three classes: “hard”, “weak” and “relative”. Architectures with hard QoS guarantees are distinguished by a quantitative service quality specification with packet-related limit values. The latter may not be exceeded or, at most, only tenuously and briefly (ms) exceeded. Architectures with soft QoS guarantees also have a quantitative service quality specification with limit values, which are only guaranteed in the statistical average. Architectures with relative QoS guarantees operate solely with qualitative service quality specifications. Architectures with hard QoS guarantees (e.g. IntServ) feature a flowbased end-to-end reservation process and a scheduling per flow or packet

QoS Architectures and Resource Management in the Intranet

185

both within the access region and within the core network. They differ solely in the processes (scheduling, access control) and protocols they employ. All architectures with soft QoS guarantees (e.g. BB) form traffic aggregates in the core network and carry out an access control within the access region. Frequently, the access control functions of these approaches are supported by distributed measurement systems that provide an insight into the state of the core network. Architectures with relative QoS guarantees (e.g. DiffServ) distinguish themselves from the previous architectures to the effect that either no access control is provided or this is only carried out at the network access point due to reasons of local state information. The architectures can be represented in a two-dimensional complexity/QoS guarantee chart (see Figure 39). Qualitative

Quantitative “low”

“stringent”

IntServ GS RM BRA

Medium

Complexity

High

“relative”

EgressAC BB

Low

DiffServ-AS Best Effort None

Weak

Strong

Very Strong

QoS guarantee Figure 39: Architecture comparison

The graph illustrates the basic relationship between the achievable QoS guarantee and the required complexity. The complexity does not increase linearly with growing service quality guarantee, but disproportionately. This behavior can be explained by two facts: • State information per flow is required in order to provide a reliable QoS guarantee; • The larger the desired guarantee performance degree, inevitably the larger the number of positions in the network for which states must be kept per flow.

186

Jörg Eberspächer, Josef Glasmann

It can be concluded that there are several approaches that ensure a hard QoS guarantee for real-time sensitive traffic. They have different degrees of complexity, however, there are no negative effects on scalability within the setting of company networks. Of primary interest for a network operator are system solutions with lower complexity that rely on standards-based technologies and require lower management overheads. Providers of interactive real-time services, on the other hand, place great emphasis on a higher level of guarantee and easy importability. For each of these applications, matching solutions that can be used within company networks have already been defined. Bibliography Adas, A.: “Traffic Models in Broadband Networks”, IEEE Communications Magazine, p. 82–89, July 1997 Amir, E., McCanne, S., Zhang, H.: “An Application Level Video Gateway”, ACM Multimedia '95, San Francisco, 1995 AQUILA Home Page: http://www.ist-aquila.org Audsley, N. C., Burns, A., Richardson, M. F., Wellings, A. J.: “Hard Real-Time Scheduling: The Deadline-Monotonic Approach”, 8th IEEE Workshop on Real-Time Operating Systems and Software, 1991 Bak, A. et al.: “Traffic Handling in AQUILA QoS IP Network” 2nd International Workshop on Quality of future Internet Services QofIS, Coimbra, Portugal, September 2001 Baker, F., Iturralde, C., Le Faucheur, F., Davie, B.: “Aggregation of RSVP for IPv4 and IPv6 Reservations”, RFC 3175, September 2001 Bianchi, G., Capone, A. Petrioli, C.: “Throughput Analysis of End-to-End Measurement-based Admission Control in IP”, IEEE INFOCOM 2000, Tel Aviv, Israel, March 2000 Blake, S, Black, D. et al.: “An Architecture for Differentiated Services”, IETF Standard, RFC 2475, December 1998 Boutremans, C., Le Boudec, J.-Y.: “Adaptive Delay Aware Error Control For Internet Telephony”, Internet Telephony Workshop 2001, New York, USA, April 2001 Braden, R. et al.: “Resource Reservation Protocol (RSVP) Version 1 Functional Specification”, RFC 2205, September 1997 Brady, P.: “A Technique for investigating ON-OFF Patterns in Speech”, Bell System Technical Journal, 44: 1-21, January 1965 Brandauer, C. et al.: “AC Algorithms in AQUILA QoS IP Networks”, 9th Polish Teletraffic Symposium, 2002”, 2nd Polish-German Teletraffic Symposium PGTS'02, Gdansk, Polen, September 2002

QoS Architectures and Resource Management in the Intranet

187

Chan, K., Seligson, J. et al.: “COPS Usage for Policy Provisioning (COPS-PR)”, IETF-Standard, RFC 3084, March 2001 Chen, S., Nahrstedt, K.: “An Overview of Quality-of-Service Routing for the Next Generation High Speed Networks: Problems and Solutions”, IEEE Network Magazine, Vol. 12, p. 64–79, Nov.–December 1998 Davie, B., Charny, A. et al.: “An Expedited Forwarding PHB (Per-Hop Behaviour)”, IETF Standard, RFC 3246, March 2002 Elek, V., Karlsson, G., Ronngren, R.: “Admission Control based on End-to-End Measurements”, IEEE INFOCOM 2000, Tel Aviv, Israel, March 2000 Engel, T. et al.: “AQUILA: Adaptive Resource Control for QoS Using an IP-based Layered Architecture”, IEEE Communications Magazine, Vol. 41, No. 1, January 2003 Escobar, J., Deutsch, D., Partridge, C.: “Flow Synchronization Protocol”, IEEE GLOBECOM'92, Orlando, Florida, USA, June 1992 Feher, G. et al.: “Boomerang – A Simple Protocol for Resource Reservation in IP Networks”, IEEE Workshop on QoS Support for Real-Time Internet Applications, Vancouver, Canada, June 1999 Floyd, S., Jacobson, V.: “Random Early Detection gateways for Congestion Avoidance”, IEEE/ACM Transactions on Networking, V.1 N.4, p. 397–413, August 1993 Glasmann, J., Müller, H.: “Resource Management Architecture for Realtime Traffic in Intranets”, Networks 2002, Joint IEEE International Conferences ICN and ICWLHN, Atlanta, USA, August 2002 Glasmann, J.: “Ressourcenmanagement für Echtzeitverkehre in Intranets”, Dissertation, Technische Universität München, December 2003: http://tumb1.biblio.tu-muenchen.de/publ/diss/ei/2003/glasmann.html Hancock, R. et al.: “Next Steps in Signaling: Framework”, NSIS Working Group, draft-ietf-nsis-fw-02.txt, March 2003 Handley, M., Schulzrinne, H., Schooler, E., Rosenberg, J.: “SIP: Session Initiation Protocol”, IETF RFC 2543, March 1999 Heffes, H., Lucantoni, D.: “A Markov modulated Characterization of Packetized Voice and Data Traffic and Related Statistical Multiplexer Performance”, IEEE JSAC, Vol. 4, p. 856–868, 1986 Heinanen, J., Baker, F., Weiss, W., Wroclawski, J.: “Assured Forwarding PHB Group”, IETF Standard, RFC 2597, June 1999 IEEE 802.3: “CSMA/CD (Ethernet)”, IEEE Standard, Last Update 2002 Istoica, I., Zhang, H.: “Providing Guaranteed Services Without Per Flow Management”, ACM SIGCOM, Cambridge, Massachusetts, September 1999 ITU-T Recommendation H.323: “Packet-Based Multimedia Communications Systems”, Geneva, Switzerland, 1998 Jamin, S. et al.: “Comparison of Measurement-based Admission Control Algorithms for Controlled-Load Service”, IEEE INFOCOM '97, April 1997 Jiang, W., Schulzrinne, H.: “Analysis of On-Off Patterns in VoIP and Their Effect on Voice Traffic Aggregation”, 9th IEEE International Conference on Com-

188

Jörg Eberspächer, Josef Glasmann

puter Communication Networks, 2000 Kaneko, H., Stankovic, J., Sen, S., Ramamritham, K.: “Integrated Scheduling of Multimedia and Hard Real-Time Tasks”, 17th IEEE Real-Time Systems Symposium (RTSS '96), Washington D.C., USA, December 1996 Knightly, E., Shroff, N.: “Admission Control for Statistical QoS: Theory and Practice”, IEEE Networks Magazine, p. 20–29, March/April 1999 Lamont, L., Li, L. et al.: “Synchronization of Multimedia Data for a Multimedia News-on-Demand Application”, IEEE JSAC, Vol. 14, p. 264–278, January 1996 Lee, D., Melamed, B., Reibman, A., Sengupta, B.: “Analysis of a Video Multiplexer using TES as a Modeling Methodology”, IEEE GLOBECOM'91, p. 16–19, 1991 Leland, W. E., Taqqu, M., Willinger, W., Wilson, D.: “On the Self-Similar Nature of Ethernet Traffic”, IEEE/ACM Trans. on Networking, vol. 2, no. 1, February 1994 Linn, D., Morris, R.: “Dynamics of Random Early Detection”, ACM SIGCOMM, S. 127–137, Sophia Antipolis, Frankreich, September 1997 Little, T., Ghafoor, A.: “Synchronization and Storage Models for Multimedia Objects”, IEEE JSAC, Vol. 8, No. 3, p. 413–427, April 1990 MacFaden, M., Partain, D., Saperia, J., Tackabury, W.: “Configuring Networks and Devices With SNMP”, IETF-Draft, draft-ietf-snmpconf-bcp-12.txt, January 2003 McCloghrie, K., Fine, M. et al.: “Structure of Policy Provisioning Information (SPPI)”, IETF Standard, RFC 3159, February 2003 McCloghrie, K., Perkins, D., Schoenwaelder, J.: “Structure of Management Information Version 2 (SMIv2, IETF Standard, RFC 2578, April 1999 Nichols, K., Blake, S., Baker, F., Black, D.: “Definition of the Differentiated Services Field (DS Field) in the IPv4 and IPv6 Headers”, IETF Standard, RFC 2474, December 1998 Nichols, K., Jacobson, V., Zhang, L.: “A Two-bit Differentiated Services Architecture for the Internet”, IETF RFC 2638, July 1999 Pan, P., Hahne, E., Schulzrinne, H.: “BGRP: Sink-Tress-Based Aggregation for Inter-Domain Reservations”, KICS, Journal of Communications and Networks, Vol. 2, No. 2, June 2000 Pan, P., Schulzrinne, H.: “YESSIR: A Simple Reservation Mechanism for the Internet”, International Workshop on Network and Operating System Support for Digital Audio and Video NOSSDAV, Cambridge, England, p. 141–151, July1998 Perros, H., Elsayed, K.: “Call Admission Control Schemes: A Review”, IEEE Communications Magazine, p. 82–91, November 1996 Ramamritham, K., Stankovic, J.: “Scheduling Algorithms and Operating Systems Support for Real-Time Systems”, Proceedings of the IEEE, Vol. 82, No. 1, January 1994 Rejaie, R., Estrin, D., Handley, M.: “Quality Adaptation for Congestion Controlled Video Playback over the Internet”, SIGCOMM, 1999

QoS Architectures and Resource Management in the Intranet

189

Riedl, A., “Hybrid Genetic Algorithm for Routing Optimization in IP Networks Utilizing Bandwidth and Delay Metrics”, IEEE Workshop on IP Operations and Management (IPOM'02), Dallas, USA, October 2002 Riedl, A., Glasmann, J.: “On The Design of Resource Management Domains”, International Symposium on Performance Evaluation of Computer and Telecommunication Systems, SPECTS 2002, San Diego, USA, July 2002 Romanow, A., Floyd, S.: “Dynamics of TCP Traffic over ATM Networks”, IEEE JSAC, V. 13 N. 4, p. 633–641, May 1995 Rothermel, K., Helbig, T.: “An Adaptive Stream Synchronization Protocol”, Network and Operating System Support for Digital Audio and Video, 1995 Rueda, A., Kinsner, W.: “A Survey of Traffic Characterization Techniques in Telecommunication Networks”, IEEE Canadian Conference on Electrical and Computer Engineering, Calgary, Canada, Vol. 2, p. 830–833, May 1996 Schelén, O., Pink, S.: “Resource Reservation Agents in the Internet”, 8th International Workshop on Network and Operating Systems Support for Digital Audio and Video (NOSSDAV'98), Cambridge, UK, July 1998 Schelén, O., Pink, S.: “Resource Sharing in Advance Reservation Agents”, Journal of High Speed Networks, Special Issue on Multimedia Networking, Vol. 7, No. 3–4, 1998 Schulzrinne, H., Casner, S., Frederick, R., Jacobson, V.: “RTP: A Transport Protocol for Real-Time Applications”, IETF RFC 1889, Januar 1996 Steinmetz, R.: “Synchronization Properties in Multimedia Systems“, IEEE JSAC, Vol. 8, No. 3, p. 401–412, April 1990 Stoica, I., Shenker, S., Zhang, H.: “Core-stateless Fair Queuing: A Scalable Architecture to Approximate Fair Bandwidth Allocations in High-Speed Networks”, ACM SIGCOM, Vancouver, Canada, August 1998 Strassner, J., Ellesson, E., Moore, B., Westerinen, A.: “Policy Core Information Model – Version 1 Specification”, IETF-Standard, RFC 3060, February 2001 Tan, W., Zakhor, A.: “Real-Time Internet Video Using Error Resilient Scalable Compression and TCP-Friendly Transport Protocol”, IEEE Transactions on Multimedia, 1999 W. Almesberger, W:; Le Boudec, J.-Y., Ferrari, T.: “SRP: a Scalable Resource Reservation for the Internet”, 6th International Workshop on Quality of Service (IWQoS'98), May 1998 Wang, Z., Crowcroft, J.: “Quality of Service Routing for Supporting Multimedia Applications”, IEEE Journal of Selected Areas in Communications, Vol. 14, Issue 07, September 1996 Willinger, W., Taqqu, M., Sherman, R., Wilson, D.: “Self-Similarity through HighVariability: Statistical Analysis of Ethernet LAN Traffic at the Source Level”, IEEE Trans. on Networking, Vol. 5, USA, p. 71 – 86, 1997 Wroclawsk, J.: “The Use of RSVP with IETF Integrated Services”, IETF Standard, RFC 2210, September 1997 Yuan, P., Schlembach, J., Skoe, A., Knightly, E.: “Design and Implementation of Scalable Edge-Based Admission Control”, International Workshop on QoS in

190

Jörg Eberspächer, Josef Glasmann

Multiservice IP Networks (MQoS '01), Rome, Italy, January 2001 Zhang, H.: “Providing End-to-End Performance Guarantees Using Non-WorkConserving Disciplines”, Computer Communications: Special Issue on System Support for Multimedia Computing, 18(10), October 1995 Zhang, H.: “Service Disciplines for Guaranteed Performance in Packet-Switching Networks”, Proceedings of the IEEE, Vol. 83, No. 10, p. 1373–1396, October 1995

IV. Technical process support

Business processes are being automated across functional areas and company boundaries. This indicates that a continuous application landscape is required and it is enabled via a common, open, intra-company network. The first step involves the transformation of application-specific data into a uniform information structure using, for example, content management systems. In addition, consolidation of the fragmented communication landscape across all media and applications is needed.

Shai Agassi

Business process evolution in real-time

New technology has always been both a driver of change and a primary tool for coping with change. In the current market place, every company is awash in computing power, network connectivity, and functionality provided by richly complex applications. To win companies must create a recipe – a secret sauce – that combines their understanding of customer needs and ability to meet them with all the power that technology offers. The central problem facing most companies is that these days even when they find it, the sauce doesn’t stay secret for long. Technological product innovation, giving half a decade advantage to companies 20 year ago, now spread through the eco-system in less than 3 months. The only sustainable competitive advantage is the ability to adapt to change. The concept of the real-time enterprise has emerged as the general shape of the solution. Yet the real-time enterprise requires continuous changes and ultimate flexibility in reconfiguring business processes to fit ever changing market conditions. As such, the Evolving Enterprise is a vision of the corporation as an organism that constantly assimilates new processes and adapts to them. In doing so, the organism changes itself to better respond to current conditions or to create a advantage changing the eco-system for other organism fighting for success. To become an Evolving Enterprise companies must organize themselves, their technology foundation and their process landscape in a new way. They need to match a shortened change management cycle required by a rapidly evolving strategy with flexibility at the IT landscape that allows them to recompose their processes almost on the fly. All this needs to be done while at the same time reducing costs. At SAP®, we feel the key to this transformation is to put business processes front and center, and to build an entire infrastructure focused on lower the costs and increasing the speed of evolution in response to market forces.

194

Shai Agassi

Switching to the future Consider the crossroads at which AT&T arrived 60 years ago regarding its handling of long-distance telephone calls. At the time, most such calls were completed manually, by banks and banks of operators working in front of large plug boards. As AT&T’s managers and technologists looked into the future, they saw that the volume of calls the company handled was about to explode. By increasing the capacity of trunk lines, technology had helped make long-distance calls much less expensive, more useful, and more attractive, which meant that usage of long-distance calling was bound to grow exponentially. But AT&T quickly realized that to handle this growth using its old methods, it would eventually need to hire the entire population of the United States to work as telephone operators. To avoid that, the telephone company came up with a new breed of automated switch and a dialing scheme based on 3-digit area codes that together enabled subscribers themselves to dial and complete their own calls. Today, every CEO and CIO can see that something similar is happening in their own companies, but at a much more dramatic pace and involving much greater risk. AT&T had the luxury of facing not even a single competitor, but what company today can say that about itself. So, instead of a decade or two in which to innovate in process and strategy, today’s corporation must bring new solutions online in a matter of months. If it doesn’t keep up with this imperative, it risks falling far behind in its markets and inevitably, irrecoverable decline. And that’s not all: Just months after that first set of solutions come online, they will inevitably require significant enhancements and extensions, and the development cycle begins again. There’s truly no rest for the weary.

Manage change or die If there’s anything that’s a constant in today’s business world, it’s change. At every level of the business, change is a defining characteristic. Whether it’s the price of raw materials or customer expectations, logistics schedules or the competitive landscape, company identities or basic business models – everything is morphing and mutating. Nothing is static. Naturally, managers look to IT for answers to many of these challenges. But IT today is geared to shorten the transaction cycle, an achievement that gave most companies an advantage in the previous century. The reduction

Business process evolution in real time

195

in transaction cost came at the expense of process rigidity. Much of the IT infrastructure currently in place cannot handle the constant process changes now required, in either business processes or business strategies. But from now on, for companies to survive the almost Darwinian selection of fittest over merely fit, IT itself must change. And it must change in a way that not only enables but actually facilitates rapid, ongoing, open-ended change. What’s needed, this essay will argue, is a fundamentally new way of conceiving of and building IT systems, a framework that marries business process and infrastructure while actually increasing flexibility and lowering cost. Sounds like a contradiction in terms? Consider this: Driving the overwhelming need for flexibility in IT is the fact that in virtually every market today, business processes are evolving at warp speed. As soon as one supplier innovates in a critical business process, all of its competitors must make essentially the same improvements, if not even greater ones, or else they risk losing market share and going into permanent decline. Examples abound: Amazon.com showed how to build a better bookstore using the Web, and ever since Barnes & Noble, caught off-guard, has been scrambling to catch up. Wal-Mart’s implementation of direct connections between store inventories and suppliers has changed the relationship between retailers and suppliers forever. Companies like Nike and Cisco orchestrate huge networks of other companies and ship entire lines of product never touched by one of their own employees. What’s common with these companies? They found a way to constantly change their business domains, and perfected the process change, leaving behind or digesting companies that couldn’t match them.

Automation in the way? Making things even more difficult, of course, is that most of the business processes that drive companies today are already automated. AT&T was lucky enough to be working from more or less a blank slate, automating a slow-moving, relatively simple and well-understood manual process. Today’s corporation, in contrast, must find ways to make layer upon accumulated layer of automation – some of it implemented in legacies of relatively archaic technologies – perform in new ways. Teaching old dogs new tricks is not something that most CIOs wish to contemplate, but that is exactly what’s needed. A fine balance needs to be struck between extending the life of old running code, and replacements of new processes. In both

196

Shai Agassi

cases though the change is a tough operation, akin to open heart surgery, done with the same precision and similar care. What all this means is that the very process of automating business processes must itself evolve and take on a new level of capability. Rather than simply enabling work, as it has mostly done in the past, the IT organization must become an enabler of flexible business strategies. IT must begin to drive innovation into business processes across the entire enterprise. IT must make change sustainable by containing cost and reducing risk. A Herculean task indeed, yet anything less is not enough.

Outsourcing: cut costs, lose control What’s a company to do? One option that has enjoyed popularity in recent years is outsourcing. Hire an outside company to take responsibility for running part or all of IT and like that, many variable costs are now fixed. Increasingly, the writing of new software under such outsourcing contracts gets done overseas, often in Asia, where labor rates are lower. Inevitably, though, outsourcing leads to a loss of control. Instead of becoming more flexible, business processes and strategies become more rigid. For example, if a company’s opportunity management & customer support process is outsourced to an offshore location, simply managing the on going life of the code supporting these processes will contain the IT cost associated to maintain them. The organization will find it harder though to reconfigure these processes to meet changes happening in their industry in improving either of these processes, as an off shore entity in most cases a third party will add complexity to change. It will get even harder to reconfigure both processes and almost impossible for the organization to build a new process on top of these processes, evolving to tie the support-to-opportunity cycle in reverse of its current opportunity-to-support cycle. The impedance mismatch between outsource agreements, based on lower cost for fixed process agreements, and change management requirement placing a premium on speed to adoption will require a new infrastructure that allows somehow for both to coexit.

Business process evolution in real time

197

Business process evolution An alternate solution is gaining increasing attention across Corporate America, namely the idea of business process evolution. Here, a new kind of IT technology, operating as much on the level of business process as on that of traditional IT infrastructure, is enabling companies to deploy, execute, monitor, and continually optimize their business processes and strategies with unheard of rapidity and flexibility. These strategies allow for mergers and acquisitions, consolidation of IT systems, and the outsourcing of selected non-strategic activities. And for once, when deciding to outsource such activities, a company needn’t lose control over them. The name of this new technology is SAP NetWeaver™, and its purpose is to help companies design, build, and deploy innovative business processes on top of their existing IT components. Instead of always having to rewrite old programs or spend large sums on replacing them with new software products, corporations can choose between wrapping legacy processes and orchestrating them or replacing these processes with new process engines and using new software components. SAP NetWeaver™ allows them to harness existing software and hardware investment in powerful and evolving new ways. The orchestration of selected and specific functions of existing components and, in effect, creating all-new applications, called composite applications, that can be designed to support any new business process a company requires. As such, it represents a fundamentally new way of building IT systems, with more mutual alignment between technology, process, and business strategy than has been possible before – and yet, designed with flexibility as a goal not an afterthought. With SAP NetWeaver the business process is extended to provide not only a transactional view, but rather a knowledge centric collaborative flow of events and information through the organization. Its workings are presented in a way that both business and IT managers can understand equally well, providing them with a common platform to negotiate business change with clarity. SAP NetWeaver’s common language for describing business processes and the applications that support them enables enterprises and their managers to react to and even anticipate changing business conditions much more rapidly and precisely than before. From now on, the business itself can evolve as rapidly as is needed, with no waiting for distant outsourcers to respond to requests or for entirely new technology stacks to be created. When change happens, process models reflect the change, and code is generated to match the new models. In the past, change was documented, code had to be rewritten, and hopefully the code reflect-

198

Shai Agassi

ed the needs stated in the original business goal.

Implications of web services One of the key enablers to the new possibilities for IT architecture is rise and acceptance of web services as a way for applications to communicate. Web services are like a universal communication plug that allows applications to plug into each other and exchange data and invoke each others functionality. SAP NetWeaver™ uses web services standards as the basis for an enterprise class, services-oriented IT and process architecture. The central idea behind this kind of architecture, which is one of the hottest topics among advanced IT thinkers today, is that existing applications are treated as a stable platform on top of which new apps are built and innovative new processes are supported. By exposing selected functions from these existing applications as well-defined, well-behaving services, in a contractual way, it becomes possible to rapidly construct new, composite apps that can meet the needs of any new processes a company may need. Instead of writing software functions from scratch each time they’re needed, those functions may be discovered in existing apps and called into action as required. Building new apps becomes largely a process of stitching together and reusing pre-existing software that’s available as services provided by black boxes whose inner workings need not be considered. In effect, the construction of apps takes place at a higher, more abstract level, which pays off in terms of greater flexibility, improved collaboration between business and IT managers, and significantly lower implementation costs. As an open, standards-based integration platform, therefore, SAP NetWeaver™ reduces complexity, lowers TCO, and improves the alignment of IT and business objectives. Making the software even more useful to customers is that unlike competing suppliers of similar technology, which seek to sell professional services related to their products, SAP intends to expand an already-robust catalog of pre-defined business processes that it delivers in the form of packaged products. Leveraging a platform like that, corporations now may choose to outsource some of the black box components, as long as they are guaranteed the flexibility to change composites on top. They should require that the black box components are maintained fresh, in effect requiring a guarantee that they represent the best practices fit to their industry for the specific

Business process evolution in real time

199

component. Doing so maintains the speed of transaction closure, while reducing the cost of IT yet not at the price of losing the flexibility that comes from the ability to recompose the new composite processes representing strategic changes.

Enterprise services architecture Services open a door to ultimate flexibility, and that is a big problem that gets in the way of their usefulness. One can easily imagine services for everything at every level. A service can provide the correct time. Another service could execute the payroll for a 10,000 person company. Both are services, but they are clearly not in the same ballpark. In order to make sense of the possibilities of services and to make them useful for IT in a particular business, you need a theory that classifies services into groups that will provide the building blocks for integration and innovation. At SAP, our theory of services is called Enterprise Services Architecture, which explains the structure of a service-oriented business application shown in the following graphic:

NEXT PRACTICES

BEST PRACTICES Another Suite

mySAP Business Suite

xApps

Composites

Composites

Composites

...

Dispute mgmt

Dispute mgmt

Invoice verif.

x...

xPD*

Order mgmt (new)

3rd Party

SAP NetWeaver

Order mgmt (new)

General ledger (new)

Components

General ledger (new)

General ledger (old)

...

Components * Product Development

Figure 40: Service-oriented business application

200

Shai Agassi

The way that services work in Enterprise Services Architecture is as follows: • The components at the bottom of the graph are services provided by enterprise applications. Programs like ERP, CRM, and so on, are broken up into collections of services that can then be used to support integration or to create other applications. • SAP NetWeaver™ abstract these components as services and enhances them with functionality from the integration tools like content management and business warehouse. It allows component services to be combined to create new services to meet special needs. It also provides development tools and common frameworks so that new applications can be created from services and new services can be created from scratch where needed. • At the top of the stack are composite applications, applications built from services provided by other applications. Increasingly, traditional applications like ERP and CRM will be reborn as collections of common components woven together through composites. At the right of the graph, you can see composite applications that will break new ground in innovation introducing what we cal next practices, not yet commonly adopted as best practices by the industry they serve. At SAP, we call these applications xApps.

Proving our point This article has laid out the path from the pressures pushing companies toward the real-time enterprise to what SAP sees as the solution in our implementation of Business Process Evolution. This vision of ultimate flexibility comes to us in each generation of technology. The question is can we realize it in a way at is affordable and provides competitive advantage. Showing that we have reduced the Total cost of operation and ownership at vanguard customers, we at SAP we feel that NetWeaver is the bridge that connects the needs of stable and robust transactional execution environment with the emerging behavior of business thrown into consistent change management. We are eager to help every one of our customers introduce the ability to rapidly evolve, and become a leader of its market.

Friedrich Kopitsch

I&C Technologies for a Real-Time Enterprise (RTE)

Gartner Group analysts define Real-Time Enterprise (RTE) as follows: “The real-time enterprise is an enterprise that competes by using up-to-date information to progressively eliminate delays to the management and execution of its critical business processes”. This definition does not discuss the methods by which central company business processes might be accelerated. The present article examines a selection of I&C technologies that can contribute to transforming a company into a real-time enterprise. Each technology will be briefly described, followed by an examination of their benefits to the RTE. To sustain competitiveness, real-time enterprises must work ceaselessly to maintain their RTE status, examining the maturity and utility of promising technologies on a regular basis. I&C technologies support companies on the road toward becoming RTEs. Picking the right choice and combination of suitable technologies is essential for accelerating business processes critical to the company, ensuring that information on events affecting the business is received quickly and reaching decisions based on this data. In general, companies will opt for mature technologies – because on top of potential benefits, the use of new technologies at an early stage also involves risks. The following technologies will be discussed in this article: • Business process analysis and modeling • Rules Engine • Enterprise Application Integration (EAI) • Enterprise Application Portals • B2B Connectivity • Web Services • Service-oriented architectures (SOA) • Wireless LAN • Wireless Messaging

202

Friedrich Kopitsch

• Instant Messaging • Location-Based Services • Real-Time Analytics • Mass RFID Tagging

Enabling technologies for a Real-Time Enterprise (RTE)

Business process analysis and modeling

Continuous change brought on by global competition, new service provision patterns and shorter product development cycles necessarily leads to an ongoing evaluation and reorganization of value chains and the underlying business processes. In addition, a comprehensive business process management that aims not only at analysis, design and optimization but also has monitoring and control as its objective becomes a central competitive factor – especially in times when economic pressure forces companies to comb through their entire range of procedures and activities in search of optimization potential. What are business processes?

Since the advent in the 90’s of the radical new approach of Business Process (Re-)Engineering, matters of procedural organization have taken precedence over those of structural organization in corporate organizational design. Consequently, studies made in the course of designing organizational structures focus on operative processes. This type of business process describes the generation of output on the basis of: • the acceptance of goods and services required for the process from suppliers; • the coordination of the entire output generation process • and the event-controlled flow of tasks that are assigned to participating instances and objects. In keeping with the concepts of Business Process (Re-)Engineering, the core processes serve to fulfill the immediate goals and are responsible for passing the outputs on to external customers. All other operative processes

I&C Technologies for a Real-Time Enterprise (RTE)

203

are called auxiliary or support processes and either support these core processes or generate output for internal customers.

Supplier

Processing steps Input

Customer Output

Figure 41: Business process structure Modeling business processes

In business process design, companies apply design, analysis and optimization in a cycle of continuous improvement to align their business processes to the requirements and needs of the market and their own enterprise. Modeling tools of various manufacturers (e.g. ARIS) enable an interactive graphic description of business processes, including their simulation and final evaluation. Process design provides answers to such questions as to the sequence of process steps or whether parallel processing is an option. It clarifies which outputs must be generated and whether and how output generation can be supported by software systems. In doing so, it describes the elements of both the procedural and structural organization, including tasks, functionaries, materials, information, volumes, costs, etc. Analysis finds the key data for cost drivers and permits the identification of capacities and capacity utilizations; weak spots are revealed and attention is called improvement potential. Knowledge gained from analysis and optimization ultimately leads to the definition of target processes, i.e. the method by which added value is to be generated in a company in the future using the described business process. Business processes defined in this way reveal where in a company existing software systems are able to provide support. Should it become necessary to purchase or develop suitable systems, business processes provide the foundation on which the requirements of such system are specified during the requirements engineering phase. Also, applications are easier to service if the control logic is not hardcoded with regard to business processes but is out-sourced using inter-

204

Friedrich Kopitsch

pretable rules and control flow instructions that are made available in realtime. This makes it possible to react with particular speed to changes on the fastest-paced architectural level, the process level.

Organization diagrams

Data cluster models

Function trees

Value chain, function assignment diagrams

Figure 42: ARIS model types Executable processes

By offering methods for modeling, monitoring and easy modification of business processes in real time, Business Process Management (BPM) makes it possible to keep pace with volatile market and technology trends, and interest in it has risen sharply. To execute business processes, BPM requires a powerful workflow component. For this purpose, the process model has to be interpreted; it must include the management of tasks and business rules, both automated and manual, as well as exception and escalation procedures. The development and convergence of technology standards in the area of web services and process logic open up a whole new range of possibilities in process management and business integration. The ideal would be if users of business solutions themselves could integrate completely different corporate applications in their own business processes. Rules engines

The success of a company is strongly related to its capacity to make consistent, prompt and effective decisions. The decision-making process is driven by corporate business rules, which are frequently-changing specifications that define interactions with customers, suppliers, employees and partners. Business rule determine how commercial enterprises set their prices and insurances set risks and premiums. And they make is possible,

I&C Technologies for a Real-Time Enterprise (RTE)

205

for example, for engineers to control processes and financial consultants to recommend investment strategies. Analyst Developer Quality assurance Define and improve

RulesEditor

Deploy

Engine Manager Analyst

Business process

Rules

Engine

Application Review Quality assurance Developer End user

Figure 43: Rules Engine

Business rules are among a company’s most valuable assets. However, in most companies they are difficult to localize: Unless they have been explicitly identified and managed, they are generally lost for any practical use. Either they are embedded in program code and distributed over a large number of old applications, or they reside in unintelligible business practices and procedures. Finding them is difficult and comprehending their business significance can be even harder. Changing them may be impossible. Just as companies keep their strategic corporate data in suitable databases, they ought to use equally sophisticated technologies to manage their business rules intelligently and strategically. Ideally, business rules are represented as algorithms or logical instructions, in the form of decision tables or trees, or in special rules languages (CLIPS, OPS5). If represented in the form of scripts, it is possible to manage the procedural business logic. It is no longer an option to describe procedural business logic simply in the form of IF-THEN-ELSE constructions with underlying formats such as decision tables and trees. Today’s rules engines offer graphic editors for identifying and modifying business rules that are suitable for use by technicians and by less technically-oriented users.

206

Friedrich Kopitsch

For facilitate efficient maintenance and the consistent expansion and modification of large rule bases, some rules engines offer components for validating sets of rules with respect to completeness and freedom from conflict and redundancy. In this context, completeness means that the existing rules will cover any situation that may arise. Freedom from conflict means that the rules do not contradict one another and freedom from redundancy means that the rules do not overlap. Organizations today must respond to changes in the market within a very short time span. Typically, reaction times required on the market have decreased from months to days. And some companies must even react on an hourly basis, if not in real time. The ability to easily find and efficiently work through a process – especially in conjunction with the reliable distribution of modified business rules to all business systems involved – generally guarantees extremely short reaction times. This generally does not require any changes to the applications. Strategic management of an organization’s business rules improves competitiveness and efficiency, reduces costs and accelerates response to external and internal requirements, leading to better customer loyalty. Rules engines provide an environment for the automatic management of a company’s business rules. Their functions include: • Identification and documentation of business rules and their distribution to technical applications and IT users; • Support for the introduction and use of standard company vocabulary; • Translation of business rules from a format understood by all participants into one that can be used by IT applications; • Provision of interfaces for those IT applications that themselves must use business rules; • Identification, tracking, recording and resolution of conflicts that arise in the application of business rules; • Introduction of a change management so that changes to business rules can only be released after a suitable degree of quality assurance has been reached. In a process-driven organization, business rules are identified and managed in parallel to the business process portfolio. In such an environment, the rules engine, as part of the execution environment for the business processes, takes on the task of working through the rules that drive the business process.

I&C Technologies for a Real-Time Enterprise (RTE)

207

Enterprise application integration (EAI)

EAI: a contribution to the real-time enterprise

An important factor in the success of a company is highly efficient and flexible business processes, as observed by Michael Fleisher (Gartner Group) in an interview in November 2002. “For years now we have noted an increase in the speed with which companies conduct business. We believe that in no later than ten years from now, most business processes will take place in real-time or at least in near real-time.” This means that it will become necessary for there to be an efficient flow of information within the company itself, from and to customers and suppliers. Even today, it is of great significance for a company to develop and implement a strategy that will achieve this goal. Efficient business processes are characterized by a high level of IT support and thus by a largely automated flow of information. “Application integration” is the central topic when the aim is to extend the business processes of an existing system landscape or to implement new ones. This is an undertaking that is becoming necessary with increased frequency and in ever-shorter intervals due to the rapidly changing conditions and new challenges on the global market. To successfully meet these challenges, specific technologies and a time-tested procedure are needed. If risks are kept down and economic aspects are taken into consideration, this should generate lasting added value for a company in a relatively short time. The concept of “Enterprise Application Integration” (EAI) refers to technologies and methods that, with a high level of automation, enable the integration of different applications to form comprehensive business processes. Initial situation

Perpetual cost pressure, continuously changing business requirements and the increasing complexity of IT systems are issues with which companies are confronted on a daily basis. The IT landscape in most companies today has been shaped by historically grown system structures. Over the years, these have naturally led to heterogeneous system landscapes. Above all else, cost pressure is what in many cases forces companies to reuse existing applications and integrate them in new business processes instead of implementing cost and time intensive new developments.

208

Friedrich Kopitsch

To ensure a lasting competitive advantage, companies must implement highly automated, cross-company business transactions. This requires that new applications and, more importantly, functionalities of the historically grown, heterogeneous system landscapes be integrated to form homogeneous business processes, a step that is necessitated by economic considerations in order to secure the long-term investment made in existing IT systems. To fully utilize the potential of the entire value added chain, integration must be broadened to include business partners, customers and suppliers, both today and in the future. In doing so, it is not only important for the IT systems be technically interconnected, it must also be ensured that comprehensive business processes are covered by IT (e.g. an ordering procedure at a supplier, including logistics and financial accounting). In today’s trend toward globalization, economic boundary conditions are increasingly being shaped by merger plans and joint ventures. Consequently, future integration concepts must incorporate organizational and technical integration. From a global perspective, this means that an intelligent and flexible networking of existing system structures will be the driving force behind innovations in the future. At the core of all activities is utility, which must lead to an increase in productivity and ultimately to a reduction in costs.

n2

n

Figure 44: Reduction of complexity and costs

According to Giga Information Group, companies invest almost 40 percent of their IT budget on the integration of existing IT systems. These investments are largely used for developing and operating proprietary direct connections between applications. Many analysts agree that the necessity to optimize business processes will continue to increase as the economic situation worsens. The “Enterprise Application Integration” (EAI) technology

I&C Technologies for a Real-Time Enterprise (RTE)

209

offers a foundation on which to successfully counter these challenges. EAI technology and its associated products promise simple, flexible and futureproof integration of applications. This will lead to the unrestricted use of data and functionalities in any application and business process. EAI platforms

To achieve integrated and “real-time oriented” business processes, investments must be made in specific technologies and procedures. Only then can there be any hope of developing faster, changeable processes with longterm cost control. Powerful, scalable application servers that support open standards are today a matter of course. With their increasing emphasis on the seamless interaction of applications for the purpose of improving process quality, companies have a need not only for application platforms but also for integration platforms with the following additional features: • Adapter for integrating heterogeneous and historically grown landscapes; • A brokering infrastructure that helps keep IT landscapes manageable and makes it possible to flexibly configure brokering scenarios; • A definition and execution environment for process control (“Choreography and Orchestration”) that is easy to modify without requiring changes to the application components in use; • Pre-designed templates (process templates) for the accelerated implementation of similar or standardized processes (e. g. the alignment of master data, ordering processes). The concept of “Enterprise Application Integration” is pertinent for several IT levels, including data, functions, services, processes and users. It is helpful to structure integration platforms within the categories of “People, Information, Process”, which are handled by all leading manufacturers. Process integration plays a central role in a promising overall approach (“Total Business Integration”), since it is the definitive factor in achieving a comprehensive, high-performance execution of processes and the prompt provision of decision-relevant information. Because of the great significance of process integration, the present specialists in this field (TIBCO, Crossworlds, webMethods, etc.) are now being joined on the EAI technology market by all major vendors of IT platforms, either with their own developments or purchased solutions.

210

Friedrich Kopitsch

IBM WebSphere • Portal Development Kit for WebSphere • Integration of Lotus Sametime

SAP NetWeaver People Integration Multi-Channel Access Portal Collaboration

Information Integration • Integration with IBM Content Manager

Knowledge Mgmt Business Intelligence Master Data Mgmt

Microsoft .NET • Portal Development Kit for .NET • Smart Client on .NET • MS Office Integration • MS Exchange Integration

• Integration of MS Content Management Server • Integration of SQL Server Analysis Services

Process Integration • Interoperability with WebSphere Business Integration

Integration Broker, Business Process Mgmt

• Eclipse Based Java IDE • Technical, standards-based connectivity (XML, Web Services)

J2EE ABAP, DB and OS Abstraction

Application Platform

• Interoperability with MS Biztalk Server

• Visual Studio .NET integr. • Technical, standards-based connectivity (XML, Web Services)

Source: SAP AG

Figure 45: SAP Netweaver architecture Enterprise Application Portal

“Enterprise Application Portal”, or “EAP” for short, encompasses: • a design for the complete enterprise portal that emphasizes: – support and optimization of business processes – integration of applications and systems – development of new corporate applications using web technologies • an architectural and platform design characterized by: – pre-developed reusable solution components that represent the portal platform; they are introduced in projects as open sources and can be flexibly adapted to the project-specific requirements; – modern, flexible and future-proof solution architecture based on open standards (HTML, J2EE, XML/XSL ...), suitable for enterprisecritical application; – openness to all major middleware and system platforms (commercial products, open-source platforms); – cost advantages regarding system development and operation thanks to a strategic use of open-source components. The EAP approach emphasizes integratability. The three levels of integra-

I&C Technologies for a Real-Time Enterprise (RTE)

211

tion shown in the following integration triangle interact: User integration (Employee, customer, partner, suppliers) Efficient support of and cooperation between users across all target groups • Business process optimization • Cost savings • Customer and partner loyalty

• • • •

Your business

Process integration (Business processes)

IT efficiency and quality IT cost optimization Investment security Orientation toward the future

IT integration (Applications, systems)

Figure 46: Integration triangle of enterprise portals

User integration strives to optimally support all application users in their tasks and workflows. For this purpose, portals make available portal access or portal views appropriate for each target groups. Every user group and every user is supported according to his or her own needs. Web user interfaces are designed for intuitive and convenient operation. The aim of process integration is attaining optimal and comprehensive imaging, support and optimization of business processes. It not only targets business processes within the company, but also those processes involving customers, business partners and suppliers. Media discontinuities in IT support ought to be avoided. In these days of continual change in the business environment, high demands are placed on the flexible adaptation of business processes and, thus, on the flexible adaptation of IT support. The aim of IT integration is attaining an optimal integration of IT in the company. Ideally, all web-oriented solutions are combined in a comprehensive portal architecture (main principle behind enterprise portals). Existing applications and systems are connected and integrated, creating a comprehensive overall application landscape (main principle behind EAI). IT integration has the optimization of development and operation costs and the gradual establishment of a future-proof IT (investment security) as its primary goals. The following figure shows an overview of the EAP solution compo-

212

Friedrich Kopitsch

nents that support these three integration levels: User integration Navigation and dialog structure

Authentication

Single Sign On

Personalized search

Internationalization and multilingualism

Authorization

Portlets

Personalized sitemap

User management (users, roles, user groups, rights)

Process integration (business process management) Management / Maintenance of technical objects

Workflow support

Document management

Collaboration / Community

IT integration (applications, systems) LDAP access

RDBMS access

CMSintegration

LDAP (Dir)

Database

CMS

R/3 access

Host access

SAP R/3

Host application

Server access

Server application

Figure 47: An overview of EAP solution components

EAP has a state-of-the-art 3-layer architecture, implements design patterns such as MVC (Model View Controller) and is therefore suitable for complex, enterprise-critical applications. EAP components are developed in conjunction with ambitious enterprise portal projects. Thus they have a practice-oriented design and are field-tested. The EAP implementation is based entirely on open standards (J2EE, XML/XSL, LDAP ...). Its basic components are open source: Struts (MVC Framework), FOP (Formatting), Log4J (Logging), Xalan (XLS Transformation), Xerces (XML Parser). All of the components are established as the industry standard and have proven their practical suitability and their effectiveness in demanding, complex and highly scalable enterprise applications. Due to its consistent adherence to open standards, EAP can be used in virtually every current system and middleware product. These products not only include standard products from manufacturers such as IBM, Bea, Oracle, etc., but also open source platforms like Apache Tomcat, JBoss, OpenLDAP, mySQL and Linux. The following figure shows an overview of the most important platforms:

I&C Technologies for a Real-Time Enterprise (RTE)

213

Basic components in use MVC architecture

Formatting

Logging

Struts

FCP

Log4J

XSL transformation Xalan

XML parser Xerces

SVG

Supported middleware and operating systems Application server

LDAP server

RDMS

Operating system

Bea Weblogic

Tomcat

OpenLDAP

Oracle

Windows NT/XP

IBM WebSphere

JBoss

iPlanet

IBM DB2

Linux

IBM SecureWay

MS SQL Server

Sun Solaris

mySQL

IBM AIX

MySAP WAS

Agenda:

Open Source

Commercial Product

Figure 48: Supported platforms

It is important to emphasize that EAP is an open solution platform rather than a product. EAP components are introduced in customer projects as pre-developed reusable components. The general philosophy is oriented on the open source principle: Components are made available to a customer project as open sources and are tailored to the specific demands according to customer and project requirements. Where necessary, they are expanded to include further customer-specific components and functions. Open solution platforms provide enterprise solutions with the following advantages: • Security and stability for project procedures through the use of field-tested basic components. • Good expansion and integration capability • Flexible adaptation to company requirements • Based on open standards rather than proprietary products • Independent of manufacturer and product • Maximum platform transparency and openness • Low-cost solution due to open source and the use of pre-developed components

214

Friedrich Kopitsch

B2B connectivity

An important factor in the success of a company is the presence of business processes that are characterized by a high degree of efficiency and flexibility. This means that within the company itself, there must be an efficient flow of information from and to customers and suppliers. Therefore, it is of great significance for a company to develop and implement a strategy that will achieve this goal. B2B transmits information between business partners without delay and thus has the potential of balancing demand and supply. In doing so, it generates the conditions necessary for continuous economic growth. This makes B2B connectivity a central issue. B2B connectivity, or the electronic connection between applications of different business partners with a claim to at least partial automation of business processes, offers considerable ratio potential and value-added potential when implemented intelligently. Although not really new and in many companies part of daily operations under the name of EDI (Electronic Data Interchange), B2B connectivity is experiencing a renaissance as business processes are progressively being permeated by new technologies and standards such as web services and BPEL (Business Process Execution Language). BPEL is an XML-based flow language that defines how business processes are to interact within or between companies. It permits process modeling and the immediate execution of the corresponding workflows. Here, classic one-to-one or one-tomany relationships are often expanded to include many-to-many relationships by integrating electronic marketplaces. An increasing number of companies are discovering the enormous potential in the new opportunities opened up by B2B connectivity. In addition, present process landscapes themselves, with their changed and more stringent demands on real-time information, reaction capacity and process transparency, also shift B2B connectivity into the focus of IT considerations. Examples of this are concepts such as Global Sourcing, eProcurement, Collaboration and Real-Time Enterprise. From a technical point of view, B2B connectivity uses suitable middleware to connect separate application systems between companies, such as the purchaser’s material management system with the vendor’s distribution system. The B2B connectivity infrastructure provides all basic services required to handle business processes on the Internet or using classical communication procedures. B2B processes are largely autonomous. They are only loosely coupled

I&C Technologies for a Real-Time Enterprise (RTE)

215

with one another. They are asynchronous, i.e. interconnected by waiting queues, and they react as services with a delay. This leads to robust and scalable systems. Today, applications that conform to these principles are called services. They can become a part of service-based architecture at any time. B2B Connectivity RTE Enterprise Business Process Controller Partnermanagement, Routing, Relationship Common Services (Security, Archiving, Tracking) Common Objects

Transformations

Transformations

Supplier Marketplace

ERP SCM

Adapter EDI

Soap/XML Customer

RosettaNet

Transport Layer

Adapter

EDINT

CRM

Custom Logistics

Figure 49: B2B connectivity architecture

At its core, the B2B connectivity platform provides the following modules, independent of the manufacturer: • Business Process Controller (job streams) • Partner management • Common services (archiving, retrieval, tracking/tracing, security, authentication) • Transformations (IDOC, VDA, eBXML, EANCOM (EDIFACT)) • Adapter for applications (Siebel, SAP, i2, Legacy) • Protocol connection (FTP, OFTP, HTTP, HTTPS, POP3, X.400, RFC) • Web services (WSDL, SOAP/XML) • Industry standards such as RosettaNet and EDIINT

216

Friedrich Kopitsch

Common services (e. g. billing) can be added on a case-by-case basis. They are “mounted” in the process control engine. The following must be noted to fully utilize the ratio and value-added potentials of B2B connectivity: • B2B connectivity is not purely an IT issue. Those participating in the business process being implemented must be integrated as well; • The processes to be implemented using B2B connectivity must be clearly defined for all participants and they must be considered automatable or partially automatable; • Opportunities for streamlining process flows prior to their B2B implementation and eliminating superfluous variants absolutely should be taken advantage of. Here, too, the 80:20 rule applies, i.e. 80% process complexity, 20% technical complexity; • No attempts should be made to use B2B connectivity to electronically “straighten out” processes that are deficient in contents. Projects of this nature, which are usually politically motivated, almost always prove to be inefficient; • The functional (e. g. data throughput) and commercial (e. g. ROI) demands on the solution and its required service life and extendability must be defined; • In business-critical areas, mature technologies should generally be given preference over innovative technologies. Web services

Web-supported services are among the most interesting new developments in recent years and should soon be widely accepted. In increasing numbers, customers are deciding for web services and paving the way for the future. Even today, many IT developers are spending much of their day implementing web services. All major suppliers of development environments and IT platforms support this new technology. However, web services are more than just a development technology. As integral communication mechanisms, they will be providing access to information and services within and between companies in the future. Service-oriented architectures (SOA) will use them to make available their services for centrally-defined process flows and other applications. Thus, a comprehensive view of business processes in the technical departments has a counterpart in the architectures and cooperation of systems in the IT area. Web services use the widespread, standardized protocols of the Internet

I&C Technologies for a Real-Time Enterprise (RTE)

217

and the company-own Intranet for communication. The language they apply is the lingua franca of inter-computer communication, the meta language XML. Structured exchange formats and protocols such as SOAP, WSDL and UDDI are defined on the basis of this language. XML text formats are simple to process and original XML is in the public domain. In addition, it is wholly platform independent and is supported by all major tool manufacturers. Web service calls are transmitted using SOAP (Simple Object Access Protocol). This standard protocol can be used to request and provide services anywhere in the world, even penetrating firewalls. Web services are explicitly described using WSDL (Web Service Description Language). This automatically permits them to be used from applications or other web services, independent of the server platform on which the web service runs and the programming language used to write it. In this way, a large number of simple services can be combined to form complex new services. If, moreover, the new Business Process Execution Language (BPEL) is used for controlling web service calls, it will be possible to define and automatically execute comprehensive business processes. In addition to describing and executing processes on the basis of a future standard, BPFL also makes it possible to describe process interfaces to business partners and customers. This permits business processes to automatically flow across company boundaries – without disclosing internal company information. Web services continue to gain in importance as a means of optimizing business processes. This is apparent not only in the widespread support they enjoy in industry. It is also demonstrated by the substantial efforts being made to provide the technical facilities needed to meet security and transaction requirements that are not yet available. This gap is currently being filled by Microsoft, IBM, SAP and other IT providers using Global XML Web Services Architecture (GXA). It employs a series of individual specifications to cover the most important aspects of a comprehensive security architecture. Additional specifications deal with matters such as transactions and the reliable transmission of messages. The modular definition by means of individual specifications follows the maxim that not all mechanisms are necessary in every context. This prevents avoidable web service complexity and keeps the learning curve for this technology at a minimum.

Grid Computing

Trust

Authorization

Privacy

Application areas

Reliable Messaging

WSRP

...

Secure Conversation

BPEL

Friedrich Kopitsch

Policy

218

WS-Transaction

... WS-Coordination

Global XML Web Services Architecture (GXA)

WS-Security UDDI

WDSL

Technical basis

SOAP XML

Web (HTTP, SMTP) Internet/Intranet (TCP/IP)

Figure 50: The “Web Services Stack” showing the most important specifications and standards

Meanwhile, providers of another technology who were long considered to be web services’ greatest rivals have discovered the services for their own purposes. Without exception, the manufacturers of EAI (Enterprise Application Integration) products now offer web service adapters and integrated web service development environments. This also reflects a change in the “classical” EAI: The emphasis is no longer on access to stand-alone applications and broker-based data adaptation and distribution. Instead, EAI is becoming the primary entity in cross-application process description and execution. EAI specific adapters and mapping mechanisms will continue to be used for “old applications”, but for new business applications created to the specifications of a service-oriented architecture (SOA), it is web services that are being implemented with ever greater frequency. They are used to control business processes by activating individual process steps as services via the web service. Also, suitable wrappers are employed to access the functions of Legacy applications, thereby securing the investment. Today, this transformation of EAI is often described in terms of the concept of Enterprise Service Bus (ESB): An ESB is a logic bus that treats all applications as event-controlled services. It can be extended flexibly – even across company divisions, business partners and customers – and controls the connected services in the form of web services. With global directories such as UDDI (Universal Description, Discovery and Integration), they can

I&C Technologies for a Real-Time Enterprise (RTE)

219

be made generally accessible and can be automatically located (discovery) and executed. For example, this mechanism enables a product manufacturer to select and contract a suitable shipper in the customer’s vicinity to deliver the products. The division between process description and execution on the one hand and web service calls of process functionalities on the other improves transparency in the implementation and execution of business flows. Business processes become more readily adaptable to new business requirements and are easier to monitor and evaluate on a process-specific basis during running operations. A suitable implementation of web services is used to achieve asynchronous, uncoupled communication in “near real time” between the separate IT systems and components. All relevant data are collected and evaluated and undergo further handling by the processes in “near real time”. Thus, the company will find itself on the road to becoming an “event-driven company” or, as Gartner puts it, a “Real-Time Enterprise“ (RTE). When a suitable overall architecture and a corresponding implementation is being selected, web services of today are already in a position to provide the basis for a future-oriented IT infrastructure. By virtue of their connection with applications according to the SOA concept and new standards of process description and execution such as BPEL, they are able to reduce the complexity of company IT, thus helping to closing the gap between the technical description of business processes and their implementation. This is a prerequisite for supporting new, more advanced concepts such as Business Process Fusion. By directly linking operative business flows with company management, control and optimization of business flows should improve. Service-oriented architecture (SOA)

Requirements

A primary challenge for real-time enterprises is to react with great agility to the requirements of company-critical business processes, which are changing in ever shorter cycles. This requires rapid, efficient and low-cost integration of new process elements in existing process chains or the implementation of new process chains. The necessity for adaptability applies not only to the entire scope of inter-company business processes, but also to processes that integrate external company resources.

220

Friedrich Kopitsch

Technological concepts

Service-oriented architectures are a future-oriented approach that meets these requirements. It generalizes the concept of web services, which form the technological core component for service-oriented architectures. Web services are defined as a group of technologies that permit an application function to be called up on the Internet as a service according to underlying standards (XML, SOAP, WSDL, UDDI). Thus, web services provide the basic technology need to harness the Internet for intra-company and global business processes. However, web service technology is not all that is needed for a serviceoriented architecture, which goes beyond this in a number of aspects: • Service-oriented architectures are process-driven, i.e. services for an architecture of this type are the building blocks for the orchestration of company-wide business processes. Thus, a service-oriented architecture always contains a layer for the explicit modeling and execution of business processes. • Services in a service-oriented architecture have an effect on the value of a company. A service always provides a partial business process, such as checking a credit customer according to BASEL II regulations. • Services are described and published according to standardized vocabularies and ontologies (OASIS, RosettaNet, BizTalk. org,..). This ensures that services will be interoperable from the perspective of business processes. Consequently, a service in the context of a service-oriented architecture is an application function that can be used in a number of ways – in the context of different business processes and companies. It can be made use of whenever required, brokered by a publication/subscription mechanism based on a standardized business vocabulary. Using this mechanism, a socalled contract is established between the user of the service and the service itself. This type of contract –a description that is independent of the implementation – is an abstraction of the service from the perspective of the business process and thus goes beyond classical interface concepts for the use of components in distributed environments. Thus, services correspond to dynamically-usable business functions.

I&C Technologies for a Real-Time Enterprise (RTE)

221

Example of an architecture: Integration of company processes

The figure below shows the application of a service-oriented architecture on company processes in a three-layer model. In order to qualify as a service-oriented architecture, the following criteria are decisive: • Service-orientation: all applications are addressed as services; • Process orientation: the business process logic as such (in contrast to the services, which implement the individual process steps) is executed in a separate layer; • Embedded SOA components: individual elements in the applications and operations layer of this architecture in turn satisfy the SOA and thus enable a mapping of business processes that is direct and close to the original in structure.

Business process

ARS

GPS

ABS

Business process

Existing (Legacy) application Existing company application ARS

Service interface

Business process orchestration ARS

GPS

BAS

Business process request

Service interface

Business process

GPS

Business process

BAS

Business process request

Applications layer/Resource layer Service interface

Business process layer

Service interface

Operations layer

Existing (Legacy) application

Figure 51: Service-oriented architecture for company-wide processes Benefits and future outlook

Service-oriented architectures meet demands for flexibility in the integration of business processes. Thanks to their greater efficiency when it comes to the IT-technical implementation of business process integration, they also lower the integration costs related to individual services. This is also supported by the mounting level of tool support now available for an SOA

222

Friedrich Kopitsch

programming model on leading application and middleware platforms. In the coming years, service-oriented architectures will develop into the dominant paradigm in the field of business processes integration. Although not all technological elements are yet available or in part do not yet feature the required function, the great economic potential of the technological concept of service-oriented architecture ensures that its use will lead to a strategic business advantage in future. Wireless LAN

Wireless LAN is a wireless extension of a Local Area Network. This technology enables access to the Internet or to company data and applications without the need for cabling. How does WLAN function?

• Communication takes place on standardized and non-billable data transmission frequencies in the 2.4 to 2.484 GHz waveband. Standardization is carried out by the IEEE (Institute of Electrical and Electronics Engineers); protocol standards are described under IEEE 802.11a to i. • Transmission speeds range up to 11 Mbit/sec (802.11b); in the near future, they will increase even further to 54 Mbit/sec (802.11g). • The transmission distance spans from 30 to several hundred meters, depending on the physical environment (buildings, unobstructed terrain). How can a WLAN be used?

The WLAN standard is a specification for two types of usage or modes of operation: the network infrastructure mode and the so-called “ad hoc mode”. • “Ad hoc” network (also known as “peer-to-peer mode”) In this type of usage, wireless communication partners such as Notebooks (also known as clients) are directly connected with one another. A network infrastructure is not required. This is a very cost-effective form of communication for a limited number of participants with a limited range (e. g. for small and medium-sized companies). It is also a flexible solution for interactive presentations and workshops.

I&C Technologies for a Real-Time Enterprise (RTE)

223

• Network infrastructure mode In this case, communication with the client (e. g. Notebook, TabletPC, PDA) takes place via a WLAN access point. The access point is the connecting element to the wired corporate network. A connected WLAN client is fully integrated in the network and has the same access and communication options as a “wired” client. The access point is used to access any network infrastructures, i.e. LAN (Local Area Network) or DSL or ISDN telephone connections. A client can switch between different access points without interrupting the connection (WLAN Roaming).

Access point

LAN

wLAN

Figure 52: WLAN within a network infrastructure

• Flexible use of WLAN and mobile wireless networks The use of WLAN communication is limited by the available infrastructure (range of WLAN access points). Many business people, forwarding agencies, courier services, service technicians, sales people, etc., require permanent access to the corporate network from outside of the WLAN network. For this purpose, a modern communications management solution will permit uninterrupted access to the Internet using the wireless network available at the particular location – either WLAN or the network of the mobile network operators (GSM/GPRS, HSCSD). For the user, “switching” is automatic and does not need to configured. The data transmission rate depends on the speed of the particular network, e.g. max. 40 kbit/s for GPRS in contrast to up to 11 Mbit/s for WLAN. This type of communication management is called “seamless connectivity”.

224

Friedrich Kopitsch

The advantages of WLAN

• In buildings, cables no longer need to be laid through walls and ceilings; on the company premises, communication with the company network is possible from any location at a high data transmission rate (e. g. from the factory grounds, the assembly shop, warehouse management and shipping). Outside of the company, WLAN networks (“hot spots”) are being set up in increasing numbers, especially at airports and train stations, in hotels and conference centers, etc. – that is, in locations where business travelers frequently spend their time. • Wireless LAN technology can be operated as a self-sufficient network but still fits easily into almost all existing network topologies, extending these by flexibly connecting “mobile workplaces”. • Company employees have the opportunity of accessing company data and applications at any time from anywhere within the company or in public areas using “hot spots”. • WLAN is an ideal supplement to LAN networks; in the long term, the cost of setting up and operating a Wireless LAN can be much lower than that of a “wired” LAN network (TCO – total cost of ownership). This particularly applies in the case of frequent reorganization and relocation within a company. The technology also provides optimal support for the flexible use of the workplace infrastructure by employees who are frequently away from their desks (“Flexible Office”). • Employee productivity can be significantly augmented since the company network can be accessed from any location. This significantly greater flexibility in carrying out assignments permits a high level of productivity, especially in locations in which wired communication is difficult or even impossible to implement, such as on the shop floor, in warehouse management or in the shipping department. Wireless LAN permits an extremely high level of flexibility in employee communication with the corporate network. Thus, it is a prerequisite for achieving increased productivity and new forms of workflows. Wireless messaging

It is becoming increasingly important for communication within a company and with suppliers and customers to be rapid. This demands that relevant information be made available independent of location and time. Today, a large part of corporate information is not found in a structured form in

I&C Technologies for a Real-Time Enterprise (RTE)

225

applications such as SAP R/3, but in the form of documents, E-Mails, calendar entries, task descriptions and faxes. Quick access to this information is important for rapid decision-making. Therefore, users must be able to access all information from any location, workplace or device, regardless of whether it is structured or unstructured. One approach to realizing rapid communication, or access to unstructured information, is wireless messaging. In this environment, employees in a real-time enterprise have access to relevant data from any terminal device (Notebook, PDA, mobile phone) in every situation. These data include: • Current calendar entries (synchronized calendar) • Project-related status information • Immediate access to E-Mails/sending of E-Mails • Access to documents • ... Access is implemented on the transport medium level using technologies such as GSM, GPRS and WLAN infrastructures. UMTS will also be coming into play more frequently. Building on these basic technologies, secure access to the corporate network will be required on the basis of VPN networks to ensure direct and secure access to the required information. Today, these technologies are already largely available, although they are often not yet entirely integrated. One example is the lack in integration of WLAN and GPRS networks. A user must still know in advance over which transport medium access to the corporate network will be granted in a particular situation (for example, via a WLAN hotspot at the airport but only via GPRS in the taxi on the way to airport). This situation still demands detailed knowledge on the part of the user. Roaming between access services is not yet supported. However, devices will soon be available on the market that will support access over various transport media and implement it depending on situation, cost and bandwidth. Instant messaging

In addition to the well-established communication routes of telephone and E-Mail within companies and to customers and suppliers, a third pillar is gaining in importance and has already become widespread in the Internet. It is known as instant messaging and is frequently used by teenagers for

226

Friedrich Kopitsch

“chatting” in the Internet. An important advantage of instant messaging solutions is seen in the ability of such an infrastructure to increasingly prevent travel costs and thus save on employee travel time. According to Gartner, instant messaging solutions will be playing an important role in the continued development of company IT infrastructures. It is projected that in the near future, instant messaging will figure in corporate communications as E-Mail does today. Furthermore, this new type of communication can be readily integrated in existing IT landscapes and supplements the functions of telephone and E-Mail. Basic user features are: • Information on who is currently available/accessible (so-called buddy lists). The user can store information on his or her status (in meeting, at the desk, at lunch, etc.). The information is then made directly visible to colleagues who have entered the user in their specific contact list. The user generally keeps specific contact lists for his work (e. g. project members, colleagues). • Option of online collaboration on the basis of Netmeeting. Building on pure instant messaging, a solution is created that makes it possible to work together on documents Figure 53: Status information of comand to exchange data. • Possibility of building simple and munication partners cost-effective conferencing solutions for voice and video communication. • Using Outlook or other portal solutions (such as HiPath OpenScape from Siemens), direct instant messages can be initiated with the members of a particular team/project. In the last few years, IBM and Microsoft in particular have established themselves as important driving forces on the market that have expanded

I&C Technologies for a Real-Time Enterprise (RTE)

227

their own E-Mail/messaging platforms by these functions. These solutions especially address the demands that companies place on instant messaging. Based on experience made with older corporate communication platforms (E-Mail, telephony), a sharp rise in this form of communication can be expected over the next two or three years. In general, investments can be assumed to be fairly low. Modern terminals (Desktops/Notebooks) are usually already equipped with the technology necessary for voice conferencing. Suitable back-end server infrastructures must be provided. Location-based services

Determining the exact position of the caller is among the more interesting mobile network services aside from telephoning itself. Location-based services (LBS) therefore are considered to be the key to the future mobile business market. Once the mobile terminal knows the exact position of the carrier, a wealth of regionally-based services become possible: from targeted assistance in car breakdowns, locating missing children, to the control of vehicle fleets or technicians. Location-based services are subdivided into four general areas: • Location-based information Typical applications of location-based information are, for example, user guidance to a selected location along the shortest path, or user solicitation according to his or her current location and interests. • Location-sensitive billing Location-sensitive billing can be used to handle payment of road and parking fees and tickets for events or even for municipal transportation. • Emergency services If an emergency call is placed, the location can be transmitted to the emergency call center, or in the event of a vehicle breakdown, the auto service can be guided directly to the vehicle’s location (in the USA, the Federal Communications Commission has ordered mobile network operators to transmit the following information to the Public-Safety Answering Points (PSAPs) in the event of an emergency call: call number, mobile cell and position within the mobile cell). • Tracking Tracking services can be used, for example, to display whether someone on a list of friends is in one’s vicinity, a feature that naturally requires user permission. The same mechanisms can be used to determine fleet vehicle positions or to optimize service runs of technicians. The poten-

228

Friedrich Kopitsch

tial user is any mobile network user equipped with a micro-browser. A study by the Ovum consulting and research company projects a total of 1.5 billion mobile network users worldwide by the year 2006. Of these, 684 million people are expected to employ services based on microbrowsers. All mobile network operators already offer their customers comprehensive location-based services. The potential area of application is broad: hotels, restaurants, filling stations, event organizers of any kind and merchants of any sector will stand to profit, as will banks, insurances, transportation companies and the health care sector. For companies, adapting LBS technology opens entirely new perspectives: • • • •

Additional information and sales channel Generation of new business transactions Tapping into new information technology (push services) Access by mobile employees to company-related and business related information • Optimized control of field service runs A prerequisite for the successful utilization of LBS functionality is the ability of LBS solutions to be integrated into existing corporate IT and application infrastructures. Connecting to the company databases and data warehouse is the precondition for integrating the technology into the existing business processes. Integrated location-based services enable companies and organizations to stimulate their business by taking advantage of the additional mobile marketing, sales and service track. What’s more, they can expand, fine-tune and personalize marketing, sales and customer service activities in combination with existing customer channels. With its location-based services platform, Siemens Business Services (SBS) has developed a reusable solution core that forms the basis of sectorindependent LBS applications, and that can be readily integrated in existing application architectures thanks to the technologies it employs. The solution contains: Basic LBS functionality • Automatic localization • Search for address • Search for nearest location • Display of location-based information

I&C Technologies for a Real-Time Enterprise (RTE)

229

• Display of road maps • Routing (by foot, by car) Architectural and platform concept that is characterized by: • Pre-developed reusable solution components as an LBS platform that can be adapted to customer requirements on a project-specific basis • Current, flexible and future-proof solution architecture based on open standards (HTML, WML, J2EE, XML, XSL, ...) • Platform-independence through use of SOAP • Access and transmission security through encryption • An interface to differing localization systems and providers • Possibility of use under .NET and JavaONE The following figure shows an overview of LBS platform architecture from SBS with interfaces to the external “mobile network operator” service – for determining geographic position – and to the “Geo-Service”, which makes available address data and map material with routing information.

Application Server

Carrier Positioning Data

• Dynamic presentation • Templates Web Server

Gateway

Geo-Service

Business Logic

GeoCoding Maps Routing

Adapter to • Carrier • Geo-Services • Enterprises

Enterprise

DB Server Location Data

Actual Data

Figure 54: LBS platform architecture

The LBS platform implementation is entirely based on open standards (Java, XML, XSL, WML, JDBC/ODBC, SOAP, ...) and, as a web service, can be used by distant applications. The platform supports GSM, GPRS and the future UMTS network. By applying WML standards, the platform

230

Friedrich Kopitsch

achieves independence from the type of mobile terminal. Consistent use of open industry standards ensures that the LBS platform can be implemented on nearly all current systems and middleware products. Real-time analytics

A new concept is currently experiencing a rise in popularity: “Real-Time Analytics”. In principle, two trends underlie this concept: On the one hand, a type of Business Intelligence (BI) workflow is superimposed over the customer’s normal business workflow. On the other, the individual business processes require or create current BI information during running operations. While in the past BI data were entered and compressed on a batch-by-batch basis in central data warehouses (Operational Data Stores), this information is now entered into the data warehouse at the time of creation, and is immediately made available to other business applications when necessary. To ensure availability of this business intelligence, which is always near real-time and continuous, two key technologies must grow closer: • EAI (Enterprise Application Integration) • ETL (Extraction Transformation Loading) Leading software providers are already reacting to this trend, which also goes by the synonym of “Zero Latency Warehouse”. SAP offers an “Intelligence Connector”, which joins workflow data and business warehouse data in equal measure and makes them available to other decision-making processes. As well, classical ETL providers such as Informatica are already offering add-ons for processing so-called real-time sources. In the open relationally-oriented data warehouse approach, the ETL and EAI processes have up to now coexisted on a limited but well-defined basis. The wish to make use of the full range of strengths offered by the entire process chain (integration, transformation, loading and analysis) has become stronger with the advent of new possibilities for real-time analytics. ETL technology was initially developed as a solution to the most critical point within a data warehouse application – namely the joining of data from different source systems for various analyses. Typically (but not exclusively), these sources are operative systems and, thus, connected databases.

I&C Technologies for a Real-Time Enterprise (RTE)

231

Real-time analytics can be used to produce a timely representation of a wide variety of business situations. Historical data from company-wide systems – such as ERP, CRM, SCM or Legacy and eBusiness applications with real-time data from EAI systems and messaging applications are drawn on to deliver company-wide perspectives in a matter of seconds. In addition, alarm systems that immediately inform those concerned even while the event is still in process can be readily integrated. The value of real-time analytics lies in the fact that overviews of current business information are available more rapidly than in conventional ETL applications. On its own, EAI is unable to furnish the entire scope of real-time features needed for effective real-time analytics. It lacks important pre-conditions such the ability to handle large data volumes and transformation and metadata capabilities. Conventional ETL solutions are unable to supply realtime analytics because they do not have integration capabilities for extracting transaction data at the end of a process chain. However, a properly equipped ETL platform can go a long way to supporting EAI by continuously integrating data from real-time sources. ETL enables access to other complex system, permitting EAI to supply users with a consolidated, company-wide view. ETL data integration Efficient moving/transforming of large data volumes Session-oriented Analysis-driven Based on metadata Transforms and standardizes data for analyses View of current and historical data Good scalability Near real-time analytic capabilities

EAI application integration Efficient moving of transaction messages Transaction-oriented Event-driven Uses few/almost no metadata Offers high transaction security of messages Automation of business processes Poor scalability Event-driven real-time capabilities

Because of enormous performance requirements in some cases, integrating these technologies presents itself as a challenge for “system integration”. Intertwined with these new technologies are the important topics of XML connections, database tuning, portals and “single sign-on”, in addition to the issues already mentioned above.

232

Friedrich Kopitsch

Mass RFID tagging

Over the last several years, intelligent labels, or RFID (Radio Frequency Identification), have embarked on a victory march. This leap in the success of an already familiar technology that had already been implemented many times over can be attributed to two factors: • The price: after beginning at two-digit Euro values, chip prices have now dropped down to the two-digit cent range and manufacturers project a price of two to five cents in the year 2005. This price will also make possible the worldwide use of chips in all products throughout the logistic chain, trade, etc. • Standardization: With the ISO 15693 and ISO 18000-4 standards, this technology can for the first time be employed in cross-company implementations. Without standardization, its use would not have gone beyond local applications in proprietary sys- Figure 55: RFID label tems. These improved conditions open up the following areas of application: • Logistic and transport companies • Wholesale und retail • Tracking and tracing of goods • Insurances, theft protection, ... • Reusable containers • Asset management, inventory, warehousing, ... • Event management • Security-related systems (access, passes, visa, ...) This essentially applies to all industries – from automobiles to textiles. A large number of projects has already confirmed the soundness of the approach and the successful application of the technology available today. And the return on investment (ROI) has been verified. By way of example, three areas of applications will be discussed in detail below:

I&C Technologies for a Real-Time Enterprise (RTE)

233

Document tracking

Paper-based archives are still among the most important sources of information in many organizations, and they often have to be maintained in parallel to an electronic archive management system for legal reasons. For this reason, great emphasis must be placed on optimizing the archival system and keeping records of persons taking out materials. Figure 56: Document with RFID For purposes of “Document Track- label ing”, new files and documents and those in existing archives are provided with an RFID chip (smart label). These RFID chips are equipped with a ‘personalization client’, i.e. an identification key. Persons taking out materials pass the documents through a reader at the office entrance that is connected with an accounting server. All authorized persons have access to the server from their PCs and, when necessary, are immediately able to track the required documents. Advantages: • Simpler labeling and archiving. • Cost savings through the elimination of a time-consuming search for documents. • Generation of a competitive advantage through faster availability of information. • Increased customer satisfaction due to timesaving and faster information Figure 57: Removal of document availability, leading to greater cus- equipped with RFID tag tomer loyalty. Asset tracking

Receiving, shipping, warehousing, etc., whether local or globally distributed, are integral components of the supply chain management processes. Therefore, goods received, stocks and good issued must be recorded whenever opportune. This is accomplished by means of “Asset Tracking”. The

234

Friedrich Kopitsch

assets (goods, pallets, systems, components) are “personalized” by an RFID chip, making them uniquely identifiable. The RFID chips are provided with an identification key using a ‘personalization client' and are then applied to the object. Outward movements are recorded by stationary or mobile readers. All data are kept on a central accounting server that all authorized person can access from their PCs, enabling them to track items such as archived documents or warehouse goods. By integrating such smart label concepts in comprehensive SCM solutions, goods transfers can be reliably recorded. This is also an important requirement of insurers for the Figure 58: Goods labeled with unambiguous proof of responsibilities. RFID tag Event management

On the basis of intelligent tags, RFID can be used to create an event management solution that covers all activities, including sending out invitations, checking in participants, entrance control, recording exhibit preferences, sending out information materials and final processing after the event. Exhibitors use the lead management component to gain an overview of all participants at the fair and the exhibits they have visited. Trade fair contacts can immediately be recorded online and are thus available as a database for later market evaluation. The core of event management is a transponder consisting of a microchip and an antenna. It is integrated in the trade fair pass and can be read without contact. Data is transmitted on the basis of high frequency technology. All system functions can be activated via the trade fair pass. There are three options for storing visitor data on the pass: • Before an event, customer data of the registered visitors are assigned to the pass. This is accomplished by storing the information on the RFID chip. • In the reception area, a web solution is employed for the rapid late regis-

I&C Technologies for a Real-Time Enterprise (RTE)

235

tration of visitors. PDAs (Personal Digital Assistants) are used for mobile registration when checking in and out. • In the exhibit area, PCs with transponder readers see to it that information requested at the exhibit is either handed out to the visitor when checking out or is sent out by post. Advantages from the perspective of the event manager: • Automated ascertainment of customer interests • Simple late registration of customers in the reception area when printing out trade fair passes • Electronic provision of trade fair leads • Image enhancement due to innovative event management Advantages from the perspective of the trade fair visitor: • Current information on the participation of contacts at the fair • Information on which booths and contacts have been visited (tracking list) • Collection of information without having to carry around piles of papers • Booking and reservation opportunities on location Summary

As the examples above show, a real-time enterprise uses technologies to bring about a lasting increase in the speed and quality of information flow, thereby improving business-critical processes both within the company and between companies, with present and future business partners. For corporate management, the challenge is to modify current business flows and infrastructures in such a way as to transform the company into a real-time enterprise. Today, companies such as Dell, Cisco, Intel and others are pioneers in the field, and further companies will be following in their footsteps in the near future.

V. Enterprise security

The proliferation of networks within and between companies allows centralized information resources to be accessed via many different channels. This makes IT security an essential task but one that is becoming increasing complex. For example, the number of potential targets has grown significantly in recent years and the types of attack as well as the channel is continuously changing.

Claudia Eckert

Secure mobile business solutions for real-time enterprises

Mobile business solutions are becoming indispensable components of realtime enterprises. Mobile services and applications open up new business opportunities for companies and can contribute to more efficient and effective use of resources, allowing considerable cost savings. Profitable use of mobile business solutions requires a modern, open company IT infrastructure that should, however, be a matter of course for real-time enterprises. An open IT infrastructure that can also be used on the move allows customers direct access to services offered by the companies. As well, business partners, suppliers and mobile employees can access the company network and its resources at any time. Apart from the resulting advantages for the company, however, this openness places high demands on company security, centrally important to modern networked companies. This article illustrates the opportunities for modern companies that can arise from the utilization and systematic integration of mobile business solutions. In particular, it describes the possible security risks to be kept in mind with regard to the use of mobile technologies, and points out solutions that permit the secure integration of mobile technologies in company architectures.

Introduction The growing use of the Internet for private and professional purposes together with the new prospects provided by modern telecommunications are leading to major changes in our working environment and everyday life. With mobile terminals such as laptops, mobile phones and PDAs (Personal Digital Assistants), new types of communication such as GPRS and UMTS, and wireless networks such as WLAN, employees have access to company resources from almost anywhere and at any time. While today mobile working is generally limited to mobile wideband access to compa-

240

Claudia Eckert

ny data, the next step is to exploit the potentials of existing mobile technologies in order to consistently refine eBusiness solutions into mobile business solutions and to establish them in real-time enterprises. Mobile business solutions include mobile Customer Relationship Management (mCRM) solutions as well as mobile office and even tracking applications such as mobile fleet management. According to a study by Berlecon Research, the market for solutions for mobile working will increase to 1.5 billion Euro by 2005. Mobile business solutions are becoming essential constituents of modern companies. They are leading to a transition to more flexible company processes, and processes that can be implemented more efficiently (e.g. no media changes in procurement procedures) and more quickly (e.g. contract copies can be produced directly on site at the customer) and in real-time around the clock. Hence, mobile working opens up interesting perspectives for future IT applications. At the same time, however, that these technological developments also result in the further opening of networks is problematic because it makes companies increasingly vulnerable to attack. This means that company security plays a central role in the establishment, acceptance and, hence, also in the commercial success of mobile business processes. Safe mobile communication with customers, business partners and company networks requires multilateral secure end-to-end communication that guarantees authentication, integrity, confidentiality, commitment and availability. Mobile business processes encompass much more than just mobile communication. Mobile devices and applications must be integrated into existing processes and infrastructures as seamlessly as possible while ensuring that company security is not jeopardized by their integration. It is immediately apparent that perimeter security with the installation of a central firewall and virus checks as is usual in present day companies is no longer sufficient to also secure mobile processes. This chapter deals with mobile business solutions and their integration in existing company infrastructure and highlights the potential that this offers modern companies. We then address the problem area of company security using a rudimentary business architecture to show the standard measures that are implemented for technical security in today’s systems. Mobile technologies are subject to additional security risks that we will study in detail and for which we will present various methods of resolution. Finally, we will show that the treatment of company security requires a holistic approach, essential for modern companies.

Secure mobile business solutions for real-time enterprises

241

Mobile business processes It is hard to imagine everyday life in today’s world without mobile communication. Figure 59 presents an overview of the technologies in use today for mobile communication and for mobile networking. This includes mobile access to the Internet via mobile phones as well as network access via a “classic” modem at the home office or via Hot-Spot access, e.g. at the airport.

UMTS/GPS

UMTS/GSM GSM/ UMTS/BT

PAN

PDA BT/WLAN

PC

UMTS/GSM

Laptop WLAN/BT

WLAN

IP Phone

Sat-Uplink

WLAN Access Point

Ethernet Switch DECT Station

Phone

VoIP fähige PBX

WAN Switch DECT

Packet Oriented Circuit Switched

Gateway PSTN

Figure 59: Diversity of mobile, wireless and wire-connected communication networks

Figure 59 illustrates the diversity and heterogeneity of today’s networks. These include satellite networks, mobile radio networks such as GSM/UMTS, public telephone networks (PSTN) and wireless local traffic networks such as WLAN to Bluetooth networks (BT) for coupling devices of the personal working environment (PAN Personal Area Network) such as PDA, laptop or mobile telephone. As well as mobile and wireless networks, there are cabled fixed networks such as Ethernet and fiber optic networks. As the figure shows, the various networks already have multifaceted connections with one another today. A huge challenge for future networks lies in the development of new concepts, processes and technologies and enabling and improving the inter-working of heterogeneous networks, i.e. cooperation despite administrative and, in particular, technological differences. A substantial development in this case is the integration of wireless and mobile networks in wired wide area networks to support unreserved

242

Claudia Eckert

communication and growing workforce mobility. However, the specific problem of network convergence and integration will not be discussed below. Interested readers are referred to Eckert, C., Bayarou, K. 2004. Mobile technology

Mobile terminals such as laptops, mobile telephones, organizers and PDAs are becoming an integral part of our everyday lives. PDAs with their diverse functions such as office packages, E-Mail services, web browsers and even electronic purses are becoming ever more popular. They provide a mobile working environment familiar to the user from the PC. In addition to cradles, which enable a fixed connection to the user’s PC, communication with the outside world is also enabled by WLAN, Bluetooth and GSM/GPRS interfaces for connection to mobile telephony. Of course, mobile telephones also count as mobile terminals. Where only a few years ago the only mobile phones were simple models that merely supported voice transmission, mobile phone technology has continued to develop rapidly. Development steps included tri-band devices that enabled data and voice services in both European and US networks, the provision of wireless transmission options for short distances such as Bluetooth and multimedia-capable mobile phones and smart phones with color display and a large number of applications similar to PDAs. Mobile and wireless communication

Mobile networks straddle countries and continents, enabling the virtually unfailing reachability of communication partners. An important role is played by the GSM (Global System for Mobile Communication) standard, a second-generation (2G) system that extends to all continents. When a user switches on a mobile telephone, the phone logs onto the respective network via the nearest mobile phone antenna with sufficient signal strength. The network checks whether the user is authorized to use its services and establishes the connection. GPRS (General Packet Radio Service) is a technology situated between the second and third generations. Because GPRS transmission is packetoriented, charges are based on the transmitted data volume rather than duration as with classical telephony. Consequently, it is extremely suitable for the transmission of smaller quantities of data or partial and compound data streams such as E-Mails. Now UMTS (Universal Mobile Telecommunication System) is ushering

Secure mobile business solutions for real-time enterprises

243

in the third generation (3G) of mobile systems. UMTS offers a distinctly wider bandwidth than both preceding technologies and enables audio and video streaming (e.g. suitable for video conferencing) as well as extended telephony capabilities (RichTelephony) with simultaneous transmission of voice, image and sound. Each GPRS-capable and UMTS-capable terminal has its own IP address and is, therefore, a computer in the Internet (‘always on’ paradigm). Wireless local networks or WLANs (Wireless Local Area Networks) are normally based on the international IEEE 802.11 standard. In specifications 802.11a, b and g, it allows data communication via the license-free ISM (Industrial, Scientific, Medical) frequency spectrum between 2.4 and 2.48 GHz. Mobile terminals communicate with an access point via an air interface which itself is connected via a wired network to other networks. An access point sends out signals at regular intervals so that mobile devices within its coverage range can make contact. The signal range is between 30 and 150 m. In contrast to infrared beams, the radio signals can penetrate solid objects and walls so that a region with a relatively large spread can be covered with a WLAN. Although this unhindered signal spread and the resulting data transmission capability support the great variety of applications of WLAN technology (unimpeded network access to many locations), it also however renders such networks very much more vulnerable than infrared networks, for example, which can only cover up to one meter. Mobile business solutions

Mobile business solutions is the term for technologies and applications that allow the transaction of business independent of location and time of day, and where any mobile devices and mobile network technologies can be employed. Mobile services and applications enable modern companies to open up new business opportunities. They can contribute to more efficient and more effective use of resources, resulting in considerable cost savings and process acceleration. In addition to omnipresent accessibility that guarantees high topicality of employee actions, mobile business solutions also have the particular advantage of significantly improved workforce reachability, a feature that contributes to the ability to act in real-time and to react quickly to bottlenecks. Further added value is derived from increased productivity of the mobile employee, improved customer loyalty through process acceleration and increased work process quality, benefits that are brought about by the widescale automation of processing steps made possible by a minimization of

244

Claudia Eckert

changes in media and subsequent transcription errors. According to a study by the Boston Consulting Group, implementation of mobile solutions in companies is expected to generate an increase in productivity of USD 520 billion by 2006. In addition to the pure business environment, mobile services play an important role in the leisure and entertainment sectors, a field that will not be covered here. Examples of mobile business solutions

Examples of mobile business solutions include support of the mobile sales force, field staff and service technicians. Mobile sales support (Mobile Sales Force Automation) can include providing the employee with up-todate product information, dispatching completed customer orders directly into the local company processes en route without time delays and without the need to carry out any additional manual operations, or effectively supporting the mobile employee with travel services such as flight information, bookings and route guides while he is traveling. Mobile solutions can also be used to locate service technicians at any time so that the system can optimally deploy technicians (resource-optimum scheduling) for service orders. In the run-up to the customer visit, the technician can retrieve the information required for repairing the configuration at the customer premises by means of mobile access to local company data. If necessary, the technician can have the spare parts required for repair on the basis of the available fault report transported to the customer or can purchase the required parts himself at the nearest authorized dealer using Location Based Services (LBS), for example. In order for this procurement process to be seamlessly integrated into company operation, an order for electronic invoice preparation can be sent to the company’s internal resource planning system (ERP) simultaneously with the mobile order. Thus it is no longer necessary for the mobile employee to fill out paper forms that would then be transferred in a further manual step to the ERP system, adding yet another source of error. The mobile terminal can also access the other databases in the company, making it possible, for example, to forward the required service report directly to the connected ERP system in the company after the repair is completed. The system processes the data further by, among other things, initiating preparation of an invoice and, if necessary, issuing a repair order for a defective spare part. Mobile business solutions must be integrated as seamlessly as possible into existing and future company-wide and cross-company business

Secure mobile business solutions for real-time enterprises

245

processes so that their implementation requires few or no additional resources and induction costs. Integration of mobile business processes in an eBusiness architecture

Integration of mobile applications presumes a structured architecture such as that currently found in modern company infrastructures. Such modern architectures (see Wiehler, G. 2004) normally consist of different levels (multi-tier) such as the communication and presentation level, the level of business logic, the integration and service levels and the resource level (see Figure 60). Communication and middleware services form the front-end of the architecture while business process implementations and company services and resources form the back-end. Standardized interfaces facilitate the simple integration of additional services in either the front-end or the back-end, and allow the layers to communicate among themselves, thereby achieving integration of the business processes. The separation between front-end and back-end services is also relevant to security issues. This is covered in section 3.

(Mobile) Terminals Mobile and wireless access networks: WLAN, GSM, GPRS, UMTS

HTTP, SOAP

HTTP, SOAP

Mobile access

Web-Server

Portal server

Communication and presentation level

Portal engine, presentation services User Mngt

Security service

Personalized areas

Application server

Business logic

Web services, Transaction services, Load distribution ... Integration server

Integration level

Message broker, Web services, Workflow engine, ... Directory service

Mail server

Web contents

Databases

Legacy systems

Services and resources level

Figure 60: Multi-tier eBusiness architecture

The communication and presentation level includes the communication between client components (e.g. mobile devices, customer and supplier systems) and the presentation of information, usually via a browser inter-

246

Claudia Eckert

face. Essential components of the architecture are web servers and protocols for mobile access. Among others, these include the services of a WAP gateway, which takes over the task of converting mobile protocols such as GPRS into Internet protocols and preparing HTML contents tailored to the display capabilities of mobile terminals. A company’s ‘legacy’ systems as well as company-wide services such as E-Mail, databases and directory services reside on the service and resource level. The heart of a modern architecture is formed by the business logic and the integration level with portal and application servers as well as integration servers. Portal server

The portal server is the central entry point for company-relevant business processes and is used in equal measure by customers, suppliers, business partners and, of course, (mobile) employees. The tasks of a portal server include analysis of incoming requests, starting of components required for processing and feedback of results (Portal Engine). A device-specific preparation and presentation is carried out via integrated coding and rendering processes. Roles, rights and, if necessary, certificates are administrated in the user management while security services carry out the required entry and access checks. Because a vast number of different processes are activated via the portal access point, effective access control is of enormous importance for the security of company infrastructure. The problems of security are examined in more detail in the following section. The efficient, secure and (as extensively as possible) automated administration of the identities of customers, employees, project partners, etc. is summarized under the concept of identity and access management (see also: Fumy, W., Sauerbrey, J. 2004). Personalized areas are particularly important for the connection of mobile terminals so that the user is efficiently and quickly directed to documents of interest. The application server provides the process environment for reusable components (e.g. EJB or .NET components) and also offers services for transaction administration and load balancing. The substantial task of the integration server is to connect with one another the functional components of applications initially developed in isolation, such as ERP (Enterprise Resource Planning), SCM (Supply Chain Management) or CRM (Customer Relationship Management). This enables them to efficiently support workflows within the company. The communication or data exchange among the different applications is supported by various broker services (Message, Object and RPC Broker) as

Secure mobile business solutions for real-time enterprises

247

well as standardized web services. Access to the front-end of the architecture and to the portal server takes place homogeneously via standard protocols such as HTML and SOAP so that all associated suppliers, customers, business partners and (mobile) employees interact in a uniform way with the platform. Differences arise, however, depending on which access networks are used. Mobile access is normally via wireless networks such as the WLAN networks of the 802.11 family or via mobile radio networks such as GSM, GPRS and UMTS while wire-connected access is made via TCP/IP-based standard networks. The different technologies affect the security of the company network.

Company security IT security plays a key role in modern networked companies. Through the accelerating digitalization of business processes, the volume of assets to be protected and processes critical to the company increase rapidly. The opening of company networks leads to direct access of services by customers, direct cooperation (including the joint processing of sensitive documents) with business partners is supported or, in the case of cross-company business processes, partners and suppliers are granted direct access to the company network and its security-critical resources. The outcome is to exacerbate the number of threats to company security. General protection objectives, threats and risks

A secure system is required to correctly check the identity of active elements (people, devices, services), i.e. to authenticate these elements. Access by authenticated elements to protected values must be checked. Confidentiality – which requires that no information is relinquished to unauthorized persons – must be ensured. Moreover, no unauthorized manipulations are allowed, i.e. data integrity must be guaranteed. In conjunction with business processes, the verifiability or binding force of executed transactions (e.g. in ordering processes) is also frequently of interest. The demand for availability implies that the resources and information of the company must also be accessible to authorized users.

248

Claudia Eckert

Threats

Among the most frequent attacks on today’s systems are spoofing attacks where an identity is faked (e.g. falsified IP address, falsified E-Mail address, falsified URL). Often these spoofing attacks are combined with man-in-the-middle attacks. Here, the attacker intervenes in a communication connection between two partners and completely takes over the role of one of the partners. Attacks on confidentiality mostly arise because information can be ‘sniffed’ by unauthorized persons, where they can very easily discover a wide range of sensitive data such as passwords. Of course, confidentiality is also imperiled if unimpeded access is possible to sensitive data stored on the hard disk of a computer and in databases. Unauthorized changes to information or unauthorized spying are frequently the result of virus or worm attacks that copy and execute the infected software on a computer. Threats to availability arise from Denial of Service attacks (DoS). Such attacks attempt, for example, to overload a victim system with requests so that it can longer provide authorized users with its services. Risks

Companies are directly and indirectly exposed to loss by the attacks described above. Direct loss or costs arise through incidental replacement, remanufacturing and repair costs. Indirect losses arise, for example, from losses resulting from penalties due to non-compliance with deadlines or the disclosure of confidential documents (including the breach of non-disclosure agreements). Strategic costs can be incurred through loss of confidence (withdrawal of orders, migration to competitors, etc.) of customers and business partners if, for example, sensitive person-related data are pilfered or abused. Further company-specific costs can arise from production standstills, the loss or changing of production data and, of course, the breakdown or non-accessibility of the portal or web server. If intra-company information reaches the competition, this can lead to considerable commercial penalties for the company (e.g. loss of market position). Security concepts for eBusiness architectures

A large number of concepts, services and protocols have already been developed to defend against the threats outlined above. A comprehensive report is presented in (Eckert, C. 2004). The security technologies current-

Secure mobile business solutions for real-time enterprises

249

ly deployed as standard measures within companies include anti-virus software, firewalls for the filtering of data traffic in or out of the company network, Intrusion Detection Systems (IDS) that promptly detect and defend against unusual user behavior, VPN solutions for the security of communication routes between external employees and the company network, authentication mechanisms such as Access Token or Smartcards, encrypted and signed E-Mail (together with a Public Key Infrastructure [PKI] if necessary) as well as SSL/TLS-protected connections to the web or portal server among others.

Customers, employees, business partners, ...

Back-end

Front-end

- Authentication - RBAC access checks - VPN (IPSec) - SSL/TLS encryption - PKI support - ...

SSL/TLS

Radius/ 802.1Xnode

Firewall

Internet

SSL/TLS

Firewall

Web-, Portal Server

Application server - Authentication - RBAC - VPN - SSL/TLS - PKI - ...

SSL IDS, AV, ... VPN DMZ

Databases Legacy systems

Internal company network

Figure 61: Secured eBusiness architecture (greatly oversimplified)

Figure 61 illustrates the embedding of standard security technologies in the eBusiness architecture shown in Figure 60. The separation into frontend and back-end components makes it possible to place the front-end with the publicly accessible web and portal servers in an internally and externally isolated sub-network, a so-called demilitarized zone or DMZ (see Eckert, C. 2004) and to monitor this sub-network by means of upstream and downstream firewalls. Depending on the configuration, a portal server can however also be relocated in the protected internal area. External users must be authenticated by the front-end, which means user identifications, roles, passwords or special credentials depending on the selected authentication process (e.g. password-based, SecureID Token, Kerberos V5) are administrated in the corresponding authentication server.

250

Claudia Eckert

Remote users normally dial in via a RADIUS dial-in node (Remote Authentication Dial In User Service – in future this will either be DIAMETER or 802.1X) via a telephone network, or via the Internet or a WLAN. The dial-in node receives the authentication data of the user and forwards them to the authentication server for validation. The front-end area also accommodates Intrusion Detection Systems (IDS), anti-virus programs and, if necessary, the required components of a PKI. The web server usually provides Single Sign On (SSO) for using web applications so that a user only needs to be authenticated once to access web services. Secure access to accessible WWW pages is achieved using the SSL/TLS protocol. A secure end-to-end link between the external user and the portal or web server is established via a Virtual Private Network (VPN) that today is usually configured on the basis of IPSec. The VPN establishes an encrypted authenticated tunnel between the end points of the tunnel, through which any application data that are automatically provided with the basic protection of the tunnel can be transferred. On the participating terminals, a VPN requires a pre-installed VPN client and exchanged (pershared) confidential information between the terminal and the VPN server, which in a company environment with a relatively permanent number of staff does require a certain level of administrative costs, but is in essence unproblematic. Mobile guests or business partners who only need temporary secure access within the company require other solutions such as the set-up of especially secure VLANs in the company. The back-end mostly performs renewed authentication through application services using application-specific identities (e.g. special roles) and proof of authentication. Role Based Access Control (RBAC), encrypted data transport, also within the local network, and decentralized anti-virus controls and IDS measures are part of the standard repertoire of security techniques used in the back-end area.

Secure mobile business solutions

Additional risks from mobile business solutions

Although mobile business solutions such as those described in the second section can represent considerable opportunities for enterprises, it is also important to be aware of possible added security risks when using mobile technologies (see also: Eckert, C. 2003).

Secure mobile business solutions for real-time enterprises

251

Wireless communication (WLAN 802.11a/b/g)

It must always be kept in mind when using radio networks that the data are transferred over the airwaves. This means that eavesdropping on these radio connections is considerably easier than in wire-connected networks. The WLAN card needed for this has become almost standard in laptops today and also integrates PDAs. Hence, cracking and sniffing and, moreover, altering of data packets in a radio network, become very easy for the attacker. Although the 802.11 WLAN standard contains a security protocol in the WEP (Wired Equivalent Privacy), it does, however, exhibit a number of security loopholes so that it cannot guarantee either confidential communication, or the unique identification of a user or, for example, reliable data exchange. In July 2004 the IEEE adopted the 802.11i standard that redresses the security shortcomings of the WEP to the greatest possible extent and is thus suitable for a company environment. A detailed analysis of the WEP security shortcomings and the security functions of the new 802.11i standard are covered by Eckert, C. 2004. The additional risks of the WLAN connection of mobile devices to an eBusiness architecture such as outlined in Figure 60 can be summarized as follows: There is a danger of unauthorized access to an internal network, including bypassing of the firewall protection, sniffing (eavesdropping) and altering of confidential information over the radio network, as well as direct access to the enabled disk drives of other mobile devices in the WLAN subnetwork. As well, the identities of authorized users can be stolen and services accessed by unauthorized persons as well as the execution of targeted DoS attacks on mobile devices (including spoofed Access Points). Mobile communication

In mobile communication via GSM/GPRS/UMTS, the data are only encrypted while they are being transferred over the air interface, i.e. up to the mobile antenna. They are then openly disclosed in both the operator network and in the fixed network. Other considerable security loopholes exist, e.g. movement and abode profiles can be created, SMS messages are not encrypted and man-in-the-middle attacks are relatively easy to carry out in GSM/GPRS networks (see also Eckert, C. 2004). While these problems may be tolerable for many private users, they can be critical for a company that exchanges sensitive internal company data or customer-related data over these networks. Even with UMTS only limited improvement can be expected with regard to security. Although a few prob-

252

Claudia Eckert

lems of GSM/GPRS have been eliminated (among other things the network is authenticated and data integrity is checked), end-to-end security between communication partners is still not provided. Typical additional risks of mobile communication are, as already mentioned for WLAN, the theft of identities, the sniffing and alteration of confidential data and the illegal usage of billable services. Mobile terminals

Characteristically, mobile terminals are small, lightweight and, of course, portable. Due to the fact that owners always carry them on their person, mobile terminals are mostly considered to be secure devices – a specious conclusion. Their small physical size means that they are very quickly forgotten, left lying unattended and easily stolen. Studies suggest that PDAs alone are lost at a rate of approximately 30 percent. In most cases the stored information is of very much greater value than the replacement of the device. Considering the fact that mobile devices contain sensitive business data, for example, about customers or company-internal matters, this can result in a considerable loss for the affected companies. The risk is very high because all market-leading operating systems exhibit a very similar deficit in relation to the integrated security services. All these systems do provide access control so that an authenticated user can only obtain access to the stored data by using a password. However, they frequently employ feeble techniques or the secret passwords are relatively openly stored on the device itself. There are already numerous programs on the Internet to crack the integrated password protection of these systems. Because access control is normally only carried out when the user logs in, any person who gets their hands on such an unguarded device can gain direct access to the data stored in it. Because of the portability of the devices, they are used in a host of different environments so that there is considerably greater danger of the user being observed when entering his PIN or password. In addition, the mobile device is exposed to the threats familiar from the PC world such as the transfer of viruses or the spying of the hard disk through the infiltration of worms over the existing communication interfaces. If the mobile device at the company workplace is synchronized with the company PC, the harmful software on the mobile device can directly propagate onto the company computer by bypassing the central firewall controls and anti-virus filters (see Figure 61). At the latest during the integration of

Secure mobile business solutions for real-time enterprises

253

mobile solutions, it will again become clear that the perimeter security common in companies today is no longer sufficient. Secure mobile business solutions

The additional risks associated with mobile technologies make clear that extra protection measures must be seized when integrating mobile solutions. Figure 62 outlines the secure connection of mobile terminals in the eBusiness architecture shown in Figures 61 and 60. For mobile access to a company network via a public WLAN access network (e.g. at a Hot-Spot), a VPN connection between mobile device and company server is essential to obtain controlled and secure access to protected company resources (e.g. E-Mails). Secure access to the web server can also be achieved via SSL/TLS. WPA or 802.11i can be used for safeguarding the air interface in the local company WLAN, although the establishment of a secure end-to-end connection is preferable. Front end

VPN

WLAN WPA, (WEP) WTLS

WLAN-AP, 802.1X node

SSL/TLS

Firewall

SSL/TLS

GSM/GPRS/UMTS network Radius/ WAP node

WTLS WAP gateway SSL/TLS

Mobile or wireless connection VPN (IPSec) connection SSL/TLS-protected IP connection

Figure 62: Secure connection of mobile terminals to an eBusiness architecture

A VPN connection should also be established to form a secure end-toend connection when using mobile networks. For GSM/GPRS/UMTSbased access of WAP-capable telephones or mobile devices with a WAP browser, the submitted data are conveyed to the front-end via a WAP gateway of the network operator. To realize this, a secure connection is established between the mobile terminal and the WAP gateway by means of

254

Claudia Eckert

WTLS. However, this connection is interrupted in the gateway because the data must be transformed into SSL/TLS packets before they are forwarded to the web or portal server. The same is true of data transfer from the portal server via an SSL connection to the gateway and, from there, by means of WTLS to the WAP device. In order to prevent security loopholes for the company from the disclosure of all data in the gateway, such a WAP gateway should be operated in the trustworthy company environment. If the mobile devices are deployed outside the protected and controlled network, it is often difficult to check whether the applicable company-wide security guidelines (Security Policy) are continuously adhered to by mobile employees (e.g. wireless connections may only be established to the company network, downloads from the Internet are prohibited). However, this is important to ensure that a mobile device does not allow infiltration of harmful software into the company network, e.g. during synchronization. The electronic Security Inspector eSI of the Fraunhofer SIT (see: Sarbinowski, H. et al. 2003; Singer, E. 2004) is a tool that permits corresponding controls to be dynamically carried out and that checks compliance with specified guidelines in both the components of the front-end and back-end and in the mobile terminals. Company data such as quotations, construction documents or strategic planning data that are saved on mobile devices may well be vital for company survival and must not fall into the wrong hands. This requires file encryption programs that are transparent to the user and that permit synchronization of the encrypted files with different platforms (Windows and Unix world) and necessitate a rigorous and, if possible, Smartcard-based authentication of the user to provide a high degree of security. Encrypting file systems today are, however, still not a standard component of PDA operating systems. Commercial products that cover some of the aspects mentioned already exist for the different systems. A transparent, platformindependent encrypting file system that can be used for both Windows and Unix environments and also supports joint use of encrypted files between workteams was developed on the basis of (Eckert, C. et al. 2000) with the Jsec tool at the Fraunhofer SIT. Setting up a secure communication channel between the mobile device and its respective environment by means of SSL/TLS, IPSec (VPN connection) as well as the use of S/MIME for mail encryption and the use of secure XML standards for signed and encrypted exchange of XML documents also require PKI support on the side of the mobile terminal. As a result of resource limitations and minimization of administrative configu-

Secure mobile business solutions for real-time enterprises

255

ration costs, it is frequently not possible to provide complete PKI support on mobile clients. In order to nevertheless securely integrate these devices in a company network, “lightweight” PKI solutions can be implemented such as those made available by the NSI PKI (see also: Hunter, B., Filipovic, B. 2002). The signing of business documents such as job orders and purchase orders, is an important component of many mobile business processes. Mobile devices are subject to considerable risks, meaning they cannot easily be considered as credible signing environments. One remedy can be the creation of a PDA enhancement such as was developed with the Trusted Pocket Signer (TPS) (see also: Hartmann, M., Eckstein, L. 2003) within the framework of the VERNET program of the BMWA. The TPS is a PDA equipped with a special viewer component that can only be executed in a secure mode in order to form a basis for the dependable creation of electronic signatures. These are created by the utilization of a signature key that is kept securely on a Smart card. For example, to sign a document that is processed on a PC, it is first transferred to the PDA via a secure wireless connection where it can be viewed and checked by means of the specially developed content viewer. The viewer component guarantees that “what you see is what you sign”. In order to cultivate user awareness in the creation of an electronic signature, the process of signing must be explicitly activated by the user by making a handwritten signature, i.e. a biometric feature. On one hand, this should prevent a user from inadvertently triggering a signing process by pressing a button or similar and, on the other hand, it should ensure that the user is aware of the consequences of making such a signature as in the case of the customary handwritten signature. The biometric module that was integrated in the TPS as a separate component checks the user’s signature. If the signature is valid, the hash value of the document to be signed is calculated and transferred to the Smartcard in the PDA software via the integrated Smartcard interface module. In addition, the Smartcard takes steps to create an electronic signature for the hash value. This also guarantees that the signature key of the user does not need to leave his signature card. The protection of mobile business solutions and their secure integration in a company architecture is not a simple undertaking; it demands a high degree of security expertise. However, the protection of mobile processes is essential so as not to jeopardize company security. Solving this task in a credible way still represents a huge challenge for real-time enterprises.

256

Claudia Eckert

Integrated security management The technical security solutions described in the preceding sections are without doubt important and essential components of a modern company infrastructure. It is, however, clear that technical security solutions that are mostly deployed in isolation are insufficient to achieve an adequate level of security for a company. The difficulties can be illustrated by the simple metaphor of a security chain because a chain is only as strong as its weakest link. The chain begins with the user who, of course, represents a weak spot in many respects because it is the user who frequently undermines the security services of the system through negligence and ignorance (e.g. careless divulgence of the password). Before a user can use services, he must be authenticated. Points of vulnerability arise if, for example, passwords are transferred in plain text over the network. Following successful authentication, access to sensitive data must be checked. The security chain is broken by the absence of corresponding controls or too many users acquiring access authorizations. The communication connections, whether wired or wireless, represent further links in the chain. If transferred data can be manipulated or if the attacker manages to take over existing communication connections, he will be able to become active in the name of the authorized user. In this case, the access checks to the end systems and the filtering of firewalls will no longer be effective. The security chain illustrates, however, that secure communication alone is not sufficient to protect data during the entire data processing. If a user acquires unauthorized access to these data as they are being processed or saved on one of the participating computers, the protection of the communication connection alone will have no effect. If a reasonable level of security for a security-critical company infrastructure is demanded, this requires a holistically characterized process; focusing on individual aspects as is currently still the order of the day proves in practice to be inadequate on every occasion. Modern companies have to face the challenge of looking upon IT security as an integral management task that must be systematically combined and integrated with technical, personnel and operational security measures. This requires a company-wide security concept with a security policy that lays down the technical and organizational rules, behavior guidelines, responsibilities and roles as well as measures to achieve the protection objectives of the company. The integrated security management has the task of securing business continuity to protect the standing and reputation of the

Secure mobile business solutions for real-time enterprises

257

company, to win the trust of business partners and customers, to minimize costs of security incidents and to make sure that laws are adhered to and that conditions are fulfilled (also KontraG, Basel II, Teledienstedatenschutzgesetz (TDDSG)). Integrated management in a modern company requires the concerted contribution of top management, those responsible for the process and those responsible for IT.

Summary Mobile applications and business solutions are indispensable constituent parts of real-time enterprises. They promise a sizeable growth in productivity, open up new business areas and help to reduce costs. Mobile solutions of the future are not limited to enabling broadband access to E-Mails and data for mobile employees. On the contrary, they must be integrated in existing business processes and applications such as ERP and CRM. Secure and reliable embedding of mobile solutions in a company architecture is an essential precondition for the acceptance and the ROI (Return on Investment) of these solutions. This article has demonstrated the uses of mobile solutions using simple examples and outlined their integration in the modern eBusiness architecture of a company. Company security is of strategic importance for realtime enterprises. The embedding of classical security technologies for the protection of company values is illustrated by the use of a generic business architecture. Mobile solutions represent additional risks for the company. This article has shown substantial problem areas and outlined standard and innovative solutions. The protection of mobile business solutions and their secure integration in a company architecture is no easy task. The establishment of a wholly dependable security management that continuously ensures company security (security is a process) presents a huge challenge for real-time enterprises. However, it is a challenge that must be met by a modern company if it does not to wish to lose its market position.

258

Claudia Eckert

Bibliography Eckert, C., Bayarou, K., Rohr, S.: NGN, AII-IP, B3G: Enabler für das Future Net?! Überblick über Entwicklungen im Bereich zukünftiger Netze In: InformatikSpektrum, Volume 27, Issue 1, February 2004 Eckert, C.: IT – Sicherheit – Konzepte, Verfahren, Protokolle, R. Oldenbourg Verlag, 3rd, revised and expanded edition, October 2004 Eckert, C.: Mobil, aber sicher! In: Total vernetzt – Szenarien einer informatisierten Welt, F. Mattern (ed.), Springer Verlag 2003 Eckert, C., Erhard, F., Geiger, J.: GSFS – a New Group-Aware Cryptographic File Syste, In: Proceedings of the World Computer Congress, SEC2000, Beijing, August 2000 Fumy, W., Sauerbrey, J.: Identity & Access Management – Faster ROI and improved security through efficient assignment of rights and access control (the following chapter in this book). Hartmann, M., Eckstein, L.: TruPoSign – A trustworthy and mobile platform for electronic signatures; In: Paulus et al. (ed.): Information Security & Business; Vieweg Verlag; October 2003 Hunter, B., Filipovic, B.: Enabling PKI Services for Thin-Clients; DuD – Datenschutz und Datensicherheit, 9/2002, Vieweg Verlag, October 2002 Sarbinowski, H., Shafi, T., Eckert, C.: Der elektronische Sicherheitsinspektor eSI: Ein Tool zur dynamischen Analyse der IT-Sicherheit eines Systems In: Security, E-Learning, E-Services 17. DFN-Arbeitstagung über Kommunikationsnetze, LNI, GI-Edition, Dusseldorf 2003 Singer, E.: Tool-unterstützte Überprüfung der Einhaltung von Sicherheitsrichtlinien bei drahtlos (spontan) vernetzten mobilen Endgeräten, Diplomarbeit, TUDarmstadt, 2004 Wiehler, G.: Mobility, Security and Web-Services, Public Corporate Publishing, Siemens, 2004

Walter Fumy, Jörg Sauerbrey

Identity & Access Management Faster ROI and improved security through efficient assignment of rights and access control

Companies are coming under ever greater pressure to exploit their existing IT infrastructure to the fullest extent. As a result, new IT investments are being examined more and more critically to see whether they are actually instrumental in the optimization of business processes, thereby playing a role in cost reduction, increasing turnover and reducing risk. Investments in the area of Identity & Access Management usually lead to a quick payback and contribute moreover to a significant increase of IT security within the company and, because of that, to a reduction of risks of the business operation. Identity & Access Management is a relatively new term which comprises, however, IT concepts and methods used in companies for a long time. What is new about it is an increasing standardization of the processes, flows, data structures and interfaces, the structured approach using existing company resources and the improved support by special IT systems in this area. The term Identity & Access Management describes concepts, methods and IT systems for the (partly) automated acquisition and administration of identities (e.g. users of IT applications) and for automated control of the use of enterprise resources (e.g. IT applications). At the same time, the concepts and methods employed make use of information already available in the electronic directories and of existing user information from IT applications. Below, the benefits of Identity & Access Management are first described from a company perspective where the typical tasks of such a system as well as the respective application scenario are presented. The subject is then examined from the technical perspective. Firstly, the technological concepts for the realization of Identity & Access Management systems are explained. This is followed by a brief introduction of the most important standards in this field because standards are crucial for system interoperability, espe-

260

Walter Fumy, Jörg Sauerbrey

cially across company boundaries. A frequent driving force for the use of Identity & Access Management systems is the realization of Single SignOn, i.e. the administrative simplification of login processes. Finally, several technological variants of Single Sign-On systems are explained.

Benefits of Identity & Access Management The IT infrastructure normally used today within individual enterprises as well as across-enterprises can be characterized as follows: • • • •

has developed step-by-step over time; based on heterogeneous platforms and operating systems; increasingly uses web technologies; comprises a large number of different applications for the support of business processes; • requires the administration of a large number of different users for these applications; • includes a large number of different directories, databases, files which store different data related to users (and their privileges), company resources and IT resources. This leads to a series of significant problems and question marks regarding the operation of the infrastructure: • How can national and international regulations for data protection be met and how can the required documentation be guaranteed? Statutory regulations for financial service providers, public healthcare system, the pharmaceutical industry and other industries require a secure access control infrastructure. • Security violations are frequently committed by former employees who can still make use of their user authorizations. A real-time revocation of access rights is essential to maintain the standard of security. • In many cases even higher administrative IT costs arise from a transfer of an employee than from new recruitment. As a whole, an increase in the administrative costs of user administration can be observed (costs explosion). • A strong increase in the number of users who require instant access to applications or resources is apparent. Usually, however, the time period until a new sales employee, for example, gains access to the customer

Identity & Access Management

261

database (CRM) can take several days. • A user is registered on different systems under different user names (e.g. r_meier, RudiMeier, MeierR, ...). In everyday practice, this has enormous consequences for the administration as well as for the users of IT systems. IT administrators require a great deal of time for routine tasks with only little value-added such as the creation of new users and the assignment or withdrawal of access rights. These tasks must normally be carried out separately for all applications that are used. Also, from the user’s perspective, access to the various IT resources required throughout the working day proves expensive. The user usually has to remember his own user identification with corresponding password for each application. These passwords must be changed on a regular basis. However, for security reasons the passwords should not be written down. As a result, it is often the case that passwords are forgotten, particularly for applications that are seldom used. This then requires a call to the helpdesk that replaces the password with a temporary one which must immediately be changed to a new password. All in all, this is not very convenient for the user, resulting in lost productivity and significant costs at the helpdesk. An Identity & Access Management system helps both the user and the administrator by automatically processing essential basic tasks: • Identity maintenance – Identity import from authorative sources, e.g. personnel administration systems, Customer Relationship Management (CRM) systems, Enterprise Resource Planning (ERP) systems, etc. – Correction, standardization and consolidation of the imported identities into a single unique identity per user • Cross-application authorization and policy administration – Administration of permissions: roles, rights, group affiliation – Automatic allocation of the authorization based on rules and policies • Application-specific user and access administration – Provisioning of users and access rights into company resources – Password management – Validation of users and access rights in individual applications in accordance with security policies • Identification and authentication of users • Access management – Decision on access permission on the basis of user identity or role and

262

Walter Fumy, Jörg Sauerbrey

access rights (Policy Decision) – Carrying out the access decision (Policy Enforcement) The following figure shows a typical scenario of how the process from the hiring of the employee up to the use of the resources required for this employee can be significantly improved by an Identity & Access Management system.

Employee is hired

Identity administration • The employee is assigned the respective roles via the IAM platform • Rules are furnished on the basis of the company security guidelines (policy)

HR administration • The master data are generated in the personnel department and automatically transferred to the central Identity Store

Identity & Access management

In minutes

Productivity • Employee has the accesses and authorizations agreed in his role

Authorization • The authorizations corresponding to the role are automatically set • Individual criteria: e.g. period of validity, are entered in the IAM

Provisioning process • Intranet/Extranet access, E-Mail account, (…and others) are automatically created in the target systems • Individual authorizations are set in the portals

Figure 63: Typical Identity & Access Management (IAM) scenario

This scenario is only one example of many where an Identity & Access Management system in operational use reduces costs, increases productivity and improves security and, thus, contributes to a rapid Return on Investment (ROI). Qualitatively, ROI potentials result • At the development and operational level through: – automation of administrative processes – reduction of development and operating costs in internally-developed applications – reduction of helpdesk/hotline costs – enabling of User Self Service • At the user level by:

Identity & Access Management

263

– reduced search and access times – real-time allocation of company resources which avoids unproductive waiting times – reduction of the number of passwords • At the security level through: – automatic, real-time revocation of authorizations – prevention of superfluous authorizations – improved password security – efficient monitoring/auditing in relation to the authorization landscape • At the asset level through: – Cost transparency (PC, mobile phone, Remote LAN Access (RLA), etc.) – Prevention of unauthorized use and waste

Concepts and technologies As explained in the previous section, Identity & Access Management systems control who is given access in which context (e.g. dependent on criteria such as location and time, quality of authentication, ... ) and to which resources (e.g. data, services, ...). Technologically, an Identity & Access Management system includes the areas: • Authentication, i.e. the verification of the identity of people, applications or processes; • Identity management, i.e. the administration of user data; • Access management, i.e. the administration of data and processes for carrying out access controls. As illustrated in figure 64, these three areas are closely linked with one another, but each of them presents a technical challenge of its own. The basic functions, main approaches and interactions are described below.

264

Walter Fumy, Jörg Sauerbrey

User

Employee

Authentication

Identity Mgt

Access Mgt

Resources

Applications

Smart Card Provisioning

Policy Decision

Customer

Biometrics

Portals

Partner

Password

Web Services Identity Store

Citizen

Policy Enforcement

Token

Processes

Figure 64: Identity & Access Management Architecture Authentication

There are a number of different techniques for the secure authentication of users. The spectrum ranges from the widespread use of user names and passwords, authentication using biometric features and goes up to PKIbased solutions. In practice, more and more Security Tokens such as Smartcards are used for this purpose that, in addition to increased security and improved convenience, also allow for additional fields of application such as time recording, electronic purse functions, or physical access. If a user wants to access an application or service, he must first declare his identity and be authenticated. This way the application knows who wants access and can decide on the legitimacy of the desired action. If the information required for checking an identity is administrated by the applications or services, in a historically developed infrastructure with its distributed applications this results in multiple digital identities of a user. Transferring these multiple identities into unique representations is one of the central tasks of Identity Management. One problem in web-based applications is that the underlying communication protocol (HTTP) does not recognize transactions. In other words, every communication step and, hence, each interaction with the application is independent of the previous one. Accordingly, the user must be authenticated at each communication step. This evidently cannot be realized via an interactive process with the user.

Identity & Access Management

265

The usual solution consists much more of the introduction of a session management function where the first authentication is also used for subsequent communication steps. The basic idea is to issue the user with a temporary identification (”ticket”) that she re-submits at each request. Session management is frequently realized using cookies whose contents are determined by the application. The realization of a session management concept throws up a series of security-related questions, for example, on the duration of ticket validity, on the transferability to further applications or on security against manipulation. Attacks on web applications are not infrequently based on an insecure implementation of session management. In contrast, Identity & Access Management systems are specialized for this task and fully meet the respective corresponding security requirements. Following the initial user authentication, they package information on the identity of the user as well as his roles in encrypted cookies. A corresponding cookie must be transferred each time a protected resource is accessed. The Identity & Access Management system can decrypt, evaluate and then make a decision as to whether access is permitted based on the contents of the cookie. If required, specific user information can also be transferred to the web application for further use. This mechanism is a decisive advantage and a distinctive feature in the face of solutions that operate solely on the basis of standard Identity Management, but do not include session management. Identity management

Identity management ensures the efficient administration of all users in a single domain (e.g. a company) and is thus a core prerequisite for the realization of electronic business processes. The basis of each Identity Management solution is a high-performance directory service (directory, database, repository) which makes the necessary data efficiently and readily available. Frequently, in practice, so-called meta directories are used for this. These guarantee adequate performance, a suitable load distribution – even for large quantities of data – and efficient interfaces for the administration and retrieval of data. The creation and administration of user data is mostly carried out by dedicated users (administrators). However, there are also solutions where users register themselves and enter and maintain their own personal data. In these cases, it may also be necessary to release this data.

266

Walter Fumy, Jörg Sauerbrey

Access to data provided by Identity Management takes place via standardized protocols, mostly the Lightweight Directory Access Protocol (LDAP). There are two main possibilities for the consolidation of user data and access of applications to these data. 1. Each application has its own internal user administration and the user data of the application are always synchronized with the database of a central Identity Management system. To achieve this, the first step is to carry out the validation and consolidation of user data such as authorizations, roles and profiles; in the second step provisioning is performed on the unified user data, i.e. especially to make available this information to applications and services. Identity Management must provide the necessary functionality for both steps. 2. The applications do not have their own user administration; they access a central user directory. All requests are directed to the central Identity Management. This situation requires an initial consolidation of the user data, thus avoiding the problem of data synchronization. However, this approach results in increased requirements of availability and performance. Both concepts lead to consistent data management for all connected systems and, overall, to a significant reduction of administrative costs. Access management

The outsourcing of user administration from the applications has the consequence that authorizations no longer need to be administrated by the applications themselves. Instead, users can be assigned authorizations on the basis of current business requirements or the current security policy within the common framework. Access Management systems control access to objects or resources. Firstly, the data must be made available, on the basis of which access authorization can be verified later. The rules of individual applications regarding the allocation of rights to users are submitted to a Policy Decision Point (PDP) and centrally assessed there. Current authorization information is then provided in a format that can be interpreted in the applications by Policy Enforcement Points (PEP). Access rules describe which subjects may access which objects (resources) in which context. Subjects are mainly users represented by names and

Identity & Access Management

267

optional attributes such as location, E-Mail address, country identification, role or group affiliation. In the case of web applications, the objects or resources to be protected can exist, for example, in the form of a web server, a URI, a file, or a URL with function call. Access rules may not only be specified directly in the form of allowed or not allowed accesses, but can also be formulated in the form of rules on the basis of attributes. I.e., there are three types of data required for Access Management: identities, resources and access rules. The capture and management of these data are carried out in an administrative process that is temporally and functionally decoupled from the monitoring process. The enforcement of the defined rules is automated in the operational section. The administrative process itself is normally dealt with by Access Management systems, i.e. they include functionality for the administration of their data. Identity Management systems make this functionality, specially for the aspect of the administration of subjects, available in a comprehensive approach. The detailed administration of object-specific data and, in particular, the access rules lies mainly within the responsibility of the Access Management system. In many cases Identity Management is coupled with Access Management to guarantee the consistency of user data for all applications. After the administrative process, Access Management in the narrower sense carries out the automatic monitoring and control during operation. Based on the authentication of subject or user, a check is made of access rights and the granting or refusal of access. Standards crucial for interoperability

The aspect of interoperability is of singular importance for Identity & Access Management systems. Comprehensive interoperability can only be achieved by the appointment and adherence to internationally accepted standards. In particular, relevant bodies have been established for standardization in the field of Identity & Access management: • W3C: World Wide Web Consortium (http://www.w3c.org) • OASIS: Organization for the Advancement of Structured Information Standards (http://www.oasis-open.org) • Liberty Alliance (http://www.projectliberty.org/)

268

Walter Fumy, Jörg Sauerbrey

Some of the most important standardization activities in the area of Identity & Access management are briefly characterized as follows (see also Figure 65). Identity Services Interface Specifications (ID-SIS) Identity Federation Framework (ID-FF)

Identity Web Services Framework (ID-WSF)

XACML

SAML

WSS

XML-DSIG

XML-Enc

SPML

SOAP

XML

LDAP

HTTP

Figure 65: Identity & Access Management Standards XML

XML stands for eXtensible Markup Language and represents a data format for structured data exchange via the WWW. XML is not a descriptive language such as HTML, but a type of meta language which can be used to create other specialized languages. XML documents are intensively used to enable the electronic co-operation between companies, i.e. the exchange of business documents over the Internet. The standardization of XML is performed by W3C (see W3C Recommendation, 2004). SOAP

The Simple Object Access Protocol (SOAP) describes one of the XMLbased mechanisms for exchanging information via objects or entities. The three major components of a SOAP message are: • Envelope: the “envelope” for a message • Header: Contains different meta data for the message • Body: Contains the actual message The main objective in the development of this W3C standard (see W3C Recommendation, 2003) was a simple and extensible concept. SAML makes use of the second characteristic in particular.

Identity & Access Management

269

SAML

The Security Assertion Markup Language (SAML) which is based on XML is standardized by the OASIS industrial consortium (see OASIS Standard, August 2003). SAML also enables secure authentication for web services and provides an important basis for the Single Sign-On solution of the Liberty Alliance. The SAML framework also integrates other XML-based standards such as XML Signature (XML-DSIG), XML Encryption (XMLEnc) and SOAP. An Open Source implementation of SAML can be found at http://www.opensaml.org. SPML

The Service Provisioning Markup Language (SPML) based on XML is also standardized by OASIS (see OASIS Standard, October 2003). SPML is an extension of the Directory Services Markup Language (DSML) and enables the efficient and standardized data exchange of provisioning systems with one another as well as between provisioning systems and Access Management systems. XACML

The eXtensible Access Control Markup Language (XACML) is a language definition used to describe the rights for access to resources (see OASIS Standard, February 2003). XACML is standardized by OASIS (see also http://www.oasis-open.org/committees/xacml); an Open Source implementation can be found at http://sunxacml.sourceforge.net. Web Services Security (WSS)

The term Web Services Security (WSS) describes a number of different types of security mechanism that are defined independent of implementation and language with the aid of XML schemata. These comprise XML Signature (XML DSIG), a mechanism for ensuring the integrity of SOAP messages, XML Encryption (XML Enc) for guaranteeing the confidentiality of SOAP messages and other security-relevant mechanisms. Web Services Security is standardized by OASIS (see OASIS Standard, March 2004). The standard is based extensively on existing specifications

270

Walter Fumy, Jörg Sauerbrey

and adds a framework for embedding security mechanisms in a SOAP message to these. As a result, Web Services Security is primarily a specification for an XML-based container for security meta data. Identity Federation Framework (ID-FF)

The Liberty Alliance promotes a concept of federally administrated and inter-company valid identities. As well as other things, the specifications defined in the Identity Federation Framework (see Liberty Alliance:ID-FF, November 2003) enable the controlled linking of different user data distributed within a group of partner companies. Identity Web Services Framework (ID-WSF)

The Identity Web Services Framework specified by the Liberty Alliance provides the technical basis for identity-based web services (see Liberty Alliance: ID-WSF, November 2003). This framework can be adapted to the specific requirements of companies and, at the same time, guarantees the protection of personal data as well as the exchanged user information. Identity Service Interface Specifications (ID-SIS)

The Identity Service Interface Specifications ultimately apply to the Identity Federation Framework as well as the Identity Web Services Framework. In the framework of ID SIS, specifications emerge for the standardized transfer of personal registration information as well as the Contact Book, Geo-Location and Presence areas as well as others. Standard-based networked services such as the login profile, guest book or calendar and different localization or messaging services can be made available on the basis of ID SIS.

Single Sign-On solutions Single Sign-On solutions form a crucial driving force for the sector of Identity & Access Management. However, depending on the characteristics of the solution, the key points are completely different. Concepts where users must authenticate themselves only once during a session or to only one instance and where any necessary further authenti-

Identity & Access Management

271

cations are made automatically are described as Single Sign-On (SSO) concepts. Single Sign-On is associated with a high level of user-friendliness (the user only needs one set of login data to hand) with simultaneously improved security because varied or complex login data and, in particular, “strong” passwords are managed in such a way that they are transparent for the user. Single Sign-On concepts have been discussed for many years and a large number of different architectures proposed for their realization. The various approaches differ primarily in the character of the SSO instance (e.g. a central server, the local system of the user, a mobile token) as well as in the concept of automated authentication (e.g. passwords stored on Token, login data saved on a server, provisioning of login information, federation of user data). Two classes of Single Sign-On systems can be roughly distinguished (see Pashalidis, A., Mitchell, C. J. 2004): 1. The SSO instance administrates the service-specific authentication information for the user and thus authenticates him to a service if required. This approach is also known as “pseudo Single Sign-On”. 2. Special relationships are established between SSO instances and services on whose basis an SSO instance can inform a service about the authentication status of a user. This is usually carried out using assurances. This approach is identified as a “true Single Sign-On”. A further important differentiating feature of SSO architectures is the local or central positioning of the SSO instance. Some examples of Single Sign-On solutions are described and discussed in the following. The currently most important central Single Sign-On solutions in the web environment are .NET Passport from Microsoft and the concept of the Liberty Alliance project. Both solutions enable the user to authenticate himself with different web services using the same access data. The user can maintain his profile centrally and, if necessary, he can release elements of it depending on the application, e.g. his address for an order. Above all, local Single Sign-On solutions are attractive because of their low costs and short-term viability. Currently, this mainly concerns “pseudo Single Sign-On” solutions. In the future, technologies such as those provided by the Trusted Computing Platform Alliance (TCPA) will also enable the realization of local true Single Sign-On.

272

Walter Fumy, Jörg Sauerbrey

.NET Passport

.NET Passport (www.passport.com) is a web-based SSO service that has been offered by Microsoft since 1999. A passport server acts as an SSO instance in which everyone with an E-Mail address and password as well as additional personal data can be registered and can thus get an online identity. This identity is represented by a unique Passport User ID (64-bit). After successful registration a user must then identify himself just once at the beginning of his session to web services which participate in the passport service. He can use other passport partner sites and services without re-registering. Substantial parts of .NET Passport are based on Kerberos technology (see Internet Engineering Task Force, September 1993). The decision as to whether the user has already been authenticated is made on the basis of a so-called Ticket Granting Cookie (TGC). A Service Granting Cookie (SGC) transports the assurance of the authentication as well as additional data from the user profile (e.g. address information). These cookies are encrypted and stored on the client system. The TGC can only be read by the passport server whereas the SGC can also be read by the service concerned. Microsoft .NET Passport is one example of central true Single Sign-On although the number and variety of passport partner services have remained relatively modest. Liberty Alliance

The Liberty Alliance (www.projectliberty.org) is a consortium that is developing open standards for web-based Single Sign-On. In particular, the Liberty Alliance has developed guidelines for a cross-company association (“Federation”) of user accounts which always includes the express agreement of the user. Among other things, the consortium is working in the medium-term to establish “Circles of Trust” with service and identity providers to be able to administrate trust relationships in the background in such a way that it is transparent for the users. The specifications of the Liberty Alliance are an example of decentralized true Single Sign-On. In the terminology of the Liberty Alliance, the SSO instance is identified as an Identity Provider. Assurances about the type and status of the authentication are formulated with the aid of the Security Assertion Markup Language (SAML). Pseudonyms are used as SSO identities for the protection of privacy. A number of specifications of the Liberty Alliance is already available in

Identity & Access Management

273

the final versions (see section 3.5). The first implementations are available. Trusted Computing Platform Alliance

The Trusted Computing Platform Alliance (TCPA) is a working group which specifies methods to make the computer more secure through additional hardware. A crypto co-processor shall convert the PC into a “secure platform” and not just identify and authenticate its user, but also assume encryption and decryption tasks. On the basis of the TCPA specifications local true Single Sign-On can be realized. Besides, the so-called Identity Credentials of the TCPA can be used as SSO identities. The Identity Credentials are concerned with special certificates which confirm that the platform in question conforms to TCPA. A major disadvantage of this concept is the poor mobility support of the user because the Identity Credentials are always bound to a particular HW platform. Local Pseudo-SSO using Tokens

In local pseudo-SSO solutions the SSO instance lies within the control of the user. In general, the user administrates a database (stored in encrypted form) of his various authentication data. The user authenticates himself once to his SSO instance during operation. This then subsequently supports his authentication to applications or services by providing the necessary (decyphered) authentication data. In its simplest form this type of local pseudo SSO solution is only a SW tool for the administration of login information. Practical solutions store the database on a mobile Token (such as a Smartcard). An example of this is the HiPath SIcurity Card Login product from Siemens which provides features such as Auto-Record (where the SSO instance “learns” the registration data of a user), Auto-Fill (where the SSO instance recognizes the application or the login window and carries out the login completely automatically), password generator (where the SSO instance creates secure passwords for the user which satisfy the security policy of the company), or the secure storage of additional personal data (e.g. credit card numbers, account numbers, PINs). This type of entry-level solution also provides a useful migration path in the direction of a comprehensive Identity & Access Management architecture through the use of Smartcards and a Public Key Infrastructure (PKI).

274

Walter Fumy, Jörg Sauerbrey

Bibliography Internet Engineering Task Force: RFC 1510: The Kerberos Network Authentication Service, September 1993. Liberty Alliance: Identity Federation Framework (ID-FF) v1.2, November 2003. Liberty Alliance: Identity Service Interface Specifications (ID-SIS). Liberty Alliance: Identity Web Services Framework (ID-WSF) v1.0, November 2003. OASIS Standard: Extensible Access Control Markup Language (XACML) v1.0, February 2003. OASIS Standard: Security Assertion Markup Language (SAML) v1.1, August 2003. OASIS Standard: Service Provisioning Markup Language (SPML) v1.0, October 2003. OASIS Standard: Web Services Security v1.0 (WS-Security 2004), March 2004. Pashalidis, A.; Mitchell, C.J.: A Taxonomy of Single Sign-On Systems, Proceedings of ACISP 2003, Springer LNCS 2727 (2003), 249–264. W3C Recommendation: Extensible Markup Language (XML) v1.0, third Edition, February 2004. W3C Recommendation: Simple Object Access Protocol (SOAP) v1.2, June 2003.

Martin Schallbruch

Real-time business requires security, trust and availability

A must – data that is secure and always available Nowadays if the complete IT system in a bank was down for a single day, the losses would probably be so severe that serious economic consequences could result. Real-time business is already a reality in the financial services sector. Information is the product that is dealt with here. A head start of the smallest time margin in information, e.g. in stock exchange trading, is worth ready cash. The basic prerequisite for time-optimized information processes is data that are constantly and permanently available and that are also reliable. Long gone are the times when it was possible to be competitive in the financial service sector without functioning electronic data processing. Besides, many new business models in this field can only be realized using information technology. Only day trading in the securities sector or virtual market places in the B2B sector are mentioned as examples here. The speed of information processing is not the only challenge. Just as important is the reliability of the data on which decisions are based. In the case of online stock trading, it would be fatal if data for a purchase or sales decision were not dependable. Particularly in professional trading, accuracy of several digits after the decimal point is required because a tiny difference can lead to substantial absolute values when stock trading in large numbers. The financial industry is certainly very appropriate for demonstrating the dependence of seamless information processing. It is, however, not nearly the only industry for which information technology that functions round the clock is indispensable. Under closer examination, ever more classical productive companies are also turning into real-time enterprises. In the manufacturing industries, companies are compelled to become

276

Martin Schallbruch

real-time capable for reasons of costs. Storage costs can be saved if raw materials are delivered “just in time”, i.e. precisely when they are required, and if final products can also be delivered immediately to the customer after assembly. There are not really any practical alternatives because storage capacity for raw materials and final products is not only expensive; it also ties up productive capital. Companies that wish to be competitive can no longer afford this luxury and must convert to comprehensive real-time business simply in order to survive. For some time public administration has also started to re-design business processes using modern information technology. Real-time capability follows as a positive side effect in this case. Thus, the government will make all federal internet-capable services available before 2005. A project group that centrally coordinates the required effort has been set up especially for this purpose under the description of “BundOnline 2005”. As a result, the possible use of over 400 services via the Internet would save not only costs, but also increase convenience and speed. Particularly the high level of convenience (administrative paperwork is not required) and the gain in speed will lead to increased utilization and higher efficiency of the services offered. Investments will pay for themselves that much quicker. A fundamental requirement for use – both here and elsewhere – is the trust of the users in the security of their data. An EMNID poll from 2001 within the scope of the “Government Online Study 2001” indicates that 85 percent of German citizens consider the transmission of personal data in eGovernment to be insecure. These security misgivings mean that online services are not used to their greatest extent so possible savings are not optimally realized. Hence, the level of acceptance rises and falls with the security of the data as do the realizable cost advantages which are dependent on this acceptance. In other words, convenience and speed alone does not attract new users if the security of the data cannot be convincingly affirmed beforehand and if it cannot be guaranteed. Security must be understood in its broadest sense in this context. Especially in the case of services that are already provided online via the Internet, concerns regarding the security of data transmissions from the user to the service provider are paramount. This can be explained by past security incidents where, for example, credit card data were “captured” by hackers during transmission over the Internet. In February 2003, it was reported in the press that the data of approximately eight million credit card holders in the USA were “stolen” by hackers via the Internet. This type of announcement leads to further sec-

Real-time business requires security, trust and availability

277

tions of the population regarding data transmission via the Internet as insecure, resulting in the cautious use of eGovernment services. Yet secure data transmission via the Internet is only a small part of the transaction and data processing chain, whose individual links must themselves be secure – because the whole chain is only as strong as its weakest link. In addition to safeguarding transmission via the Internet, the secure storage of data must also be guaranteed. This doesn’t just mean security against the unauthorized reading of customer data records, but also physical security against theft, fire, water damage and any other eventualities. Furthermore, secure data processing also means error-free data processing which must be ensured through appropriate measures. Another requirement is nonstop operation. Even if the applications being dealt with are not time critical, a minimum level of operation must be guaranteed at all times. Interruptions in online operation that are all too apparent lead to loss of customer confidence. Such psychological effects play a somewhat subordinate role in purely business terms, but must nevertheless be taken into account. The subject of “Business Continuity and Recovery Services” (BCRS) is of just as great importance as data integrity or protection against unauthorized access. A detailed treatment of BCRS would go too far at this point. It is important to note, however, that all the aspects mentioned must be considered to the same degree in order that real-time business not only functions, but is also accepted. All security aspects have both a technical and an organizational component, where each must be considered in a measured way. In order to prevent a hodgepodge of individual measures being created here, all security aspects and measures must be included in a comprehensive IT security concept.

The security concept – the basis of credible data processing Those who wish to make electronic process data processing secure often have to deal in practice with existing mature systems or at least partial systems, where security aspects were not always taken into account at the outset. Each individual component of the entire system must be analyzed and each must itself be secure. Even at the end when all individual components are secure, the complete system must checked once more to expose any errors which are, for example, based in the organizational process flow and the interaction of the components themselves. Support for these in a prac-

278

Martin Schallbruch

tice-oriented procedure is provided by the Baseline Protection Manual of the Federal Office for Information Security (BSI). IT baseline protection

The Baseline Protection Manual has a modular structure and provides the user with the possibility to use only those modules that he requires for his system. Taking the Baseline Protection Manual as the basis, a lean security concept is achieved that contains only the modules which are relevant for the respective system. However, some preparation must be made using the Baseline Protection Manual in order to create a sound security concept. Consequently, before the creation of such a concept, an IT structure analysis must be carried out to identify the components of the overall system. The overall IT system comprises all infrastructural, organizational, personnel-related and technical components which are connected organizationally or technically (i.e. networked). The results of the IT structure analysis must be documented so that the dependencies of the individual components, their interaction and the interfaces can be clearly identified. The next step is to determine the protection requirement for the IT system. The central question is, therefore, how great is the maximum level of damage if the IT system is impaired in its proper function. When these preparations have been carried out, the appropriate modules must be selected from the Baseline Protection Manual and a check made as to whether the recommended measures have already been implemented and, where necessary, supplemented. The Baseline Protection Manual specifies measures in all relevant areas such as, for example, client-server networks, fixtures, organization, communication and application components. This means all possible aspects for the respective IT component are taken into account. Cross-references from one measure to another must be followed where necessary. This is indicated in the Baseline Protection Manual by corresponding cross-references which assist the user in including all necessary protection aspects. The actions recommended in the Baseline Protection Manual are standard measures according to the current state of the art (i.e. technology) which usually can also be realized economically and stably. Detailed descriptions and instructions are included for concrete implementation of the respective measures. They enable the responsible persons to easily understand and implement the nominated measures. Depending on the addressee to whom the measures are oriented (system administrator, end user), the technical terminology also varies. As a universal instrument of IT

Real-time business requires security, trust and availability

279

security concepts, the Baseline Protection Manual supports implementation through concrete measures right down to the end user. The basic protection tool (GSTOOL) is available to keep the IT security concept up to date and to constantly provide a current overview of the realized protection measures, the requirement for protection and the costs. It allows all tasks and data concerning the IT security concept to be effectively and economically administrated. The GSTOOL supports the following areas in particular: • • • • • • • • • •

Modeling and layer model of IT baseline protection Capture of information about IT systems/IT structure analysis Application compilation Implementation of safeguards Evaluation of costs Assessment of protection requirements Report generation Audit support Basic security check IT Baseline Protection Certificate

Consequently, the development of an IT security concept is supplemented and simplified by an effective tool. To round this off, it is recommended to acquire an IT Baseline Protection Certificate. This also signals to the outside world that a certain level of protection can be guaranteed. The certificates range from self-explanatory up to certification by the BSI on the basis of an audit report which is carried out by a licensed IT Baseline Protection Auditor. A clear overview and basic introduction into the subject is provided by the “IT Security Guidelines” (IT Baseline Protection in brief) issued by the BSI. This describes the importance of protection measures in the IT sector in a clear way and recommends basic security measures.

Ensuring competitive advantage by checking security in real time Security is an essential prerequisite for successful real-time business. In order to be real-time capable, the required security measures must also be designed in such a way that the necessary dynamic security checks are possible in real time. If the business process is founded exclusively on the pro-

280

Martin Schallbruch

cessing of information (e.g. stock trading, commodity futures, virtual market places, online auctions), real-time capable checking of security is essential. Security checking must guarantee the technical security of data (protection against viruses and other damaging programs) as well as ensuring data reliability. This requirement must already be incorporated in the approach for a real-time protection concept. The following points in particular must be observed: Redundancy

Critical IT components must be designed to include redundancy so that, if a component fails, a replacement is instantly ready for use and can also be put immediately into operation. It should never be possible for the whole system to be down for repair or maintenance work. Components that are hot-swap capable must be used as far as possible which means they can also be changed during operation without having to switch off the system. Standby systems

In the case of essential systems, where protection against failure has the highest priority, the redundant design of individual components is not sufficient. Redundancy is ineffective if major catastrophes such as fire, flooding or earthquakes occur. This requires the operation of a complete second system which can immediately replace the first system in case of emergency. Data synchronization is of particular importance for the immediate switchover to a backup system. In the case of a catastrophic event, it is no longer possible to assume there will be sufficient time to transfer the current data to the reserve system. The constant matching of data is required in this instance, which necessitates a functioning and possibly also redundant data connection between the two systems. The systems should, therefore, be at a minimum spatial distance from one another to ensure that the reserve system is not also affected by the catastrophic event. Mobile standby systems

If a certain downtime is tolerable, mobile standby systems can provide a practical alternative. At the same time, however, it is essential that the mobile standby systems contain all the necessary components required for handling the business processes. Mobile standby systems have the advan-

Real-time business requires security, trust and availability

281

tage that they can also be rented as the case requires. However, if a great number of IT systems are affected in the case of a catastrophe, there may be insufficient mobile standby systems available for all those affected. In any case, mobile standby systems require current datasets to always be available as a backup, allowing them to be rapidly read into the substitute system. Compatibility with the substitute system should be checked regularly. Backup strategy

A major component of any concept for real-time business is a suitable backup strategy. This should be oriented to actual requirements and adapt itself to changes of the IT system. Regular testing is therefore essential to check whether the backup corresponds to the requirements of the process, particularly in relation to the topicality of data as well as whether the data can be read in without any errors and at the necessary speed. Security patches

The increasing exploitation of weaknesses in the software by so-called ‘exploits’ makes it necessary to frequently download the latest security patches. They prevent security loopholes being exploited, thus compromising the IT system. Virus scanner

The use of an up-to-date virus scanner is absolutely essential. The number and, above all, the rapid pervasive spread of computer viruses have constantly increased over recent years. This can also be attributed to the increasing online propagation of computer viruses. This occurs primarily through the sending of files by E-Mail, but also through no fault of the user by automatic propagation such as, for example, in the case of the so-called SQL Slammer and W32.Blaster worms. Enormous importance must be given to the constant rapid download of up-to-date virus signatures. Data checking must already take place at the input mail server where infected EMails must be intercepted. To guard against other infection routes in the network for example, all client systems should always have the most recent virus scanner available. Only in this way is it possible to protect against viruses which infiltrate through the use of end-to-end encryption.

282

Martin Schallbruch

SPAM filters

The mass flooding of the Internet with SPAM mail, frequently in combination with viruses and worms since the beginning of 2004, presents a serious threat for the disruption of business processes. Intelligent SPAM filters are becoming increasingly imperative. Firewalls

Just as essential for defense against the perils of the Internet is the use of a firewall. Viruses such as W32.Blaster have shown that a virus scanner alone is not enough. The proper configuration is decisive for the effectiveness of a firewall. Important here is that only required data are allowed in and out, and all other external data are kept away. Documentation

One point that is often neglected is the documentation of the IT system. This does not just mean the documentation of the status at installation, but also real-time maintenance of this documentation. All changes to the IT system must be reflected in the documentation – not only the hardware and software in use, but also all updates and patches, current settings and configurations. Only thus is it possible to create a speedy overview and to solve problems in real time should faults occur. Training courses

A central element of a functioning and secure information technology is the level of training of the user and the person who looks after the system, i.e. the system administrator. Operating errors by the employee should be excluded by the relevant training courses. Particularly high requirements are placed on the training of system administrators because they must immediately initiate suitable countermeasures in the case of a fault. This is only possible with a complete understanding of the technology. This particularly includes the knowledge of conceivable faults and breakdown constellations and the appropriate reaction.

Real-time business requires security, trust and availability

283

Confidentiality and integrity

In addition to the guarantee of security and availability, the confidentiality and integrity of data must be ensured. Encryption and signature are the media used to ensure this goal is achieved. Powerful encryption can guarantee that unauthorized persons do not gain access to data. Care must be taken here when dealing with the confidential data of customer invoices, also however protecting the company’s own business interests. The IVBB as a good example

A good example of a fast and secure IT system is the Berlin-Bonn Information Network of the federal government (IVBB). The reason for the formation of the IVBB was the relocation of the federal government and parliament to Berlin. The safeguarding of the division of labor of government functions between Berlin and Bonn had to be supported by secure information and communication technology. The Information Network is of vital importance for federal authorities with offices at a number of locations. Users of the IVBB are the Lower House of the German Government, the Federal Assembly, the Federal Chancellery and Federal Ministries, the Federal Audit Office as well as junior Federal authorities in Berlin and Bonn and at further locations. The concept for the medium-term and long-term realization of the broadband IVBB was already completed in 1995. In parallel to this, the Federal Office for Information Security (BSI) developed the IT security concept for the IVBB. In March 1996, the Federal Cabinet decided to further develop the expansion and operation of the IVBB from the existing Bonn Federal authorities network (BBN) together with the operator Deutsche Telekom. Today, more than 30,000 members in more than 180 premises use the IVBB. The high requirements for availability, confidentiality and integrity of communication are fulfilled by a host of competent security measures. Security was a mission statement even in the concept phase of the IVBB. Technical disruptions as well as rather improbable major catastrophes were deliberated. The necessity for comprehensive measures was once more confirmed by the aftermath of 11th September, 2001. Berlin and Bonn are connected by so-called trunk lines with a transmission rate of 2.5 Gbit/s. SDH rings (Synchronous Digital Hierarchy) each with 622 Mbit/s were built in Berlin and Bonn; a maximum of three locations are connected on each ring. Communication in the IVBB is therefore essentially encrypted. With its primary locations in Berlin and Bonn, the IVBB is constructed

284

Martin Schallbruch

as an exclusive, separate network with its own junctions to the telephone network and to the Internet. This means that, in the event of crisis, the secure telephone supply and an independent information platform with IVBB Intranet are available for crisis management. The junctions to public networks (Internet and telephony) are protected by special security components. A central firewall, virus scanner and SPAM filter protect the junction between the networks of the user, the IVBB Intranet and the Internet. This is all the more important because the impact of the joint Intranet and Internet access is constantly increasing and the IP service platform has developed into one of the extremely important modules of the entire IVBB. Today, a multitude of IP-based services are transacted via the IP service platform. The complete network is continuously monitored for its functional serviceability and security against manipulation. The failure of major components is intercepted by the annular connection (duplex conductors) as well as the duplicate construction of major components. Even the failure or complete destruction of one of the central switching nodes would only lead to minor and short-term restrictions for the user. To understand the IT-supported perception of government functions in today’s fast moving times as a real-time business, the IVBB is a good example of how appropriate safeguarding of real-time business can be built on.

Local network Bonn

CSN

Local network Berlin

CSN

CSN

CSN

CSN

Authority-sided network connection SDH network basis (Trunk: 2.5 Gbit/s; Ring: 622 Mbit/s) Central Switching Node

Figure 66: Topology diagram of IVBB

VI. Outsourcing

Outsourcing tasks that are not part of a company’s core competence is a growing trend. This results in business processes that are leaner and more efficient. It also facilitates fast reactions to changes in the marketplace. Outsourcing does equate to a networked company and more complex overall coordination, but innovative communication solutions have addressed this issue. They enable the successful management of valueadded networks, including those having many ecosystem partners.

Clemens Jochum

Intelligent IT sourcing in the financial industry: background, preconditions and requirements of future IT organization design

Context: why are alternative sourcing concepts necessary? Although it seems that the worst of the overall economic slump is now behind us, the German banking industry still faces difficult economic conditions – falling margins, rising loan loss provisions in the credit business and uncertainty on the stock markets. A swift and vigorous recovery is not expected.1 However, the unsatisfactory return on equity compared internationally and the consistently low market capitalization of German financial service providers are not the result of external factors alone. In fact, the present situation highlights the need for structural innovations in the financial industry. Instead of relying on economic conditions to improve, service providers can adopt a resolute entrepreneurial approach to generate opportunities for a lasting upswing in earnings. In addition to the necessary consolidation of German banking structures – without a doubt, Germany can be considered “over-banked” in a European comparison2 – the key structural challenges also lie in overcoming inefficient value-added models. Compared to other sectors, the financial industry continues to be characterized by a high degree of vertical integration. Over 80 % of bank products are produced in-house. In comparison, other sectors such as the printing and automotive industries already began decon1 2

See Deutsche Bank Research, Economics, No. 39, August 2003. Regarding the structure of the German financial industry in a European comparison, see European Central Bank, Structural Analysis of the EU Banking Sector. Frankfurt 2003.

288

Clemens Jochum

structing their value chains back in the 1960s and ‘70s as international competition began to grow. The automobile industry has gone to the extent of having suppliers deliver entire modules instead of parts, and the modules are then assembled on location by supplier personnel. The banking sector, in contrast, is only now beginning to restructure its value chains as an instrument of a competitive corporate policy. In spite of the traditionally strong competition between individual banks, many institutions still considered themselves to be largely self-sufficient. After all, customer loyalty had always been relatively high. By the beginning of the 1990s, competition in the financial industry had intensified dramatically. This was due to the advancement of new technologies and e-commerce as well as changed customer requirements. Today, customers do not wish to be restricted as to where and when they do their banking. The growing number of specialized providers such as direct banks is a clear indication of this trend. By the mid 90s, simple and fast access to bank products due to improved technologies had led to increased transparency between employees and, ultimately, to greater competition. First used in the 50s and 60s as a means of rationalizing workflows in bank operations, information technology has become a major production factor in banking due to its fast-paced technological development. Without IT, the enormous growth of international capital markets would not have been possible. It is only with advanced IT systems that financial institutions have been able to handle today’s transaction volumes quickly and reliably while maintaining an overview of the complex financial operations. Many of the financial innovations we have become familiar with over the last few years would be unthinkable without IT.3 This trend applies not only to banking within the international capital markets, but above all to retail banking. IT runs through the entire customer process – from acquisition and handling of business transactions to the continuous availability of customer information (e. g. on the Internet and self-service terminals). Bank customers increasingly take advantage of the features that this technology offers. For example: Between 2000 and 2002 alone, the number of online bank transfers more than doubled to over 760 million.4 Internet and online banking have become the basic strategy for customers and product groups and this has represented an important competitive factor for 3

4

See Jürgen Rebouillon/Stanley Bauer, Optimierung der Wertschöpfungskette durch Outsourcing, in: Luca P. Marighetti, et al. (publ.), Management der Wertschöpfungsketten in Banken. Wiesbaden 2001. See Deutsche Bundesbank, Statistiken zum Zahlungsverkehr in Deutschland 1998-2002. August 2003.

Intelligent IT sourcing in the financial industry

289

some considerable time. Bank products without digitization are inconceivable today. Invasion Front office rationalization

Expansion Branch automation

Integration Customer self-service

Virtual organization

Need for IT

Technology waves

1970–1980

1980–1990

1990–2000

2000–...

Figure 67: IT trends: Transition from process support to a driving force for success

However, as a consequence of the changed status of IT in banking (from process support to value driver), the IT production factor has has turned into an enormous cost pool in the material costs of banks. Compared to other material costs which have doubled since 1990, IT spending has risen to three and a half times the level of 1990.5 Thus, when it comes to improving the cost structure in the financial industry, IT is of great significance. However, costs cannot be lowered by simply making sweeping cutbacks in IT spending. Pressure to substantially reduce costs may under no circumstances result in lower quality IT systems. On the contrary, we are in the middle of the IT revolution. Demands placed on IT will certainly continue to grow in the future, not only with regard to volume, but especially as concerns quality: scalability, reliability, accessibility, new services for customers, etc. Cost reduction and continuing IT innovation are the two challenges that bank IT organizations now face. Potential solutions for these apparently irreconcilable requirements lie in the development of alternative IT sourcing concepts. Intelligent IT sourcing – sourcing is the procurement of internal or external services – can enable companies in the financial industry to rise above 5

See Christoph Dolzanski, Strategien zur Kostenreduktion bei Banken und Sparkassen, in: BIT, Issue. 4, Volume 3, September 2003.

290

Clemens Jochum

the effects of cost savings and increased productivity and to concentrate on their core competencies in banking while at the same time participating in an innovative, future-oriented and competition-directed IT development. By now, many companies are actively considering the use of alternative procurement concepts for information technology. Outsourcing as a management strategy is becoming increasingly important in the banking sector.6 However, full utilization of existing potential structures requires a considered approach. The nature of such an approach and what consequences the development of intelligent IT sourcing concepts will have on the future structure of the IT organization will be discussed in the following two sections. Purchasing

Co-Sourcing

Sourcing Mgt

Purchasing Service

Service

Project Teams

Service Provider

Service Recipient

Outsourcing

Service Provider

Service Recipient

Joint Venture

Sourcing Mgt Service Service Staff

Service Provider

Staff

Service Recipient

Service Provider

Figure 68: Sourcing strategy: selection of various models

Preconditions, preparation and first steps In the banking IT sector, outsourcing of IT services and reduction of costs cannot just be performed arbitrarily. There are strict constraints which must be adhered to: bank sustainability must be ensured, operative risks must not increase and regulatory specifications must be complied with. When developing a sourcing strategy, the first decision must be which IT services should categorically not be outsourced. Strategic IT core competencies 6

See Accenture/International Bankers Forum e.V., Mehr als Geld und Zinsen: Outsourcing im deutschen Bankensektor. September 2003.

Intelligent IT sourcing in the financial industry

291

such as IT security, risk management, IT architecture and sourcing management should certainly remain in-house. All other areas of IT can in principle be examined for their potential of being optimized by sourcing. Deutsche Bank applies the sophisticated approach of smart sourcing. Smart sourcing cannot be reduced to the simple alternative of “outsourcing or in-house”. Rather, it is a matter of finding the best sourcing model for each individual element of the IT value chain. To do so, the various processes in the IT value chain must first be identified, differentiated from one another and disassembled into clearly defined, manageable elements. A “make-or-buy decision” is then made after an exact analysis based on these disassembled internal processes. A wide range of sourcing models is available for selection (on-shore, near-shore, off-shore and in-house, outsourced or co-sourced), all of which must be seen as essentially equivalent. For each individual situation, it is important to weigh up whether the service is faster, more innovative or more cost-effective when it is internal, external, or sourced through a co-operation, and whether specific risks or dependencies arise from that particular type of sourcing. The first big step in strategic IT sourcing was made by the Deutsche Bank in December 2002 with the decision to outsource its continental European IT infrastructure. The 10-year outsourcing contract encompasses data processing centers and smaller server locations in Germany, Belgium Italy, Luxembourg, Portugal, Spain and Switzerland. From a cost perspective, the bank estimates a drop in the fixed level of operating costs by up to 80% in the next 10 years. Overall, it is expected that there will be significant cost savings. In spite of the considerable cost savings resulting from this decision, the preliminary considerations that led to this step were not limited solely to cost effects. Far more important than absolute cost savings is the connection between efficient operating cost variability and increased flexibility. By using a variable billing scheme geared to actual transactions, the bank is able to purchase its IT services flexibly according to momentary capacity requirements. IT services can thus be purchased like electricity or telephone services. This gives the bank the decisive advantage that it is no longer forced to maintain maximum resources needed to handle transaction peaks (e.g. during stock issuance and dividend payment) while these resources remain underutilized at other times. Unlike in-house operations, the external IT provider is able to achieve economies of scale because the capacities that remain unused by Deutsche Bank can be made available to other customers during that time. Another crucial aspect is the use of technological core competencies of

292

Clemens Jochum

the sourcing partner. Through its partnership with an external provider, Deutsche Bank participates directly in the expert knowledge of a leading global service provider specialized in the operation of IT infrastructure. In light of future demands on IT services in banking (volume, improved service quality, accelerated standardization), this opens the door to new technological opportunities that could not be implemented in this form if sourcing remained within the bank. Finally, an increase in innovative potential can lead to a competitive advantage. In cooperation with a partner who is among the innovative leaders on the IT market, access to new technologies is faster and more direct and suitable applications can be implemented promptly, all of which gives the bank a leading edge in the “time-to-market” competition. In view of these varied aspects, all of which determine the quality and benefit of the outsourcing relationship, it becomes clear how important it is to select a suitable provider. In the example cited, special care was therefore taken in designing the selection process. The objective was to find the provider on the market who was best able to meet our requirements in terms of the quality of services offered, payment conditions, security standards, potential for a future-oriented IT technology development and finally the prospects for employees affected by the outsourcing (transitional and employment conditions, advancement prospects). An important factor was also a certain degree of compatibility with the company’s own corporate culture. Based on these considerations, a decision was made for a twophase selection procedure.7 In the preliminary phase, five companies were selected on the basis of a market analysis and their offers were evaluated by applying a scoring model. Four evaluation criteria of equal weight were applied: finances, strategy, employees and technology. In addition, the providers were closely examined with regard to their market position, reference projects, financial strength, etc. After this pre-selection, it was decided to enter into a mutual “due diligence process” with two of the providers. At its core, this selection phase consisted of Deutsche Bank and the IT service provider assessing each other for the purpose of jointly building a sustainable framework for success and developing an operational concept.

7

For more detail, see: Peter Lassig / Hermann-Josef Lamberti / Clemens Jochum, Scoring- und beidseitige Due-Diligence-Prozesse im Rahmen der Lieferantenauswahl beim Infrastruktur-Outsourcing, in Wirtschaftsinformatik, 2/2003.

Intelligent IT sourcing in the financial industry

293

Smart sourcing and the future organization of the IT department – Impact The careful selection of sourcing partners is only one factor contributing to the success of a project. In fact, how intelligent sourcing concepts are applied has more far-reaching consequences for the future design of the IT organization. To be able to effectively and efficiently utilize the opportunities offered by existing sourcing potential, it is necessary to develop new expertise that is only partially offered by the present IT organization. The first step involves the creation of suitable coordination and control capacities to cope with the risks invariably connected with every outsourcing. The simple conclusion of a sourcing contract in no way automatically guarantees that the associated benefits will actually take effect (cost reduction, higher flexibility, better service quality, technological innovations). If large sections of a company are outsourced, a fundamental responsibility exits from the very start to systematically guard against the possibility of a loss of control and coordination. The challenge here is not so much a complete failure of the outsourcing relationship – there are only a few known cases of this in the field – but rather poor contract performance. Above all, it is essential to regularly adjust the subject matter of the contract to changing business requirements, since it is impossible to anticipate every modification that may become necessary over a contract period of ten years. In view of these considerations, a so-called “Sourcing Management Team” (SMT) that is able to meet these requirements should be set up at an early stage – ideally as early as the selection phase. As part of its infrastructure outsourcing program, Deutsche Bank set up a sourcing management team within its IT organization. As an interface between the organizations, this team offers a specific and in this form unprecedented set of skills. In particular, these skills include highly developed contract and negotiation skills, understanding of contract mechanisms and the insourcing business model, and the ability to implement the contract in the most economically advantageous way. Using these skills, it is possible to prevent a loss of control by closely monitoring the contract stipulations, in particular the service level agreements. For contract monitoring to be successful, there must be a clear concentration on basic principles and on defining, establishing and optimizing the core processes of documentation, measurement and change. By introducing homogeneous and standardized processes and guidelines, effective control can be ensured.

294

Clemens Jochum

The field of responsibility of the SMT extends far beyond simply controlling contract performance. The SMT is responsible for achieving the next stage in IT development and for ensuring that the technology areas are coordinated and integrated, whether they are internal or external to the company. In particular, technological know-how relocated to the external service provider with the transfer of company employees must continue to be coordinated and harmonized with all other developments inside the bank. For Deutsche Bank as an organization, the SMT acts as a link between the business operations and “production” on the part of the IT service provider. In the SMT, the organizations are brought together again and the complexity is reduced. Inevitably, this extends the classical boundaries of a company. The SMT is the starting point for continued development in strategic IT sourcing. It is the core competencies in sourcing management developed here that will make it possible for the next steps, namely the use of the overall existing sourcing potential. Associated with this is the implementation of synergies. It will no longer be necessary to build up a similar organization from scratch when it comes to monitoring future sourcing projects. In this way, it is possible to achieve economies of scale in strategic IT sourcing. The activities of the SMT and the new skills it encompasses are an expression of the fundamental change in IT department profiles that must take place in view of the application of intelligent IT sourcing concepts. In the future working model, the IT organization will focus on conceptual tasks and IT governance. The challenge will be to develop technological concepts tailored to the requirements of the business operation – with respect to the infrastructure and the applications – and to coordinate and generally monitor their development at the various sourcing partners. Thus, sourcing places new and greater demands on IT employees, whose active involvement in shaping this change will require them to review their own responsibilities. No longer is the employee’s role that of a craftsman; it has become that of a businessman or an architect. The future model offers employees a more varied and complex sphere of influence and new professional opportunities, but it also demands a great willingness to learn. At the same time, the employer is obliged to equip employees with new skills that are better suited to the changing business environment. Training and advancement programs tailored to the particular requirements and open and continuous communication are the best routes to take. The goal of implementing intelligent sourcing concepts throughout the entire IT sector can only be realized if the basic organizational conditions

Intelligent IT sourcing in the financial industry

295

in companies are suited to this endeavor. To implement a consistent and sustainable IT sourcing strategy, the IT organization must have a highly functional structure. Only in this way can existing potential structures be efficiently exploited and the diverse relationships to IT service providers effectively developed (coordination and control). This is one of the reasons why an integrated structure was developed for the Technology & Operations division at Deutsche Bank. As part of a functional redistribution, the technical departments of the corporate divisions of Corporate Investment Banking (CIB) and Private Clients & Asset Management (PCAM) were unified to form a single department. This generates synergetic effects in the respective technological fields of activity. Moreover, the responsibility of looking after in-house customers and coordinating different IT service providers now falls on a single department. We are convinced that this new organizational structure will bring us several steps closer to our concept where IT organizations will be going in future.

Jürgen Frischmuth, Christian Oecking

Outsourcing as a strategic management decision

Preface When the Internet hype faded and the economic downturn set in, many companies were forced to consolidate. In the days when business was booming and innovation flourishing, they had had little time to ask searching questions such as: What is our core competence? What are our cost drivers? What factors affect our industry and, therefore, our core processes too? How can we cut costs and minimize risks intelligently? And how can we focus on the essentials? Now, however, concerns about shareholder are bringing the subject of profitability back center-stage. Growth for growth's sake has become less important. Since industry today has to focus on different issues, information and communication technologies (IT) have a very important part to play. Business processes are increasingly being mapped onto or assisted by electronic systems – a clear indication of the close links between corporate business and IT strategies. The problem is: This trend is only adding to the burden on IT departments that are already straining under the heavy burden of dayto-day operations. Internal IT resources are often quite simply overstretched as a result. They have neither the capacity nor, in some cases, the skills that are needed. This is where the outsourcing of information and communication technology takes on added significance, as suitable models can supply both experienced management capacity and the ability to handle highly complex IT challenges. A similar procedure can be adopted to handle a company's entire process chains.

298

Jürgen Frischmuth, Christian Oecking

The aims of IT outsourcing projects1 Outsourcing information and communication activities is a strategic management decision. As a rule, companies decide to entrust a certain set of tasks to experts in order, first and foremost, to secure for themselves a strategic business advantage. But there can also be other reasons why a firm decides to outsource IT operations: • • • • • • •

To cut costs and add flexibility To gain access to expertise, skills and technology To improve quality and performance To trim the balance sheet To transfer risk To focus on core business To speed up innovation and improve execution.

Whatever reasons are behind a decision to outsource, it is important to be clear about the goals – and about exactly how these are to be measured – before the decision is taken. Clear goals and clearly defined measurement criteria are indeed the key factor of success in IT outsourcing projects. Around 40% of firms nevertheless still link the choice of this or that provider primarily to issues of price and technical capabilities (see figure 69). Criteria by which outsourcing partners or service providers are selected Price

44%

Technical skills

39%

Partnership/culture

39%

Service quality

33%

People/professionalism

28%

Innovation Agility and flexibility

22% 17%

Source: Forrester Research Inc.

Figure 69: Criteria by which outsourcing partners or service providers are selected

Outsourcing as a strategic management decision

299

Finding and selecting the right service provider is of paramount importance. After all, every company wants to commit to long-term collaboration that works smoothly in an atmosphere of mutual trust. If outsourcing is intended to serve strategic management objectives, it is not enough to base the decision merely on projected financial outcomes and a technical specification of the services to be provided. It is also crucial to spell out what a positive business relationship between the parties concerned should look like. A definition of the core processes that will be affected and a detailed account of the opportunities and risks associated with outsourcing are equally critical to the success of the venture. On a more fundamental level, it is important to pay sufficient attention to the quality of the relationship between the outsourcing customer and the service provider. All too often, people seem to forget that service providers too have to make money on an outsourcing project. Outsourcing is the business that pays their wages. In the course of such a relationship, the service provider and the customer should regularly sit down together and ensure that each one's aims still line up with those of the other. If outsourcing is not a win-win situation, something is wrong.

Outsourcing in transition

Origins of the term “outsourcing”

There are many different aspects to the concept of outsourcing. Perhaps the simplest and most useful interpretation is to see it as a composite of the words “outside”, “resource” and “using”. This “use of outside resources” can cover every product or service that might conceivably be purchased externally. Outsourcing is certainly a management method whose roots go back way beyond IT outsourcing as it is known today. Variations on the same theme are standard practice in classical manufacturing in particular. Management strategies such as make-or-buy decisions and the reduction of vertical integration effectively refer to the same topic. Today, outsourcing models encompass a wide range of corporate services. Examples include: • Catering services (canteens) • Payroll accounting

300

• • • •

Jürgen Frischmuth, Christian Oecking

Building security Fleet management Business travel management Cleaning services.

Looking at even this brief list, it becomes clear that the method can be applied to all kinds of different businesses and business activities. Each line of business nevertheless has its own specific angle on the same theme. Aspects of change

IT outsourcing is following the transition away from traditional business activities and toward more complex issues of corporate information technology development. Information technology today gives companies a strategic competitive advantage – in traditional markets, of course, but also and especially in new markets. Accordingly, such competitive advantages are no longer measured on the basis of conventional ratios such as IT costs as a proportion of sales, but are, ideally, reflected by the performance indicators used in balanced scorecard systems, for example. In recent years, the practice of outsourcing information technology and/or business processes has therefore become more widely accepted as a valid tool of management. At the same time, the focus is visibly shifting away from traditional factors of success to the business value that this tool provides to the company. One of the metrics used is Economic Value Added (EVA), which defines the actual value that a service adds to a company in the form of a management ratio. Companies that examine the issue of outsourcing frequently get into deep discussions about their own core competencies, qualities, strengths and weaknesses. Wrestling with this subject matter on an internal level often enables a company to redefine and, ultimately, reinvent itself. For companies intent on shaping their markets in the long term, understanding this process as a permanent, ongoing challenge will prove to be a crucial factor of success. What does this mean to the companies that provide outsourcing services? They have no choice but to consistently focus on increasing the business value they add for their customers and for their customers' customers. As a result, outsourcing is no longer regarded merely as taking care of one aspect of a customer's operations. Instead, it also involves actively shaping the customer's business activities in the marketplace. This practice of shoul-

Outsourcing as a strategic management decision

301

dering responsibility for one or more business processes is, therefore, logically referred to as business process outsourcing (BPO).

Adding value by outsourcing the management of complexity Full-service IT providers are accepted as such only if they apply the same professionalism to every link in the value chain that consists of designing, building, operating and managing IT systems. Companies can and do outsource every link in their value chain. These links can, however, only be fully exploited if the provider genuinely masters the underlying complexities and can leverage them to generate business value. Value generation and value delivery are thus the central challenges in the outsourcing business. A brief outline of the core elements of these challenges is provided below. Variant forms of outsourcing

It makes sense to begin by explain a number of specific terms used in outsourcing. In many cases, outsourcing is in effect used merely as a synonym for service provision. When they provide “information technology outsourcing” (ITO) services, for example, suppliers generally assume responsibility for a large chunk of the customer company's IT systems, which they then transform and operate in accordance with business principles drawn up and agreed in advance. Selective or partial outsourcing constitutes a further alternative. Along the lines of the business process-centered BPO strategy mentioned earlier, this model involves delegating very narrowly defined IT activities to an outside provider. Examples include call center operation, desktop management, R/3 operation and R/3 application management. In themselves, these individual activities are usually less complex. In the context of transitional outsourcing, service providers are made responsible for a transition or transformation process. They do not merely take what is given to them and commit to comply with agreed targets. This model can be applied when migrating from obsolete to new, more modern technology, for instance – especially where the customer company has neither the resources nor the skills needed to effect such a transition alone. In this scenario, an excellent mastery of both the old and the new technologies

302

Jürgen Frischmuth, Christian Oecking

plus the ability to efficiently manage the transition are the key factors of success. Such developments often necessitate changes in corporate culture, as well as alterations to the framework within which internal employees operate. These are the considerations that make this model so complex. The most sophisticated form of outsourcing, however, remains business process outsourcing (BPO). BPO enables one or more business processes, including the associated IT and administrative services, to be made available from the outside. Here, successful realization hinges on the optimized design and management of the outsourced process or processes, especially with regard to their integration across the companies concerned. Process metrics and the methods of measurement must be designed with great care in this line of outsourcing business. The customer expects outsourcing arrangements to add value. And process metrics are the basis on which value-added targets can be agreed, measured and continually benchmarked. That is why they are so fundamental to this kind of business.

Complexity

Business Models for ORS Business Process Outsourcing Transitional Outsourcing

Responsibility for total customer process

Full IT Outsourcing

Transition management

Transition management

Selective Outsourcing

Total IT responsibility

Total IT responsibility

Total IT responsibility

Selective responsibility

Selective responsibility

Selective responsibility

Selective responsibility

Business value contribution

Figure 70: Variant forms of outsourcing Areas of application for corporate customers

Let us now explore how these various outsourcing models can be applied in corporate practice and what benefits they yield in terms of added business value. The relevant literature identifies distinct clusters of significant business processes, splitting them into core processes on the one hand and support processes on the other. The general assumption is that core processes do not lend themselves to outsourcing, but that outsourcing is definitely a viable option for support processes. As it stands, however, this view is valid only if the design and operation of a company's core processes have already been optimized.

Outsourcing as a strategic management decision

303

The outsourcing of support processes is an obvious consideration whenever companies are looking to sharpen their focus or to restructure. Any of the models discussed can be applied in such cases. Provided the level of complexity is manageable, they can all deliver the expected business value. Realizing such a promise is far more difficult for core processes, however. A company that is already best in class in all its core processes will doubtless feel little compulsion to outsource those processes to a partner firm. Conversely, if not all core processes qualify as best in class, the company should consider selectively outsourcing individual processes or subprocesses and having these run by external providers. One size does not fit all

Different companies that are looking for outsourcing partnerships have differing needs: • Company size within a given market Is the company a local producer of small components or an international manufacturer of complex industrial systems? • Current status of the internal IT landscape Does the company already have a made-to-measure, end-to-end application environment in place, or is it still using proprietary and piecemeal IT systems? • Process reengineering requirements To what extent has the company already invested in (re)designing its business processes? There are so many factors that differ from company to company. Logically, therefore, complex outsourcing partnerships require tailor-made solutions that match each specific situation. Objectives of corporate transformation

Roughly speaking, the objectives of outsourcing partnerships tend to fall into one of the following three categories: • Technology partnership Relatively small contribution to customer benefits and business value • Process partnership Significant contribution to the performance of value-adding processes

304

Jürgen Frischmuth, Christian Oecking

• Enterprise partnership Redesign of the company or one of its markets The more business value an outsourcing partnership is supposed to add, the more vital it is to align this partnership with the customer company's corporate vision. IT always plays an essential part, laying the foundation for extensive and positive change. The scope of outsourcing also determines the nature of the business relationship between the partner companies.

Living out an outsourcing partnership3

Customer value awareness – the fundamental principle

What factors determine the success or failure of an outsourcing partnership? To answer this question, Siemens Business Services introduced the principle of “customer value awareness”. The principle helps to ensure that the value generated by outsourcing always remains the focal point of the outsourcing relationship. Equally, it is important for both partners to remain consciously aware of the value issue. The principle is therefore designed to foster open communication between the outsourcing customer and the outsourcing supplier such that both parties always know exactly what each other's goals are and how these can be met. Strictly speaking, “value awareness” would be more accurate than “customer value awareness”, precisely because the principle applies to both sides of the equation. By using the term “customer value awareness”, however, the service provider underscores its keen understanding of the value it must continually deliver to the customer. The term “value” is used here in its widest sense. Value and perceptions thereof, innovation and flexibility, trust and the monitoring of performance – these are the parameters that stake out the playing field within which customer value awareness must be lived out (see figure 71). Value and perceptions of value in an outsourcing context

Given the sheer variety of business situations, there can be no hard and fast rules about what form of value is most important in an outsourcing partnership. Again, to provide a rough classification, let us refer back to the two fundamental goals of outsourcing projects that we defined earlier:

Outsourcing as a strategic management decision

305

Value/value perception

Customer value awareness Trust/ monitoring

Innovation/ flexibility

Figure 71: The parameters that define customer value awareness

• Cost performance improvements • Business performance improvements Outsourcing projects that aim to improve cost performance generally attach most importance to tactical value, i.e. value that can be realized at short notice. This means cutting costs, improving cash flow and/or liquidity, freeing up capital or avoiding the need to invest, for example. It is also important to improve performance, capacity and/or the quality of services in this context. Increasingly, however, long-term and more strategic issues are also rising to prominence in the expectations of outsourcing customers. This shift of focus lines up with the current trend toward outsourcing projects that deliver more concrete, measurable benefits. In such projects, customers expect to benefit from the operational experience, innovative strengths and execution capabilities of the service provider. Alternatively, they demand flexible access to world-beating capacity and skills. New technologies or the service provider's local presence in different countries might also make it easier for the customer to access key global markets. One frequently cited benefit is that outsourcing frees up customers to concentrate on their own core competencies. In addition, it can give them a keener competitive edge by revamping processes, optimizing IT architectures, and creating optimal conditions for organizational changes with the support of the service provider. At the same time, outsourcing can also be linked to the customer's business goals in a way that realizes short-term objectives. In recent years, it has become more common for the cost of outsourcing to remain variable as a function of the customer's business performance. This gives companies a flexible way to buy in services on demand.

306

Jürgen Frischmuth, Christian Oecking

Tactical (short-term) value

Strategic (long-term) value

•

Lower or variable IT costs

• Focus on core competencies

•

Improved cash flow/liquidity

•

•

Freed capital resources

Access to the resources of a global service provider

•

Avoidance of capital spending

•

Easier access to global markets

•

Better performance and quality, more capacity

•

Keener competitive edge

•

Fast innovation and execution

•

Ability to revector, focus and revitalize the company

•

Greater customer satisfaction

•

Costs linked flexibly to business parameters

Figure 72: Types of value that outsourcing business seeks to increase

When customers entrust responsibility to a service provider, they expect to get better service in return. During the transitional and/or redesign phase, it is nevertheless possible that the services provided may need to be coordinated more closely or fine-tuned. Thorough preparation for this transitional phase can help both the customer and the service provider to avoid any of a number of potential pitfalls. The latter can be rooted in contractually agreed standardization, the technological renewal of the customer's infrastructure or the need to reengineer processes, for instance. These examples clearly illustrate the extent to which constantly changing market conditions influence the success or failure of outsourcing projects. Both parties must be aware of this fact and make the necessary preparations. Innovation and flexibility

Outsourcing contracts are long-term arrangements. Taking the status quo as their point of departure, they define a target status that is to be realized by the service partner and the customer acting in concert. Precise contractual agreements are drawn up on the basis of a detailed due diligence analysis of the conditions currently in place. In an age where everything is constantly in flux, however, these conditions too are shifting permanently. The customer's business situation may change, for example. New technologies may alter the market. New ways of executing certain processes may be discovered. Both parties must therefore mentally prepare themselves and be in a position to respond flexibly when the need arises. Outsourcing partners have certain expectations of each other. Customer

Outsourcing as a strategic management decision

307

companies usually expect to be regularly informed of innovations in the relevant technologies and IT service areas. On the other hand, service providers want their customer to keep them up to date on any relevant changes to their business situation and context. In many cases, such changes quickly mean that existing contractual agreements no longer optimally serve the goals of one or other of the outsourcing partners. Accordingly, outsourcing contracts should always include a “flexibility clause” that specifies the procedure to be adopted in such an eventuality. The following issues, for example, should be dealt with in this clause: • A list of those issues and areas that are likely to experience the most frequent changes (e.g. technology, business processes and markets) • A standard process to be adopted to respond to such changes • Details of who is to be involved in the decision-making process • Provisions concerning the nature of the change • A description of the impact of change on the contract's original goals • A description of the new goals to be realized as a result of the change • Details of action to be taken to realize the new goals • Details of/agreement on what other contractual terms must also be adapted. Given such a dynamic context, service providers must evidence a strong commitment to innovation. Yet they must also remember that technological innovation is no longer regarded as an end in itself. Service providers should therefore constantly screen new technologies to see how far they can be deployed in practice to better serve the customer's business objectives. Technological innovation is, however, only one of the change issues that must be addressed in the context of outsourcing. Equal importance must be attached to innovations that relate to outsourcing business models and business processes. Siemens Business Services pays close attention to both aspects as it tracks relevant changes and developments. Nailing down the flexibility cause takes a lot of time. But it is worth the effort. A well-formulated flexibility cause builds a framework within which the contractual relationship can regularly be adapted in response to constant changes. By providing sufficient room for flexibility, this clause also builds a solid platform of continuity to underpin the ongoing success of the outsourcing relationship. Either party to the contract must closely examine its own activities if this clause is to be filled with useful, relevant content. The subject of innovation in particular must be discussed at regular intervals. And again, these discussions should not focus solely on technological innovations: New developments in business models and processes must also be analyzed

308

Jürgen Frischmuth, Christian Oecking

in light of their importance to the goals defined in the outsourcing contract.

Factors of success once the contract has been signed Once an outsourcing contract has been signed and sealed, a number of factors determine whether it actually produces the desired outcomes. First and foremost, top management must stay rigorously focused on strategic corporate development. It is imperative to describe and define expectations and performance as precisely as possible in the contract. This applies not only to the individual IT services themselves and to detailed documentation of service level agreements, but also to any transfer of personnel or infrastructure, for example. Trust and communication are the bedrock of every outsourcing partnership. Without them, service providers cannot hope to quickly identify the changing needs and requirements of the outsourcing customer. And if they cannot do that, they cannot adopt a proactive approach and give the customer company the benefit of innovation. Without deep trust and open communication, neither party will be genuinely satisfied with the outsourcing relationship. The management of an outsourcing project should be entrusted to experienced staff so that mistakes that have caused large and complex projects to fail in the past can be avoided in future. Senior management on either side should give high priority to cultivating this partnership.

Conclusion Outsourcing business is part of a whole new way of looking at economic activity. Perhaps “business value generation” is the term that best sums up the underlying philosophy. For many of the major strategic challenges facing today's companies, professional, appropriately dimensioned information technology is a critical factor of success. Each company is therefore called on to focus sharply on its core business areas – especially on those that revolve around mission-critical knowledge. All other processes can be outsourced more or less in their entirety to external service providers, whose execution must measure up to the company's own business performance ratios. At the same time, the business relationship must be structured in such a way that the service provider can realistically have a positive impact on these ratios.

Outsourcing as a strategic management decision

309

Outsourcing, in other words, is the best-in-class way to achieve IT excellence, to enable companies to realize their strategies more efficiently and thereby to sustainable increase their value. Another important “side-effect” is that this arrangement helps relieve the one true bottleneck faced by every company: the time that top management can devote to customers and employees. In all its different nuances and variations, outsourcing has gradually emerged as a strategic management issue – an issue that thus demands the attention of every managing board. Whether a company ultimately decides to farm out its desktop management, to outsource its telecommunications or data center services, or to establish strategic partnerships to increase the value of the company, the key factors of success always remain the same: the efficient management of complex business processes and the provision of external management resources. As information and communication continue to converge, it is more important than ever for companies to strive for technological excellence, efficient execution and fast business delivery. The logical conclusion is therefore that every company should carefully explore the potential benefits of a strategic outsourcing partnership and validate its findings in collaboration with market-leading providers. Bibliography 1. Oecking, C., Westerhoff, T.: Typische Problemstellungen und Erfahrungen in IT Outsourcing-Projekten, in: Gründer, T.: IT Outsourcing, Munich, 2004. 2. Ferussi Ross, Christine: Key IT Outsourcing Lessons From Reference Clients, Forrester TechStrategy(tm) Research, Cambridge 2003 3. Oecking, C., Westerhoff, T.: Erfolgsfaktoren langfristiger OutsourcingBeziehungen, in: Köhler-Frost, W.: Outsourcing, Berlin 2004

Johann Csecsinovits

Less costs, more functionality

The intensive discussion on the subject of outsourcing in the management literature of past years, which is also reflected in this chapter, is closely connected with the potential for change in business processes as a result of new communication technologies. There are a multitude of new technologies that permit or favor the outsourcing of processes and functions which were previously integrated in the company.

Outsourcing and the transaction cost economy A graphic explanation of the trend towards outsourcing and the role of communication technologies is furnished by the transaction cost economy introduced by the British economist and Nobel Prize winner, Ronald Coase. The transaction cost economy essentially means that companies form because the standardized co-operation of the individual employees in a company ensures efficiency and cost reduction due to lower transaction costs. Consequently, the amount of transaction costs determines the tendency towards agglomeration and integration. In an industrial world with little networking, and from the point of view of higher transaction costs, it was necessary to integrate all company-relevant processes, transactions and interactions in the company – even those that do not belong to the actual business objective or to the key competencies. The decline in transaction costs resulting from the communication technologies has, however, created a new situation for many companies: It has become cheaper to outsource certain company functions. An appropriate outsourcing strategy can allow a company to concentrate completely on its core competencies and to delegate all other functions to a third party. As a result, the costs for processes which lie outside of the actual main business and reduce the profit that is produced by the primary business line can be considerably reduced.

312

Johann Csecsinovits

Outsourcing – a subject for telephony and IT Outsourcing strategies also play a role in the sector which – in the sense of the transaction cost economy – had initiated many outsourcing projects: the telecom and IT sector. The outsourcing of this sector must be seen as an opportunity to increase the qualitative level of performance at the same time as reducing costs. Today, many large companies are facing the subject of outsourcing the telecom sector: Telecommunications is the key competence of only a minority of companies. Usually, it much more a means to an end. Particularly in times of intensive international competition – and this trend in the economy looks like it’s here to stay – it is not economic to commit resources in the company to telecommunications, which would be better concentrated on the core activities and competitiveness. Another argument in favor of outsourcing is the enormous complexity in the telecom and IT sector. The increased requirements in mobility and rapid technological development, for instance, also play a role as do the question of costs. The market liberalization in the telecom sector has led to considerable cost reductions, but also to a lack of transparency and loss of security. It is extremely difficult for companies in the remarkably diverse provider scene to attain a valid overview of the cost structures which, in reality, are incredibly difficult to compare. Here, too, outsourcing to a professional service provider can make economic sense. Ultimately, in the sense of transaction cost economy, this concerns a major company-strategic decision on whether and in which form the telecom and IT sector is outsourced. Companies who wish to concentrate on their core competencies cannot avoid direct confrontation with this issue. The fact that such a strategy is not only incorporated in theory (although, with regard to outsourcing, this theory has been rather exhausted), but also in business practice, is demonstrated by MAGNA STEYR in its outsourcing of telephony.

Built to order in grand style Particularly in a globally positioned company such as MAGNA STEYR, the question of focusing on central proficiencies is a question crucial to its success. The convincing performance of the company is closely linked with the strategy to strengthen and to further develop its core competencies. The primary business of the company is clearly defined: The development and manufacture of complete vehicles, the development and produc-

Less costs, more functionality

313

tion of components and systems as well as innovations for the automotive future. The spectrum of services and products is unique in its composition. The objective: To be the most efficient and successful company in the automotive supplier industry, distinguished by continuous, disciplined growth. Milestones on this path: For the first time, the provisional “soundbarrier” of 1,000 manufactured vehicles was achieved in one single day of production in May 2004 at the MAGNA STEYR factory of Graz Thondorf. Yet in 1995 the average daily production in the Graz factory was 101 vehicles. The year before, an average of only 30 vehicles per day was built there. This corresponds to a threefold increase in daily production. With around 9,000 employees in Graz, six different types of vehicle are currently built for three major clients, including DaimlerChrysler (the allterrain vehicles of the Mercedes Benz G class, the four-wheel drive 4MATIC models of the Mercedes Benz E class, Jeep Grand Cherokee and Chrysler Voyager), Saab (9-3 Cabrio) and BMW (X3). These data and facts of business activity make clear that the requirements of the telecom and IT sector within the company are challenging and highly complex because these orders can only be successfully transacted by superior networking with the respective clients.

Cost reduction, cost transparency and innovations through outsourcing For a number of reasons we have resolved to outsource telephony within the framework of an operating model. Our central objective: To combine all relevant services of the telephony environment in a single integral model that should be made available by one prime contractor who is also responsible for the service. In future our company should buy in additional services, but no hardware or infrastructure. Our service provisions should be monitored via service levels and the billing should be based on our internal conventions for each service user and his respective cost center. Infrastructure investments made from the side of Magna Steyr should also be incorporated in the operating model – i.e. the future operator must purchase the existing telephony infrastructure from MAGNA STEYR. Our most important motives for the selection of an operating model were and are, cost reduction and cost transparency. But the participation in company-relevant innovations in the telecom and IT sector was also an impor-

314

Johann Csecsinovits

tant reason for us. Following these guidelines, we searched for a partner who developed innovative flexible solutions for telephony that would allow us to fulfill customer wishes quickly, to qualitatively high quality and with absolute reliability to an even greater extent. Our outsourcing project was carried out on the basis of detailed technical and economic requirements specifications. With respect to the economic requirements, our target is for the annual communication costs for our company in the first year of adopting the system to be 10% lower than the reference year of the 2002. This concerns the costs for the entire telecommunications with a comparable call volume and comparable user numbers. Increases in the call volume and in the number of users or extensions were excluded. Above all, substantial cost reductions for MAGNA STEYR should be achieved through the use of optimized solutions and technologies (e.g. GSM VPN) and selective provider choice.

Complex requirements With respect to our technical and organizational requirements, a multitude of complex guidelines were listed in the specification document for the potential operator, including: • Integration in our existing telephone infrastructure, call number portability and the integration of external companies; • The timely provision of necessary personnel capacities; • Provider management including analysis, market monitoring and contract negotiation; • Analysis of future developments within the framework of strategy workshops and a continuous improvement process; • All questions concerning our billing (units, costs for moves, adds, changes, upgrades, call cost reports, separation of service and private calls); • The requirements and key data of our telephony (number of users, cordless telephony, availability of voice channels, network coverage, GSM VPN implementation, landline telephony, emergency scenarios, traffic density); • The requirements of the Help Desk: – the Service Level Agreement with the required service level, reaction times, troubleshooting times, maintenance work, escalation management, training)

Less costs, more functionality

315

A consortium, consisting of the general contractor Siemens as well as TMobile and UTA, was able to completely fulfill our comprehensive and very specific profile of requirements. Since April 2004 this consortium has been responsible for the management of the telecommunication infrastructure for MAGNA STEYR. Siemens Information and Communication Networks (ICN) is available to us as a central contact partner for all telephony matters.

Positive outcome of outsourcing The result of this outsourcing model, which follows the Siemens customer philosophy of “Managed Services”, is extremely positive for MAGNA STEYR. The most important data for success: • A new billing model ensures maximum transparency: Extensions are billed on a monthly price per participant, i.e. hardware and software are only paid if they are actually used by the employee. • Provider costs are also completely plannable: Together with its partners, Siemens was able to develop a flat rate charging model which is one of the remarkable innovations of this project. There are only two rates, one for the mobile network and one for the landline network. Trends in cost development can therefore be recognized at an early stage. • There is also a substantial innovation in the virtual GSM network. For many years our site in Graz has been supplied with a Hicom solution from Siemens. An expanded DECT system guarantees mobility throughout the entire business premises for approximately 5000 participants. The new system also makes it possible to reach employees away from business premises in the same way as on the company campus. A Virtual Private Network based on GSM enables the “virtual extension”. There is a standard dial plan and mobile telephones can be dialed using internal direct dialing. Internal telephone calls are generally free of charge. Thus, with this operating model, we benefit not only from savings, but also from an extended functionality. • The future is also assured: The introduction into Voice over IP and UMTS is technically prepared. A task force, made up of members of the consortium and MAGNA STEYR, continuously monitors the market as well as the newest technical possibilities in the mobility sector and investigates new developments with regard to possible applications. This is an important service so that we can promptly anticipate and take advantage

316

Johann Csecsinovits

of new technologies or new solutions. Due to the outsourcing of the telephony sector, MAGNA STEYR can now confine itself to monitoring, controlling and quality assurance measures of services. For us, this means specifically: • We no longer pay for equipment, but just for the functionality. • We have one contact partner and a future-assured telecom solution from a single source – on site we have one operations coordinator from Siemens available with a service team prepared for the respective requirements. • We know from regular reports how, for example, piece counts and call behavior develop, where there are conspicuous trends and can jointly implement countermeasures or take steps to achieve the required developments. As a result, the company can concentrate better than ever on its real core business, i.e. vehicle production and development.

Success factors in outsourcing As international discussion shows, specific success factors play a role in outsourcing projects: This is the reason why there are successful and less successful outsourcing projects. According to studies, the failure quota of outsourcing is around 70 percent. Such data show that the complexity of projects is frequently underestimated in outsourcing. From our experience the following success factors, which have played or play a special role in the success of our project, should be highlighted: Realistic expectations: The expectations of an outsourcing project must be realistic. If expectations are set too high, the foundation stone for the failure of the project is already set with the signing of the contract. Expectations are only realistic if they embody a win-win situation: both the customer and contractor must profit from the outsourcing project. A second aspect is the precise definition of the requirements: We have determined these point for point in a jointly developed set of agreements with our partner. Partnership & networking: As in many aspects of economics, it is also true in this case: outsourcing is primarily a question of trust. Outsourcing partners must be trustworthy partners. Particularly in the telecom and IT sector it is true that the insight into internal business processes and proce-

Less costs, more functionality

317

dures is an essential incentive for optimum solutions. If the provider does not have this knowledge or he cannot provide it, the best technological solutions will fail due their lack of compatibility with business processes and structures. The dovetailing of the personnel by a single operation coordinator and a powerful service team from Siemens at MAGNA STEYR demonstrates the close personnel networking. Another important dimension of this partnership is also the instructions to our partner to advance the company’s development with the solutions and strategies employed in the telecom and IT sector. This requires a partner who holds a leading position with regard to technology and the future. Cost cuts & planning security: The third key factor for success in an outsourcing project is the question of cost cutting and planning safeguards. The central economic requirement in this context is clear: After outsourcing, the outsourced area must result in long-term lower costs for the company than beforehand. We have achieved absolute cost and planning indemnity with our outsourcing project, which is a something really innovative for an operating model in this field. The flat rate model for telephony costs developed by Siemens and implemented together with its partners is an example of how cost reductions can be achieved in the areas where it hardly seemed possible at first glance. Cost and planning indemnity means that we pay a fixed contribution for the use of each terminal or that there is a monthly flat rate in the call center service per agent, etc. Today we already know, in the framework of our five-year contract, exactly what we will have to pay out for telephony in the future. This is an absolutely central factor for the success of outsourcing. Information policy & marketing: A fourth very important factor in our cooperation model is the continuous information work and the marketing for new telecom subjects (e.g. new services, new equipment). The IT department of MAGNA STEYR is, after all, a service provider in the company. It is vital to market products and services, which are new or additionally available within the framework of the operating model, in such a way to customers that the positive effects related to the products and services can be achieved. Siemens takes this important function in the company seriously and uses different communication channels (e.g. Internet) for this purpose. Our IT department can be reduced to the function of an initiator which prioritizes upcoming projects and, as a security policy for the correctness of the selected solution, trusts market forces. The benefits of outsourcing concepts depend not least on the fact that the communication of the related services and advantages is designed correspondingly professionally and results-oriented.

318

Johann Csecsinovits

Professionalism is decisive From our point of view the outsourcing question is clearly answered in the affirmative by the concrete example of telephony. The outsourcing of this sector represents a real gain for the company: We have a solution partner, who provides solutions for our communicative requirements from a total perspective. An essential condition for the success of the outsourcing is doubtlessly also the professionalism of both partners: Whoever outsources company areas must also be aware of the precise reasons why he is doing so and what his objectives are; whoever undertakes this task must be able to optimally solve the technological, conceptual, economical, communication-political and company strategic requirements. This is the exact basis for outsourcing, which brings cost savings and greater functionality. The outsourcing of telephony will play an increasingly important role in the future. The use of mobile radio and new communication technologies will also become more important for smaller and medium-sized companies, but due to strong competition, it will become ever more difficult to see through the large numbers of providers and the challenging technologies. It is totally uneconomical and inefficient to expose the company to risks with regard to the choice of technologies or to tie up personnel resources for telephony when attractive operating models are available on the market.

VII. Challenges facing qualification and management

Markets have become more complex and company relationships more intricate. At the same time, product life cycles have become shorter and customers more demanding. Real-time companies must therefore cope more often with anomalous situations and make important decisions in the shortest periods of time, in both the strategic and operational areas. As a result, the ability to transform information into negotiating power and speed of decision are decisive factors for success.

Dieter Spath, Rita Nøstdal, Martina Göhring

Visual management Concept for the support of strategy, planning and control in real-time

Real-time companies must cope more and more often with extremely anomalous situations and make exceptional decisions in the shortest periods of time in both the strategic and operational areas. As a result, the ability to transform information into negotiating power and speed of decision are decisive factors for success. This article describes how a center for visual management – also known as the “Business War Room” – effectively the nerve center of the company in the form of an electronic environment where people from different locations working together and concurrently using a multitude of information technologies, can make a significant contribution to management in real-time.

Introduction In today’s globally-oriented business world, markets have become more complex and company relationships more convoluted. At the same time, product life cycles have become shorter and customers more demanding. Consequently, companies are often obliged to cope with exceptional situations, and to take unusual strategic and operational decisions more and more frequently. The ability to make real-time-based decisions and to implement these immediately in targeted measures is one of the key factors for success in real-time companies. One premise for this is efficient performance management that provides the management and decision takers with the relevant, operational and analytical information “on time”. If this is initially provided by bringing together and integrating the frequently heterogeneous information and sources of knowledge within the company, the management is called on to transform this information into actions and ultimately into customer benefits. As a catalyst, a center for visual management can

322

Dieter Spath, Rita Nøstdal, Martina Göhring

support the unfettering of the creative analytical potential of human beings with this inherent management task.

The center for visual management as a control center A center for visual management (the expression “Business War Room” is also used for this) is the nerve center of a company – namely in the form of an electronic environment where people from different areas and locations work together and simultaneously use a large number of information technologies.

Figure 73: The control center and the nerve center of the company

The objective of this extremely focused and intensive co-operation is to take decisions on the basis of constantly changing conditions in order to be able to quickly design complex programs, for example, or to develop strategic plans. So that decision situations can be analyzed and the effects of these decisions simulated, it is necessary to prepare information in the form of structured data, to clearly visualize it and to combine it with the knowledge of the persons involved. The simultaneous interaction of human and electronic networks promotes “extreme collaboration” which gives rise to “real-time” exceptional services through the interaction of human and electronic networks. A series of studies has proven that the use of a center for visual management results in gains in productivity that are well above average as well as increases in

Visual management

323

performance and considerably more rapid results than in conventional working environments.1 Such company control centers have their origin in military applications and, especially in big business, have been recognized and in use for quite some time. While applications were rather analytically-strategically aligned up to now, modern centers for visual management are based on real-time control and the integration of operational and analytic-strategic worlds in the enterprise as well as the connection of distributed and mobile users.

Business Performance Management • Competition analysis • Risk and crisis management • Strategy development and management

Software development • Distributed process management • Cooperative SW development • Assessment and selection Application fields Technologies

Integrated

Visual

VISA

Production Business Performance • Digital production planning • Maintenance management • Performance Product development management • Management of complex product development projects • Technology analysis Mobile

Information analysis • Analytics (mining, Cooperative modelanalysis, reporting..) Characteristics ing techniques and • Simulation and tools Interaction and navigation scenarios Visualization techniques: • Spatial interaction • Immersive 3D visualization • Cooperative interaction and • Information walls navigation Collaborative

Figure 74: Visual management at the Fraunhofer IAO – Interaction of technology, information and human beings

Company application fields extend from the monitoring and control of the creation of services or delivery (real-time process management) up to strategic development and evaluation. In addition, they comprise subject areas such as software development, production management, product development and company control. Figure 74 shows important features and selected application fields of a center for visual management, which are currently emerging at the Fraunhofer IAO.

See, e.g. Covi/Olson/Rocco (1998), Teasley/Covi/Krishnan/Olson (2000) or Williams/Kessler (2000).

1

324

Dieter Spath, Rita Nøstdal, Martina Göhring

Company management in real time by performance management The model of corporate or business performance management describes a comprehensive management concept for the unified control of a company. This full integration is expressed by the fact that all relevant management disciplines are considered for the success of the enterprise. Performance management describes the concrete methods for the control of these management disciplines. Performance management tools

In a long-term study (see Nohria, Joyce, Roberson 2003) it was proven that most management instruments and techniques influence the company performance indirectly at best. The most successful companies are more likely distinguished in the following primary management disciplines: strategy, execution, culture and structure. They are supported by at least two of the following secondary management disciplines: talents, innovation, leadership and mergers/partnerships. Specific key factors for success are behind each management discipline, which can be fulfilled well or poorly. If these should be fulfilled particularly well, e.g. to be more successful than the competition, this can be encouraged by performance management. The objective of performance management must be to support these management disciplines so that they can be successfully integrated in the real-time enterprise. For this, they must undergo continuous assessment, they must be coordinated and the results must be continuously and consistently communicated. This section should illustrate the possibilities that are provided by performance management in order to support the management disciplines responsible for company success raised in this study. For this purpose, performance management building blocks that affect the management disciplines in different ways are available. Thus, individual instruments, technologies, systems or procedures (i.e. tools here) can exercise its specific effect. An optimum potpourri of deployed tools ensures the best degree of effectiveness of performance management. A selection of important tools for the control of management disciplines is shown in table 4.

Visual management

325

Table 4: Building blocks of performance management and the tools for the support of management disciplines Building blocks of performance management Management dis- Business Intelliciplines (accord- gence ing to Nohria and others 2003) Primary: Reporting, Goal Strategy Metrics (BSC), Upto-Date Info, Dashboards

Intellectual Capital

Competitive Intelligence

Cultural Transformation

Knowledge Management, Relationship Management

Market Level, Trend Analysis, Clustering

Action Model

Benchmarking, Stability Analysis, Potential Analysis Self Organization, Change Management

Execution

Process Indicators, On-demand Analysis for Business Problems, Ad-Hoc Reporting, Anomalies

Value Added Analysis, Customer Value

Culture

Availability of Data at AP level

Structure

Knowledge Management, Cooperation System

Appreciation, Value Benchmarking Measurement, Usage Criteria Scorecard, Best of Location Factors, Breed Early Warning Systems

Secondary: Talents

Innovation

Leadership

Knowledge Management, Training Management R&D numbers, Analytical System

Knowledge Assessment, Intellectual Capital Due Diligence, Technology Scorecard Value Analytics

Early Warning System, Analytical System Mergers/partners Financial Analysis, Due Diligence, Decision Support Competence Criteria

Self Control by Employees Cooperation System, Change Management

Benchmarking, Personnel Codes Recruitment Assessment Benchmarking, Innovation BaromRisk Management eter Trend monitoring

Relationship Management Industry Analysis, Change ManageBenchmarking ment, Risk Management

Management disciplines as application fields for performance management APPLICATION FIELD ‘STRATEGY’: The key factors for the success of the strategy are its clear definition, standard communication of the strategy to customers, employees and shareholders, and maintaining the strategy for a sustained period. Performance management supports this with instruments which, on the one hand, can provide transparency and clarity such as scorecards, and, on the other hand, through the coupling of strategy with business results by means of metrics, whereby successes and deviations

326

Dieter Spath, Rita Nøstdal, Martina Göhring

from the strategic plan can be identified more quickly. So-called action models provide the methodical basis required for this. APPLICATION FIELD ‘EXECUTION’: It is obvious that there is no proven relationship between the outsourcing of a company and its overall financial performance. On the contrary, it is more a question of concentrating resolutely and routinely on the business processes and to continuously optimize the operational sequences. Each year, market-leading companies have to raise their productivity twice as much as moderately successful companies. The use of technologies and instruments only plays a decisive role if this can be assessed later as to whether it can contribute to the lowering of costs or to an increase in operational performance. Performance management contributes to the optimization of the efficiency of processes and discovers the suitable product and service quality for the customer. It is a matter of regulating the speed of processes so that – in the figurative sense – it is possible to brake just in time before a bend as well as to accelerate out of it in time. Normally business processes no longer run sequentially, but in parallel and networked as, for example, in supply chains or virtual companies. Here it makes sense to synchronize business, support and control processes that are running in parallel. The improvement of the quality of implementation by performance management can best be illustrated by the use of an example. A company recognizes, for example, a weakness in their process because customers complain of excessive implementation times. Typically, process analysis is now carried out, the weak points are identified and optimized. In the best case, this state continues for a while until major weaknesses surface again. Perhaps the customer complains again about overly long implementation times even if they have not really changed – but the expectations of the customers have simply shifted. Or he has developed other requirements and new wishes that are either not fulfilled at all or only fulfilled to an unsatisfactory extent. Application example of product development

In the framework of the “VISA in product development" project at the Competence Center R&D management of the Fraunhofer IAO, a central, integrated and transparent management information system with corresponding visualization option was developed by Frank Wagner and Holger Eckstein in co-operation with a southern German vehicle manufacturer. “VISA in product development” uncovers deviations within complex product development projects in real-time and supports flow

Visual management

327

improvements. Accordingly positioned metrics as well as linked technologies and applications create transparency in the development process and inform the decision makers of the status and progress of the product development project. In this context, it has turned out that the visualization of immediately accessible and aggregated data and information plays a prominent role. There is an opportunity here to semantically represent complex situations and relevant multi-dimensional datasets and to present it in a simple, tangible display format for the decision makers. Due to the globalization of markets and the internationalization of companies, it is necessary to support the increasing mobility of decision takers through the integration of mobile I&C technologies and applications as well. Consequently, the taking of short-term decisions, e.g. changes of product specifications due to unexpectedly poor product tests or problems with development partners, are considerably eased or even made possible. Altogether, this substantially minimizes the risk of the development project for the company. Also, real-time access to all project specific data and information allows the early detection of risks, reducing the appearance and frequency of the often short-term – and thus acutely risky – troubleshooting. Available reference solutions and previously defined alternatives or case studies can be called up at any time and support the decision to taken in the short term by contributing to the finding of a solution. As an outcome, a specific cipher system and supporting methods were developed which enable the contents-related navigation and coordination of complex development projects and the partial processes or work packages contained therein. In the widest sense this also affects the interfaces to preliminary projects such as product planning and preliminary design and follow-up actions such as production planning as well as sales and service.

Again a process shortcoming exists which needs to be identified. Without performance management, this can only be recognized if the complaints are actually put on the table. With performance management, the signs are identified within the process and the point when the behavior of the customer and his requirements change is recognized earlier. For this purpose a process analysis focusing on these performance indicators must be carried out. In other words, within the framework of a process analysis, the performance indicators are systematically identified first, followed by the determination of target variables or benchmarks for these indicators and, consequently, deviations or anomalies within the individual process steps

328

Dieter Spath, Rita Nøstdal, Martina Göhring

can be detected and improved through continuous controlling. As a result, customer changes can be detected by characteristic changes of the indicator. Time consuming, comprehensive process analyses can therefore be saved or should only be carried out if there are massive strategic changes in the plan, e.g. mergers, innovations, new market fields, etc. In “normal” business operation the indicator values must be available at the touch of a button because there is not enough time to carry out continuous costly analyses. Business Intelligence or Performance Support Systems enable this information to be available “on demand”. APPLICATION FIELD ‘CULTURE’: The objective here is to create a working environment for the maximum performance of services. This means motivating individual employees and teams to achieve their peak performance, to develop and support them and to reward their success as well as to ensure the fulfillment of guidelines in addition to raising the yardstick each year. In short, the objective of control is the optimum exploitation of available talents. APPLICATION FIELD ‘STRUCTURE’: It is neither the specific organizational structure, nor the degree of centralization or decentralization that makes a company successful. The key factor for the success of a well functioning organizational structure are simple processes and guidelines which ease the work of each individual and customer, i.e. to a certain extent a debureaucratization of each process step. As a rule, flat hierarchies are more promising here. This includes a high degree of responsibility and flexibility of the individual employee. In order for the employee to make the right decisions, however, he must be able to access all relevant and current information. Performance management can support this by improving the transparency of information across all sectors, hierarchies and company boundaries. Business Intelligence and Knowledge Management Systems enable a comprehensive exchange of information which is reflected in improved cooperation and increased productivity. However, a structure – once optimized – must be regularly checked to eliminate bureaucracy in good time. Early Warning and Change Management Systems can trigger timely warning signals. APPLICATION FIELD ‘TALENTS’: The objective of the company is to invest in efficient employees. On the one hand, this means the further development of the company’s own employees in order to achieve an effective workforce and to be able to staff any free management positions from

Visual management

329

its own ranks. On the other hand, it means targeted external recruiting to advertise for additional top executives. Performance management relies here on an effective and efficient qualification system, supporting the control of the qualification process through Education Controlling and relying on the opportunities of cost reduction and performance measurement of ELearning systems. APPLICATION FIELD ‘INNOVATION’: The aim of the company is to be innovative – either through new products and services, the use of new cost saving techniques or the recognition and mastery of technological advances. Within the scope of the control of innovations, there are insufficient findings in order to exert concrete influence through the tools and instruments of performance management. APPLICATION FIELD ‘LEADERSHIP’: The key factors for success identified in this management discipline for increasing company performance are good relationships with employees and the motivation of the management team. An additional feature is the swift recognition of opportunities and problems. This concerns market changes, politics and population dynamics. Performance management can provide the necessary indicators in the observation area and point in time for this purpose. APPLICATION FIELD ‘MERGERS AND PARTNERSHIPS’: The objective of mergers and partnerships is to both establish values through growth as well as to benefit customer relations and to complement core competencies in addition to exploiting new market sectors. The resources of performance management offer themselves here to a particular degree in that the selection of the correct partners for mergers and partnerships requires profound analyses, evaluations and research. A great number of resources are invested in this as a rule so that the use of systems for decision-making is worthwhile. Visualization of performance management

According to Aristotle “Man never thinks without a picture”. It is no longer possible to imagine the action of performance management on management disciplines across the board – even for a strategist. The empirical search for the relations between strategy and environment for positioning, the checking of static and dynamic influences on a company, the inclusion of additional aspects not considered so far, the networking of strategies or strate-

330

Dieter Spath, Rita Nøstdal, Martina Göhring

gic maneuvering of competitors often make it necessary to design a number of strategic models in order to understand the effect of this interaction and to derive discernable measures for the enterprise. Transferring this to strategic visions involves looking to the future with the knowledge of the past in order to learn from earlier mistakes. Strategic thinking not only requires looking out in all possible directions, but also the simultaneous perception of parallel processes as well as their reactions on the environment. Adding all these perspectives together results in a very complex vision for the strategist, which is barely understandable without visualization techniques. Outlook: Immersive 3D information visualization

In order to intuitively and quickly portray and process company information in an understandable way for people of different technical backgrounds, new types of human-machine interface are provided in many fields. 3D visualization techniques help to meaningfully couple complex product structures with company data. Immersive 3D systems promote constant spatial awareness and 1:1 representation of prototypes, products, processes and information as well as spatial navigation and interaction in product and information structures. These 3D representations provide a basis for decision-making, allowing facts on the product to be debated quickly and efficiently in interdisciplinary teams. The freely configurable 3D visualization unit ‘Compact Display Unit’ developed in the CC Virtual Environments of the Fraunhofer IAO permits immersive visualization technology to be flexibly integrated into larger onsite information environments. On the one hand, Figure 75: 3D representation of a parts list single workplaces can be set up and, on the other hand, team discussion information systems can also be established by combining individual components. This flexibility increases the number of possible applications, thus reducing the investment costs of immersive technology which continues to present a stumbling block to its wider use.

Visual management

331

Integration and architecture

Real-time performance management as described here requires the integration or consolidation of a multitude of applications and information systems within the enterprise. Nevertheless, the objective is not the ultimate technical integration of all information systems, but the provision of all essential and decision-relevant data and information for collaborative interactive analysis and control in the center for visual management. (Existing) Business Intelligence applications constitute a suitable starting point for the realization of a control center for performance management. The analytical applications are coupled with the operational systems (buzzword Business Activity Monitoring) by means of Enterprise Application Integration (EAI) technologies in order to fulfill the real-time requirement. Quantitative information from knowledge management, rival information systems, online services and information, etc. can be made available via corresponding text and web mining technologies for analytical questions. Simulation and scenario technology enable the evaluation of action alternatives while cooperative navigation and interaction concepts ensure simultaneous control in the multilayered information area.

Visual management

Visualization and Navigation MIS Cockpit, Scorecarding

Reporting, CharPlanning, Scenario Technology acteristic Values, OLAP

Data/Text/ Web Mining

Interaction and Communication

Quality Management

Process Management

Simulation

Data Integration

Data Warehouse

Operative Systems (ERP, CRM etc.)

Figure 76: Information and analysis components

Intra/Internet, Knowledge Databases, File Server etc.

332

Dieter Spath, Rita Nøstdal, Martina Göhring

Conclusion In the face of today’s concepts of performance management, the real-time control of companies signifies an expansion in several dimensions. In addition to the requirements of the technical and informational sectors, management has the task of adapting existing management models and methods to make them real-time capable. Particularly in the conversion of a time-specific to an event-specific company control, a balanced relationship must be ensured between (planned) action and reaction. This balancing act on the tightrope between hyper turbulence and stabilization measures is one of the decisive management tasks in the real-time enterprise. The center for visual management brings together and combines the pertinent management dimensions. Nevertheless, despite all the technologies in the world, the human being alone remains responsible for walking this tightrope.

Bibliography Nohria, Nitin; Joyce, William; Roberson, Bruce 2003: Managementmethoden: Was wirklich funktioniert. Eine gründliche Untersuchung zeigt, auf welche Aufgaben sich Manager in ihrem Unternehmen konsequent konzentrieren müssen, um auf Dauer Erfolg zu haben. In Harvard Business Manager 10/2003 Pages 26–43 Martin, Wolfgang 2003: Business Performance Management und Real Time Enterprise. Auf dem Weg zur Information Democracy. Strategic Bulletin BI 2003 ITResearch, Sauerlach near München 2003 Spath, Dieter et. al 2003: Marktstudie Corporate Performance Management. Bausteine, Konzepte, Produkte. Fraunhofer IRB Verlag, Stuttgart 2003

Ferri Abolhassan

From CIO to Chief Process Officer

The Real-Time Enterprise (RTE) consists of more than bits, bytes and technology. According to the current definition, the CIO of an RTE is primarily an administrator of technology. This will also be the case in the future, but his job will be greatly expanded by the networking of systems, technologies, processes and companies. The present internal perspective of a CIO will also be expanded to include the external communication of companies. The CIO of the future will be a generalist who, in particular, unites the complex processes of companies through technology and brings together the most varied elements and processes not only among departments, but also among a number of companies. In this regard, four qualities will be of special importance for the CIO of the future: • • • •

A highly developed ability to communicate A refined process reasoning A high level of social competence The ability to become an agent for change within the company and to actively pursue change management.

This will only be realized in the long term if the CIO becomes a CPO or Chief Process Officer. The technical department owns the process, appoints the IT management and administers the technology. However, with the increasing importance of business processes on business applications and the integration of processes and IT across departments and company boundaries, the question arises of who exactly in the company is responsible for business process management. The real-time enterprise requires new management which promotes a collaborative, dynamic and innovative IT approach. The trend has been noticeable for some time: the focus of IT tasks is moving in the direction of flows and processes. Inevitably this is coupled with a change of the role and tasks of the IT manager. While the IT depart-

334

Ferri Abolhassan

ment has been responsible for information management and the administration of technology up to now, in future it will identify and steer business processes throughout the company. This will cause the Chief Information Officer to become a designer of business processes, providing the information technology platform and infrastructure to enable their execution. What is behind this change of role? In November 2002, an American market research institute called the Gartner Group defined a new era of the information age under the term “Real-Time Enterprise” (RTE). With the concept of a real-time economy, the IT analysts formulated a pertinent answer to the requirements of globalization, which leads to dramatic changes in the reactions and flexibility of companies. At the same time, they described a collaborative type of company which has become possible as the result of new technical developments such as web services and EAI (Enterprise Application Integration). In the real-time enterprise different IT applications and systems are connected on integration platforms, which flexibly steer integrated business processes. The Chief Process Officer whose main task will be the integration of departments, partners and customers will be responsible for the process and IT infrastructure.

Catalog of tasks of the CPO In particular, the CPO must: • Identify, describe and analyze relevant business processes from the point of view of economics and chronology. • Expose and correct weaknesses such as excessive idle periods and unnecessary delays, and optimize processes. • Establish integrated, cross-department and cross-company business processes in a complete value chain with the inclusion of external partners. • Organize business process management with responsibility for processes and partial processes by the process owner. • Ensure integration of internal and external software applications. • Develop and introduce an efficient real-time platform with hardware and software. • Organize continuous control of the RTE processes. • Expand and further develop the RTE system.

From CIO to Chief Process Officer

335

The Chief Process Officer is the enabler of the real-time enterprise which is possible through a series of new developments, but was already formulated as an organizational idea years ago. For the insider, the RTE is a logical step derived from the technological development of the past few years. The theoretical basis for the real-time enterprise was already created in the 70s and 80s with the Management Information System or MIS. But there was then a shortage of powerful software tools and affordable computer performance in order to implement the ideas in economically sound and user-friendly applications. Consequently, real-time concepts remained just sandbox games. Even if realization was only possible following the technical innovations of recent decades, three basic considerations that are still among the essentials of today’s real-time enterprise were already behind MIS: • Information technology as a tool for the implementation of business processes, • Automation of operating and decision-making procedures, • Integration of in-company processes across department boundaries. A great advance in the direction of realization since the 90s has been provided by company-wide ERP (Enterprise Resource Planning) systems which were set up at the workplaces in the course of the networking of companies with Local Area Networks (LAN). They allow IT support of business processes via a standard database which provides the information required for processes and individual operations, and thus ensures uniform information at all points. The advantage of these systems is enormous: they prevent the duplication of work, misunderstandings and co-ordination problems between departments and personnel to a great extent, they simplify and accelerate processes, and improve the efficiency of decisions. New technologies such as web services and modern integration tools such as workflow systems and EAI enable the further development of standard software into a component architecture, where the manufacturer ensures their integration via a single platform. The modularization institutes a greater degree of flexibility for the implementation of business processes. New approaches take the closer meshing of business processes and applications into account, such as those SAP and IDS Scheer are developing with the integration of the ARIS business process software on the Netweaver ERP platform. This process-oriented further development of company software lays the foundations so that business processes can be dynamically changed and new requirements can be quickly adapted – in

336

Ferri Abolhassan

short, that real-time enterprise is possible within the company.

Real-time enterprise requires real-time business processes The emergence of the Internet has not only extended the possibilities by the inclusion of external partners such as customers and suppliers in company processes, but, at the same time, has enormously intensified the time dynamics. If a customer can enter an online order in seconds, he quite rightly expects not just prompt delivery, but also information on the availability of the ordered goods as well as an order confirmation with details of the delivery data. He also expects to be informed of any delays. Technology alone cannot fulfill these requirements; the organization must be adapted to meet such requirements. This example shows that RTE is first and foremost a reaction to the growing importance of customer orientation. The individualization of the buyer’s market has the result that the activities in the company must be increasingly aligned to the customer. E-Business solutions such as Supply Chain Management or Customer Relationship Management are an expression of this trend.

Enterprise Management

Supplier Relationship Management (SRM)

Own business processes

Procurement process

Supplier of the suppliers

Customer Relationship Management (CRM)

Engineering process

Direct supplier

Invoicing

Quality control

Own logistic processes Plan Source

Make

Service & Delivery processes

Customer of Direct customer the customers

Deliver

Supply Chain Management (SCM) Figure 77: Process management across company boundaries

From CIO to Chief Process Officer

337

Renaissance of the process Companies feel this pressure and subsequently attach an ever greater importance to Business Process Management (GPM). This is shown by the results of the Business Process Report 2003, an empirical survey of IDS Scheer in co-operation with the market analysis and consulting company Pierre Audoin Consultants. 150 companies were questioned about the quality and the need for action in their business process management. The most important results: • Four out of five companies are intensely occupied with GPM and are working on the optimization of their processes. • 82 percent of those questioned are happy with the quality of their business processes. This means an increase of 12 percent over the previous year. However, a moderate average score of only 2.88 percent would suggest a sizeable potential for improvement. • Above all, the companies see the benefits of GPM in higher customer satisfaction and the improved quality of the services offered. • Two thirds of those questioned are of the opinion that the company management should take care of business processes. Over half the companies use so-called process owners for the control of business processes, who have the necessary resources and expertise. • Every second company favored the appointment of a Chief Process Officer (CPO) with central responsibility and decision making authority.

The CIO – an out-of-date model? Information management itself is losing importance because IT services are for the general good. This also manifests itself in the trend to outsourcing of IT services and to hosting services. Thus the wheel continues to turn. The CIO replaces the EDV manager who was responsible for the smooth functioning of the hardware in the company. In the following period, the “technician” turns into a manager whose task plays a role in many other areas. Meanwhile, IT becomes diminished in the sense of an IT department, but is considered as a higher-level management task. This revaluation is linked with the growing strategic importance of IT systems in competition. The strong technologically oriented role of IT management has remained. The Harvard Business School has provocatively tackled the subject: A meeting with the title “Should you fire your CIO?” compared IT manage-

338

Ferri Abolhassan

ment with the administration of telephone systems or electricity supply and referred to the fact that no Chief Telephone Officer nor Chief Power Officer can be found in any company. Ten years ago, so it was argued, hardly anyone had even heard of a CIO but, in the subsequent ten years, the knowledge of how to efficiently use information became very widespread and no one required a CIO any longer. As radically as it was posed by the Harvard Business School, the question is posed in the fewest companies. But the changes make it necessary to separate themselves from the many activities and deep-rooted reasoning of IT management. Thus, it was assumed up to now that the CIO had to ensure an IT that is as efficient as possible, because this automatically leads to efficient business processes. This IT-centered viewpoint determines the line of activity of the CIO. However, increasingly more IT resources are administrated today by external data processing centers and software can be leased. The supply of information comes increasingly from the wall socket which makes a costly internal IT management partly obsolete. The position of the CIO as an “administrator of technologies” becomes increasingly superfluous as a result. Instead, an integrated viewpoint of organization and technology is required, exactly like that in the networking of systems and processes – and, in fact, both in-house as well as across company boundaries.

From CIO to CPO – more than just a change of name As a result, the CIO turns into the CPO – Chief Process Officer. A change of name alone is not enough. The company management and the technical departments expect that the CPO recognizes the innovation potentials of new IT applications and technologies, and can transfer this into business processes. The technical departments will also remain the process owner in the future. No one knows the processes in sales, logistics or purchasing better than the responsible departments. However, the CPO must judge the interaction between IT and the processes. A three-level model can help explain the structure and organization of the business process management and the integration of the CIO transformed into a CPO: • At the top level, the so-called C-Level management (with CEO, COO, CIO, CFO, etc.) decides on processes of strategic importance. Above all,

From CIO to Chief Process Officer

339

this concerns core competencies which companies put into their products. A basic condition of business process management in the real-time enterprise is the certitude at board level that is best documented with the set up of a CPO position. The general alignment of business process management and the conception and introduction of methods, tools and platforms belong to the central responsibility of a CPO. • The business processes linked with IT are carried out at the 2nd model level. Here the decentralization determines events because, above all, the process knowledge is lodged in the heads of the employees who work in the processes. The CPO must ensure that this decentralized process knowledge can be used centrally for process improvements and, hence, is available to all process owners. As a result, process management is the task of each individual employee. • At the 3rd level, the cycle is closed, the results of business processes are recorded and evaluated, and prepared for correction and control decisions. Here technological developments could be seen to emerge which merge levels 2 and 3 and lead to new architectures. This bundles recognized technologies such as workflow systems or EAI systems and combined them with execution systems to innovative RTE systems (see following points). Controlling aspects also gain importance as increasingly more process owners become interested in them.

Figure 78: Three-level model of business process management

340

Ferri Abolhassan

RTE platform as the main responsibility of the CPO • The establishment, expansion and further development of a real-time platform are one of the central tasks of the CPO. It consists of three components: • Process design using business process optimization tools such as the ARIS Process Platform from IDS Scheer. • Process execution in the application system. Usually this concerns an ERP system such as the mySAP platform for example. With the use of software from various manufacturers, integration is decisive for fulfillment of the RTE conditions. EAI tools or a web services platform can be coalesced to form heterogeneous IT landscapes – which, however, are intensely different from the process point of view. Consequently, EAI technologies are actually in a position to simplify interfaces, but still do not support execution of functions. It is different in web service concepts such as Microsoft .NET or SAP Netweaver. For example, Netweaver enables companies not just to integrate SAP applications in their IT environment, but also third-party systems. The integration of ARIS in Netweaver ensures the integration of processes across all systems. • Process controlling with recording, measurement and monitoring tools such as the Process Performance Manager (PPM) from IDS Scheer. These tools permit the measurement of Key Performance Indicators (KPI) such as processing time, key quality data and process costs. Thus, for example, the software automatically registers the deadlines of process throughput such as the processing of orders by time stamp. The allocation of results to the business processes provides the possibility for proactive action so that management can still take prompt corrective measures to “real-time” recorded deviations in many cases. With the integration of the Aris process platform in Netweaver, the two previously separate steps of process modeling and process execution grow together via the standard software. That’s why the organization of business processes moves to the foreground for users and the software technology is oriented to optimized company processes.

Four skills of the CIO of the future The trend in the direction of flows and processes blatantly requires a strong orientation from IT management towards economic aspects. An integral

From CIO to Chief Process Officer

341

part of this is an increase in the scope of responsibility because market changes, competitive requirements and technological developments demand the permanent restructuring of business processes and this, in turn, requires more competencies than merely the responsibility for IT systems. The new CIO subsequently becomes an agent for change within the company, who takes care of adaptable, dynamic and collaborative processes and systems. Four management skills are particularly important for this role: • The ability to communicate. The real-time enterprise only functions if internal and external partners pull together. Information and clarification are therefore prerequisites, and communication is thus a basic condition for successful realization. It is known from studies in American companies that many CIOs have an unfulfilled requirement in this respect. According to a study by the CIO magazine, the senior IT manager spends only 15 percent of his time on discussion with his own employees. • Process reasoning. Examining the value chain with technical, economical and organizational know-how and interpreting the whole in relation to dynamic processes is part of the stock-in-trade of a real-time CIO. • Social competence. RTE means teamwork and this is turn signifies that business processes only run successfully in the long-term if the employee sees himself as a real-time worker. This is particularly true if the CIO wishes to implement change processes. Active change management requires a personal “touch” for the creation of teams and the deployment of process owners. • Motivation and innovation. As the senior real-time worker, the CIO of the future depends on the whole organization – from production and marketing and sales to management – to set a real-time mood and to incur a passion for permanent optimization. The new profile of the CPO – whether he is called that or not – also demands changes in training. In future, the fact of whether he can develop elaborate concepts for network management, server consolidation or information engineering will become less important for the employment of an IT manager. Decisive will be his economic appreciation and the ability to recognize and exploit the innovation potential of IT for business processes.

342

Ferri Abolhassan

Five steps to the real-time enterprise

1. Design of real-time processes

Current business processes are optimized from the point of view of the customer and time perspective. Software tools such as ARIS from IDS Scheer support the design of processes. 2. Implementation of a real-time platform

Efficient IT systems must enable direct recording, forwarding and interaction with customers and partners. 3. Adaptation of software to real-time processes

Application systems for ERP (Enterprise Resources Planning), CRM (Customer Relationship Management), SCM (Supply Chain management), E-Procurement, reporting, etc. are optimized with a view to realtime information. 4. Control of real-time processes

Up-to-date measurement of business processes, forwarding of information, comparison with default values and benchmarks. Software tools such as the Process Performance Manager (PPS) from IDS Scheer can be used for this purpose. 5. Change management within the company

Permanent change to an agile, innovative and flexible company, continuous training of employees.

Rob House

How to be switched on – without being switched off

It is not enough simply to have fast and constant access to the data and the people that you need in order to run your life. It is also essential to know what is socially acceptable when using technology – that is, if you are to avoid the dangers of being treated like an outcast. Human relationships are still vital to success – and people who forget this truth run the risks of hostility, exclusion and discrimination. As our personal and professional lives become increasingly dependant upon a variety of devices, it is critical to understand and observe the protocols of etiquette for using technology today. We are a mobile society: business professionals and technicians tote wireless phones, mobiles have become fashion statements for teenagers, phones ring in meetings and loud conversations are conducted on public transport. The use of data is different – less intrusive – but the ease with which emails can be sent to PCs and PDAs has introduced another issue: information overload. We now receive 64 times more information than we did than 25 years ago and growth continues: in fact, the curve is exponential. As a result, people may give the pretence of taking notes when, in reality, they are working their way through a mountain of messages. The devices that empower both the business and personal side of our lives have led to a number of changes in the way we communicate in the workplace and on the move. But people are now starting to question what is appropriate and what is not – particularly in business situations. Research recently undertaken by the University of Surrey sought to determine attitudes and opinions with regard to acceptable communications practices in the digital age. The researchers’ report underlined the need to establish workplace guidelines, with management taking the lead in setting standards. Commissioned by Siemens Communications in the UK, the research looked into the way in which modern business communications are affecting workers’ attitudes, performance and interaction.

344

Rob House

Described as a study into the etiquette of business communications in the digital age, the research clearly demonstrates that an over-reliance on communications is becoming a friction point in offices, causing stress which is affecting personal relationships both at work and at home. Undertaken by Surrey Social and Market Research (SSMR) at the University of Surrey, Guildford, England – with additional analysis by the University’s Digital World Research Centre – the research set out to: • explore the experience of people in workplace settings with regard to social interaction • examine views of usage of IT and communications equipment while interacting with others • investigate perceptions of what is seen as acceptable behaviour in the workplace • study perceptions of changes in levels of acceptability over time • highlight any differences in terms of gender and age • assess the findings and provide guidance on appropriate behaviour Both quantitative and qualitative research was used to assess and study opinions. The project began with a series of qualitative discussions that were used to tease out opinions and allow participants to develop their ideas. These group discussions were then used to shape the focus and progression of the questions used in survey of more than 500 business users. As most people have experienced, the research reported on the ‘annoying’ use of mobile devices. More interestingly, it also identified an underlying demand for the better management of availability and for integrated communications systems. Ironically, it highlighted the fact that many office workers resent the interruptions that communications caused to meetings and workflow, but at the same time demanded almost instant contact when trying to reach colleagues. The report found that this conflict is straining relationships – thereby affecting employee performance and output and requiring management attention. The research showed that the demand for instant, almost constant, communication is adding to workplace Stress, can cause Anger among colleagues and strangers alike, and is proving to be a Distraction for all, both in meetings and in public places. In short, office workers are becoming increasingly SAD (Stressed, Angry and Distracted). The research sample was drawn from a cross section of business people, almost all of whom used computers and mobile phones, both at work and at home. The majority felt that the use of new technology was increasing –

How to be switched on – without being switched off

345

particularly the Internet and email. Mobile phone usage was common at work and seen as a cause of some concern, with only 11% of those surveyed believing it was acceptable for these devices to be on during a meeting. Usage in public places was accepted as long as the way conversations were conducted did not cause offence. Email was seen as the foundation of modern office life – being used by more than 90% of business users – and is used as a quick and easy method of transferring information between colleagues and companies. However, there was a perception that email and mobile phones have reduced people’s patience at work and could lead to stress and anger in certain circumstances. Respondents recognised that we live in an age where we expect to be able to contact colleagues, clients and suppliers almost instantly. There was spontaneous recognition that some kind of etiquette is required – otherwise responding to a call when talking to another person implies that the phone call is more important than that person. Similarly, answering a call in a meeting implies the meeting does not deserve your full, undivided, attention. Interestingly, only slightly more than 50% of the respondents felt that it was inappropriate to use any form of IT equipment in a meeting or when talking to another person at work. Text messaging was generally felt to be too casual for a business context, except by prior arrangement. Subject matter, location and relationships were all factors in determining how someone behaves in a meeting – with relationships being perhaps the most critical. Meetings now cover a wide range of discussions and many are informal and relaxed – on these occasions, interruptions are more acceptable if they are sufficiently important. The report concluded that while we are now keen to be in easy contact with our friends, colleagues and loved ones – and gain a feeling of comfort from modern communications – there was an underlying trend of resentment with regard to communications interruptions and distractions in the workplace. Unsurprisingly, failures of communication are significant factors influencing stress and anger – in the workplace and, indeed, at home.

Research highlights Many of the remarks made by the business people who took part in the qualitative study provide valuable food for thought and make interesting reading. The following statements are typical and indicative of stress, anger and distraction among office workers:

346

Rob House

• “As a society we want everything now and we’re not prepared to wait” • “Years ago, before mobiles, if I went out and my daughter got ill my wife would deal with it and just tell me what happened – now if I’m not contactable, all hell breaks loose when I get home.” • “If you leave your phone on (during a meeting) you’re saying you’re not worthy of my full attention” • “It winds me up immensely… irrespective of the conversation” • “I spent four hours on the phone on the train…. the people beside me were quite rightly irritated” • “If you’re in a meeting… and someone’s phone goes off, it’s frowned upon. It just shouldn’t happen.” • “I find it annoying that you can be ringing someone, you know they are there but they are not picking up the call” • “People just sit there and let their phones ring” • “There’s one guy… it’s constantly an interruption… it’s becoming a disciplinary issue… it’s just not acceptable” • “In the workplace email has increased stress levels” The quantitative study also supported the SAD findings: • Only 11% felt it appropriate to break off and use a mobile phone when talking to someone else at work • A massive 88% did not think it appropriate to use an electronic device for note-taking in meetings • When meeting in someone’s office, three quarters of respondents felt it unacceptable for that person to read email or take a phone call • More than 80% felt it inappropriate to look at or send text messages (during a meeting) • However, 53% of respondents felt it appropriate in some circumstances to receive a text message in a meeting • Around 90% felt that the only acceptable reason to receive a mobile phone call during a meeting would be in an emergency

Why are we so SAD? Technology has given us myriad ways of communicating. We have desktop and mobile phones as well as PDAs and notebook PCs that can be used as softphones. We use the Internet, intranets and extranets, as well as public and private networks, both wireline and wireless. It is hard to imagine a communications landscape more fragmented than the one we currently

How to be switched on – without being switched off

347

endure. This is a chaotic way of working, one that only makes sense for historic reasons. The core problem could hardly be simpler. We spend more time messaging than communicating because we cannot reach colleagues the first time around. ‘John’ calls ‘Jane’ and gets her voicemail because she is on the phone. Jane calls back and gets John’s voicemail. Telephone tag wastes time and money – Siemens estimates it amounts to some 30 minutes each day for each knowledge worker, which equates to a UK national wage bill of some £22bn per annum. Furthermore, this scenario is immensely irritating and frustrating. If John’s call is urgent he may ring Jane’s mobile, or send an instant message (IM) or email but he still does not know the best way to get through quickly or when Jane will be able to call him back. Now imagine that John’s urgent issue requires a quick conference call with several parties. No wonder he is SAD. That process could take all day – and fail. The University’s research supports Siemens own findings concerning availability management, a key feature in OpenScape – the company’s recently developed application that enhances the functionality of Microsoft’s Exchange Server. OpenScape allows users to flag to their colleagues their degree of availability and their preferred method. It integrates voice, email, mobile, voicemail and text messaging systems to maximise contactability while minimising intrusion.

There has to be a better way Surely, it is time that technology went back to the future and gave us the equivalent of everybody being in the same location at the same time and being able to communicate in real time? A complete makeover, however, is clearly impossible. So how can we enable better, less stressful ways of communicating and collaborating? The answer starts with voice-data convergence and instant messaging represents the signpost. IM is an application that uses icons to show the ‘presence’ —- on-line or off-line — of nominated colleagues (buddies in Internet parlance). Note that IM is actually a misnomer since it’s used to communicate in real-time; email is a messaging medium. IP phones are – in effect – data devices that are visible to the network. What Siemens has achieved is to develop and integrate with IM applications that use icons to indicate presence and availability. This means that you do not call parties who are busy and you spend more time communicating and less time messaging.

348

Rob House

IM programs also allow users to display additional availability information alongside the presence icon. Availability denotes a person’s willingness to communicate and it is based on preferences and policies, i.e. it is managed at both the individual and corporate levels. In presence-aware telephony availability is managed via personal profiles having more functionality than that of popular IM applications. Calls and messages are received or rejected on the basis of: who is allowed to reach the called party; what device is defined as being the most convenient at that time; and when the party would like to be reached. Managed availability therefore solves the ‘who, what, when and where’ of communications. Presence and availability management tools put users firmly in control of their communication devices. People can even leave phones on during an important meeting confident in the knowledge that it will only ring in a genuine emergency. Why? Because users define their own rules that help them decrease interruption while simultaneously increasing their availability. Presence and management are powerful communications parameters that minimise telephone tag, thereby boosting personal productivity, yet at the same time reducing stress. They also allow one colleague to talk to another immediately when there is a need to react to an event or an urgent issue. The presence and availability status of nominated colleagues are displayed in a communications portal. Colleagues can also be grouped into workgroups. Adding content to a communications session is easy; users simply drag documents into a new window. Imagine that a manager is responsible for several projects. If an issue arises he/she can select the relevant portal, which displays the names of all members, their presence and availability status, and the file names of the relevant documents. The issue can then be addressed immediately by clicking on the relevant names to set up a conference call in which all the participants have immediate access to the same set of documents. In this way, real-time communications means goodbye to the SAD office worker.

VIII. The steps in evolving into an “E-Enterprise”

The development of a company into a ‘Real Time Enterprise’ is an evolutionary process. Not only does the company’s own communication infrastructure have to be included in this development, but also those of authorized third parties and service providers. This is enabled via open network and system architectures, which in turn allow the step-by-step introduction of open application platforms and process supporting tools.

Steve Blood

The RTE: it starts with early warnings

RTE models tend to focus on the need to make rapid responses to events and issues and overlook the fact that early warnings need to be an integral part of the process. Indications that something is amiss should be flagged in real time. Gartner’s initial definition of a real-time enterprise was: “an enterprise that competes by using up-to-date information to progressively remove delays to the management and execution of its critical business processes.” However, in light of several important findings that were uncovered during a recent RTE research project, we concluded that an update was needed. The revised definition is: “The RTE monitors, captures and analyzes rootcause and overt events that are critical to its success the instant those events occur, to identify new opportunities, avoid mishaps and minimize delays in core business processes. The RTE will then exploit that information to progressively remove delays in the management and execution of its critical business processes.” The new definition is based on the contention that there is always prior warning before every major favorable or unfavorable business event. Let’s explore this hypothesis by first investigating non-business surprises such as the weather, e.g. hurricanes, tornadoes, earthquakes, volcanoes, floods and tsunamis. It’s clear that today's ability to issue early weather warnings is largely based on capturing the identifiable and measurable root-cause events such as falling barometric pressure. After determining the consistent presence of vital information prior to natural and man-made surprises we move on to business surprises. In this recent research project we analyzed bankruptcies, positive corporate earnings results, negative corporate earnings results, sudden market share losses, and other types of business events that surprised executives, shareholders and government regulators. As in the case of natural and man-made surprises, we concluded that business surprises also arrive after an ample supply of warning signs. We then reasoned that if these warning signs had been detected and acted on prior to the surprise events, the harmful effects of all

352

Steve Blood

categories described here could have been reduced or avoided. The take-home message is significant and somewhat disturbing. The effort undertaken by enterprises, and especially IT vendors in pursuit of an RTE capability, is profoundly incomplete. Almost all of these efforts pertain to responding faster to events by removing delays within and between each business process, while remaining silent on the early warning detection of vital information in real time. In most cases the automatic generic of an event — the indication that a problem has arisen — will come from the process. For example, SAP’s NetWeaver platform has a business process manager that performs that task. The generation of warnings can also be introduced into business processes by system integrators or in-house IT staff. Once generated, the event/issue has to reach the relevant parties, and that is where the delay starts and escalates unless a second, formal process resolves the shortfall. The ability to locate the right people in real time and supply the information needed to make fast, informed decisions is a key feature of presenceaware software and its relevance to the RTE has been covered in other articles. (Editor: “Real Value in Real Time” and “Delivering the Promise. Making it Pervasive”). A key message is the fact that while technology can minimize the delay between detecting, reporting and responding to an event, the ability to act on information in real time involves people. How many business processes can replicate the Amazon.com model, i.e. implement end-to-end automation? Products are marketed via email and ordered via a linked, secure Web site that lets customers make on-line checks about status right the way through to actual delivery. This is a seamless process that gives Amazon the lowest possible transaction cost and thereby a significant competitive advantage but what if there is an error in the process that needs intervention by a human? How happy is the customer now with the order to delivery process? Many companies have automated ordering procedures, but delays and other issues are reactively resolved (hopefully) via contact centers. What is missing is the ability to pro-actively alert the relevant stakeholders to the process breakdown in order that it may be resolved in a timely manner. Seamless flows of information are the foundation on which the RTE is built. They are clearly required for sales and manufacturing processes, but ‘events’ that require human intervention have yet to be integrated into the workflow. For example, in the business-to-business domain, a good customer may place an order but an overdue account payment flags a manual intervention before the ordering process can be completed, potentially putting a valuable business relationship at risk. However, if the account infor-

The RTE: it starts with early warnings

353

mation could be automatically sent to the right party in the company then authorization to ship the order may be given through one of a number of channels – audio using speech recognition, instant message or email. Adding people to the decision process efficiently enables more customerfriendly responses , e.g. there may be a valid reason for late payment. That represents a value judgment that computer systems cannot make. So far so good, but the synchronization of an issue with the people needed to resolve it is a dynamic process. The relevant parties may change their position or leave the company. Thus, names cannot be an integral part of event management software; instead they should be linked to a skills-based directory. Similarly, events should not be linked to a specific medium, e.g. SMS. This is a good choice since messages sent this way are non-intrusive, but mobiles will normally be switched off in important meetings. This is where managed availability enters the equation. Presence-aware software in the IP Communications space allows users to define who can reach them, when they can be reached, and on what medium. They can also nominate one or more ‘replacement’ colleagues in the event that they are not available, e.g. in yet another meeting, with a client, or on vacation. However, these benefits are generic, i.e. they are not restricted to knowledge workers and specialists. This indicates the need to extend the functionality throughout the enterprise and to make presence and managed availability pervasive so that it can be employed by anybody, at any time and be applied to any business process. When we start to pull these developments together it becomes clear that the event ‘push’ functionality of software such as NetWeaver needs to be integrated with applications such as HiPath OpenScape. This link has been made by Siemens to SAP platforms as well as those of IBM (WebSphere) and Microsoft (LCS). Similarly, Alcatel has integrated with WebSphere and Avaya with Siebel’s platforms. Thus, players in the real-time communications space have recognized the importance of the RTE. In turn this means that as more business processes become communications enabled, we can start to consider the possibility of entire business activities becoming real time, within and beyond the enterprise. Activities like cash management, pricing, service delivery, new product design, and even mergers and acquisitions can be moved toward real-time performance. “And beyond” was italicized in order to indicate the importance of realtime communications within ecosystems, particularly those involving offshore manufacturing. When an event occurs within a process that involves suppliers and other third parties, the technology that locates and informs

354

Steve Blood

the relevant individuals has to extend beyond the enterprise. Security is an obvious concern and for the RTE model to work effectively presence, availability and the skills-based directory need to be networked to all authorized parties. This represents a significant opportunity role for service providers. They are ideally placed to extend the functionality of the ecosystem ‘owner’ to those parties via a hosted service — ideally one that is Web-enabled, i.e. no client software is required. The benefits that accrue from the management of events have to exceed the investment, i.e. there has to be an acceptable ROI. That is clear, however, less obvious is the fact that many companies do not recognize the RTE opportunity. The figure is hard to quantify but take-up is low, in the order of 10 to 20%. There are various reasons why the figure is not higher, one being misunderstanding about the concept, another failing to recognize ROI opportunities that are less than obvious. Research conducted by Garner has indicated that many IT managers blur the distinction between “real time” and “online.” A clear understanding of what real time means is an essential foundation for real-time enterprise thinking. And if a third of all IT managers are unclear then the RTE potential will be lost. One area of misunderstanding is the terminology. “Real time means “real-world time” and not “sub-second” or “instantaneous”. They are good ways of describing what it means for a system to be online, but they do not mean real time. Thus, diverting attention onto a few low-latency systems misses the point. IT must enable cross-enterprise processes to detect and respond to critical business events at the rate they occur. This is the definition that Garner employs: A real-time system is one that executes its input, computation and output sequences fast enough to monitor and respond to some real-world events at the rate they occur. The key point is that the system designer does not have the luxury of processing at a rate that minimizes system costs or maximizes some internal efficiency. The system must respond within a time limit that comes from the real world outside the system — where events will move on at a certain rate. If the system does not respond fast enough, it cannot be used to control a real-world process. Real time cannot be defined in terms of a single threshold value, such as a response in less than 30 milliseconds or even 30 seconds; it is relative and context specific. At one extreme, the actuator of an automobile airbag must detect rapid deceleration, decide it is significant and activate the inflation device before the chest of the driver hits the steering column. In this context, real time

The RTE: it starts with early warnings

355

means that the entire detect-to-deploy cycle must completed in about 30 milliseconds (it takes you about 100 milliseconds to blink). At the other extreme, a system designed to monitor the rate of movement of a glacier (perhaps to warn of floods) will not need to report more often than daily. When real time is misconstrued to mean “instantaneous” or “sub-second,” the power of the RTE concept is diminished in two ways. First, we start to think in terms of specific, discrete systems, which usually support only part of an end-to-end process chain, and second, we start to think in terms of online response times of user interfaces. The problem of misinterpretation is a subtle one, because both of these aspects are important to progress in the RTE. But improvements to these things alone will not bring about competitive advantage. In some cases of real-time misapplication, we see millions of dollars being wasted on incremental improvements to systems that will never achieve payback. Investments predicated on an “e” suffix to indicate that existing systems and processes had gone online often amount to no more than a change to the user interface and the infrastructure of middle-tier application integration. The same business processes, operating on the same foundation systems, only allow the corporation to detect and respond to business events at the same pace as before. User access via a Web browser might be cheaper and more convenient, but these bring relatively small efficiency benefits to the company, and not business performance breakthroughs. Almost without noticing it, business cases and project portfolios, which have been built on an incorrect understanding of the RTE, will become biased toward a focus on IT systems efficiencies, rather than business efficiencies. Real time simply means real-world time, which makes contrived terms like “right time” redundant. For the RTE, the real world is the external business environment, with its shrinking windows of opportunity and increasing turbulence and uncertainty. This sets the tempo required for the RTE to compete. To keep pace, the cycle times of business processes and the systems that support them must be continually improved. Anecdote

An industry colleague was sitting for one hour on a plane waiting for it to take off. The pilot came on the air, apologized, and made a refreshing honest statement: “Sorry he was late – He didn’t know he was going to fly that particular plane.” The operational cost of this type of delay to air-

356

Steve Blood

lines is typically around $15K/hour and on top the company had antagonized a few hundred passengers. That type of scenario had certainly happened before and it will happen again to different airlines in different airports around the world. But introduce some RTE functionality and the pilot’s presence in the airport could be detected and flagged. Failure to arrive at a defined point within a certain time could result in an alert and the event would escalate if a positive response were not received. The system would then check on the availability of other pilots, they would be contacted, and the plane would take off on time. It’s not rocket science. In the UK the drivers and guards of Virgin’s trains check in and are authenticated using DTMF tones on a telephony platform. If this does not occur the system is alerted and an escalation process starts.

An airline that recognized the true cost of delays and then went on to do something about it could leverage the new RTE functionality. For example, a pilot might be alerted to a potential problem with the landing gear, although not one that would impact on the plane’s ability to touch down safely. If the pilot forwards this information a trouble ticket can be generated indicating the problem and the estimated time of arrival. The necessary maintenance staff can then be informed, an inventory check made on spare parts, and everything would be ready when it arrives at the gate. The plane could then be turned round on time, or with a minimal delay, thereby saving another $15K/hour and preventing a few hundred passengers venting their anger and frustration on front-desk staff. Ideally the potential problem would be diagnosed in flight and data sent ahead automatically. The process would then start via machine-to-machine and machine-to-man communications and there would be no latency, i.e. no need to wait for somebody to do something. However this can further benefit man-man processes A maintenance engineer arriving on site to deal with the landing gear alert may require further support from third line engineering. Technical support personnel are rarely office based so using instant messaging the engineer can locate a specialist that would be able to assist in resolving the fault – the system needs to know their presence and availability, i.e. free to take on this task or not. Contact could be established using a secure Wireless LAN infrastructure. The specialist could start off by chatting, then launch a voice call by

The RTE: it starts with early warnings

357

clicking on the telephony icon; white boarding and video could also be introduced into this hypothetical communications and collaboration. The scenario for fixing landing gear is the same as that of a contact center when the agent is asked a non-standard question. It represents a generic requirement. And to return to the tarmac, if the best-qualified technician is not available (physically), he/she could communicate with the colleague using streaming video and talk them through the part replacement process. Yes, you can address this type of issue on a mobile phone, but the use of a central, unified communications platform allows the entire event to be recorded. Thus, if there is a related incident in future, then there is a historical record that can be accessed. If our hypothetical plane did need to make a subsequent crash landing then this record would be a de facto black box. The more one examines the potential of the communications functionality that underpins much of the RTE concept the more one can see opportunities. A wireless operator in New York sends trouble tickets to its engineers using IM, but if they are not on-line the system wraps the information up using VoiceXML and plays it to the cellular phone. Thus, the system ensures that the trouble ticket is received and it does so using the appropriate medium at that time. Conclusion

Unlike earlier computer telephony concepts such as Unified Messaging (UM), the RTE is predicated on the need to act in real time, not the need to save time. Saving time is a very vague concept, one that does not resonate with management. Maybe UM would have saved 15 minutes a day, but the figure could not be quantified. And there are myriad ways of ‘saving time’. The take-home message is that RTE does offer a quantifiable financial saving and time, in the real-time sense, really is money. Up to $15K an hour or may be more. The cost of delay through poor communications is not well understood in the majority of companies.

Wolfgang König, Tim Weitzel

Towards the E-Enterprise: standards, networks and co-operation strategies

Modern networked types of enterprise range from Supply Webs and flexible company alliances to fully decentralized and highly modular service architectures. They all require a well coordinated network design between a large number of actors for the systematic exploitation of network effects to eventually provide the basis for ad-hoc supply chains or “real-time enterprises”. But there are network coordination problems that are characteristic of cooperation networks that constitute a considerable challenge for cooperation strategies and IT governance. This article systematically discloses the technical and cultural challenges a “real-time enterprise” is facing. An economic analysis of equilibria in cooperation networks reveals the key coordination requirements. Computer-based simulations show promising results for the resolution of “classic” network problems such as start-up problems in particular. At the same time, the value of modern co-operation strategies like especially ‘co-opetition’, i.e. co-operation with competitors, becomes clear.

Introduction “If the dominant symbol of the industrial economy is a factory, then the emblem of the modern economy is a network.” [Maub+00] The 20th century was characterized by an increasing importance of demand-sided economies of scale that are often described as positive network effects. The significance of network effects results from the need for compatibility. As a consequence, network effects play a key role for E-Business and are primarily discussed in the literature on standards [Weit04]. According to a general definition, Electronic Business describes business transactions where participants interact electronically. In contrast to Electronic Commerce solely focusing on the relations to the eventual consumer

360

Wolfgang König, Tim Weitzel

(Business-to-Consumer, B2C), E-Business comprises the entire value creation and therefore pinpoints digital and automated business processes within and between firms [Weit+01] [Schi00]. A basic challenge for any sound E-Business strategy is the development of advantageous networks of partners and systems for realizing the many network effect potentials. Until the middle of the last century, driven by the First World War, the main focus of most standardization activities was improving national productivity with standardization as a means of reducing the heterogeneity of products and processes in order to obtain production side economies of scale. The focus later changed in towards compatibility standards. Brian Arthur called this the change from “Alfred Marshall's World” to an “Increasing Returns World” [Arth96]. Because standards enable compatibility and are thus the basis for the utilization of many synergies [Gabe87; Gabe91; FaSa87; Nigg94; Klei93], it is said that standards constitute networks of users [BeFa94; DaGr90]. Prominent examples of the standardization literature include railway track width, screw diameter, nuclear power plants, keyboard design, VCR systems, web browsers, SSO software, EDI networks and office software [WeWe02; Weit04]. In the railway example, a network effect arises through the standardization of track width as railway companies are then able to exchange wagons and boxcars [Kind83, 384–385]. The screw example describes early standardization efforts to reap compatibility benefits between the U.S. and the U.K. harmonizing the pitch of the screw thread between them during World War II to make screws exchangeable [Kind83, 392]. Network effects can also be found between electrical devices and voltages, automobiles and gas, video recorders and cassettes [Plum37, 21; WaWa83, 125; Heme75, 3; GeHö95, 21–29; Bund98]. While other prominent examples such as EDI networks for the automated exchange of business data and directory services mainly focus on direct information exchange, the examples including nuclear power plants (light water vs. heavy water vs. gas-graphite reactors) [Cowa90, 559–560] rather rely on learning curve effects of technology development. Habermeier [Habe89; Habe94] differentiates between effects due to network participation (“joining a well established network”), informational economies of scale (e.g. brand awareness), complementary effects (video recorder and cassette) and dynamic learning effects (the more a technology is used, the more effectively it is optimized and the more perceptible are its advantages). Goal of this article is to disclose challenges that are characteristic of cooperation networks and to evaluate ways how to cope with them. Eventually, a firm’s technical and cultural capability to develop networks is a

Towards the E-Enterprise

361

necessary step towards a “Real-Time Enterprise” as the network capability enables a firm to “participate in networks that allow them to share databases, have access to large selections of compatible software, exchange documents (...) or simply communicate directly” [BeFa94, 117]. While the network metaphor is widely used in theory as well as in practice to describe the most diverse interaction structures between actors, there are complex, network-specific coordination problems that result from the existence of network effects and that require new coordination concepts. The subject matter of this article is therefore an economic analysis of the existence and efficiency of equilibria in cooperation networks with different institutional coordination modes (centralized versus decentralized control). Based on a critical review of the literature on networks and standards, in chapter 3 a research framework for cooperation networks is developed that can serve as the basis for an IT governance for E-Enterprises. Building on this, a computer-based simulation in chapter 4 shows promising results with regard to the general solvability of “classic” network problems such as the start-up problem in particular. A key result is that regularly many individual network agents are better off with centralized coordination and that fundamental coordination problems in the E-Business domain can be addressed by quite simple approaches such as in particular ‘co-opetition’, i.e. cooperation between competitors.

The literature on network effects

Network effects as externality

While it is common in many markets that the buying decision of one consumer influences the decisions of others (e.g. bandwagon, snob, and Veblen effect are broadly discussed in economic literature [e.g. Leib50, CeKa82]) there are markets subject to network effects or demand-side economies of scale. This implies a positive correlation between the number of users of a standard (‘user network’) and its utility [KaSh85] [FaSa85]). Renowned examples are information and communication technology markets and the telephone. The utility derived from a phone depends on the number of persons you can reach with it. Parallel with the growth of the telecommunication and information technology markets in recent years the discussion has gained considerable momentum. The externality property implies coordination problems in markets sub-

362

Wolfgang König, Tim Weitzel

ject to network effects. In other words, network effects are said to possibly constitute market failure. Katz/Shapiro differentiate between direct network effects in terms of direct “physical effects” [KaSh85, 424] of being able to exchange information (as in the case of telephones) and indirect network effects arising from interdependencies in the consumption of complementary goods [BrWh85; ChSh90; ChGa92; Teec87]. Examples are computer operating systems and available application software, or video cassette recorder systems and the format of the tapes. Kindleberger (1983) describes free-rider problems due to the public good properties of standards. Arthur (1983; 1989) shows that technologies subject to increasing returns (in contrast to constant and decreasing returns) exhibit multiple equilibria and will finally lock-in to a monopoly with one standard cornering the entire market. Since this standardization process is non-ergodic (or path dependent) the ultimate outcome is not predictable. A frequently used example is the battle between VHS and beta [Weit04]. Analogously, Besen/Farrell (1994) show that tippiness is a typical characteristic found in networks describing the fact that multiple incompatible technologies are rarely able to coexist and that the switch to a single, leading standard can occur suddenly. An example of this is Microsoft’s Internet Explorer battling out Netscape navigator. The pattern of argument for standardization processes is always the same: the discrepancy between private and collective gains in networks under increasing returns leads to possibly Pareto-inferior results (market failure in the sense that too many/too few/the wrong technologies are adopted). With incomplete information about other actors’ preferences, excess inertia can occur, as no actor is willing to bear the disproportionate risk of being the first adopter of a standard and then becoming stranded in a small network if all others eventually decide in favor of another technology. This start-up problem prevents any adoption at all of the particular technology, even if it is preferred by everyone. In corporate reality this often shows as aggressive awaiting, an often-witnessed strategy of agents trying to avoid the risk of being the first – and possibly only – adopter of a new technology that then does not offer enough benefits to compensate for the costs by sitting out software updates, for example [Weit04]. On the other hand, excess momentum can occur, e.g. if a sponsoring firm uses low prices during early periods of diffusion to attract a critical mass of adopters [FaSa86]. In the case of complete information on the part of all actors concerning their symmetric preferences for a certain standard, a bandwagon process will overcome the coordination problem, with actors who stand to gain relatively high stand-alone utility or private benefits from adoption (as

Towards the E-Enterprise

363

compared to network effects) starting the adoption process. Nevertheless, in the case of heterogeneous preferences, Farrell/Saloner (1986) show that due to strategic behavior even perfect communication might not be able to overcome excess inertia or momentum. In the case of sponsored technologies the situation is somewhat different. Here there is a possibility of internalizing the network gains, which would otherwise be more or less lost, by strategic intertemporal pricing, for example [KaSh86]. There are private incentives to providing networks that can overcome inertia problems; however, they do not guarantee social optimality per se. Various perspectives can be distinguished in the literature [Klein98; Yang97]. The authors using empirical approaches mainly try to prove the existence of network effects and estimate their values by using regression analysis to estimate the hedonic price function of network effect goods [HaTe90; Gand94; EcHi95; Gröh99]. Theoretical approaches mostly use equilibrium analysis to explain phenomena such as the start-up problem [Rohl74; OrSm81; KaSh85; 1994; Wies90; BeFa94; EcHi95], market failure [FaSa85; FaSa86; KaSh1986; KaSh92; KaSh94; Gröh99], instability (also called “tippiness”) of network effect markets [Arth89; Arth96; BeFa94; FaSa85; KaSh94; ShVa98], and path dependency [Davi85; Arth89; BeFa94; KaSh94; LiMa95b]. Summary of the main findings of network effect theory

An extensive overview of network effect theory together with related literature e.g. from actor network theory, infrastructure theory, and vendor strategies is provided by [Weit04], a rich online bibliography can be found in [Econ00; Agre98]. Common results found in the literature on network effects are the following: • Network effects imply multiple equilibria and the market will finally lock-in to a monopoly with one standard gaining total market share. In many cases, market solutions are Pareto-inferior [Arth89; KaSh86; Thum95]. • Instability is a typical property describing the fact that multiple, incompatible technologies are rarely able to coexist and that the switch to a single, leading standard can occur suddenly (“tippy networks”) [BeFa94, 118; ShVa98, 176]. • The start-up problem prevents the adoption even of superior products; excess inertia can occur as no actor is willing to bear the disproportionate risk of being the first adopter of a standard [FaSa85; FaSa86].

364

Wolfgang König, Tim Weitzel

• On the other hand, excess momentum can occur, e.g. if a sponsoring firm uses low prices in early periods of diffusion to attract a critical mass of adopters [FaSa86]. Reconsidering network effect theory

An elaborate criticism on network effect theory is provided in [LiMa94; Weit+00; Weit04]. The main drawbacks are: • No sufficient distinction between direct and indirect network effects is made in the models although it can be shown empirically and analytically [KaSh94] that they have different economic implications. • Generally, in a world with incomplete information and rational agents (which is a common assumption in traditional models) the notion of redeemable market failure seems problematic. Problems of finding unambiguous empirical evidence for market failure due to the existence of network effects make this fundamental discussion especially interesting [LiMa90]. • The frequent proposition of indefinitely increasing and homogeneous positive network effects at the same time neglecting costs of network size is problematic. If network effects were exhaustible, multiple networks could efficiently coexist [LiMa95]. • Traditional literature lacks a differentiation between centralized an decentralized networks and particularly offers few decentralized solution proposals, making inefficiencies appear more dramatic, for no cure seems available. At the same time, no decision support for corporate standardization problem can be derived [Weit+00; West03]. Besides their limited descriptive power, these problem prevent an application of most traditional models to realistic network problems. Also, they cannot account for the individuality of network agents deciding on the design of Electronic Markets, the development and coordination of supply webs or intranets. In this context, interdisciplinary approaches incorporating important findings from sociology, geography or medicine into network research proved to be very promising for developing a unified view on networks and for increasing the explanatory and managerial power of network effect theory [WeKö03; Weit04].

Towards the E-Enterprise

365

A research framework for coordination networks Our main hypothesis is that an asymmetry between individual and collective network participation incentives disturbs the self regulating efficient allocation one would expect from traditional neoclassical economics. From a theoretical perspective, this is not surprising since the externality property associated with network effects disturbs the automatic transmission from local to global efficiency implying welfare problems. In order to identify a sound theoretical foundation for developing a methodical view on the strategic situation of network agents, in this section important concepts from network effect theory (network effects and their implications) and from economics (externalities, existence and efficiency of equilibria, neoclassical welfare mechanism) are applied. We will call systems of business partners who are tied together by mutual dependencies like especially network effects and who therefore strive to reap their common synergies cooperation networks. The terms “network” and “standard” are used in a vast variety of semantics. In this work, we use the term standard to refer to any technology or product (software, hardware) incorporating technological specifications that provide for compatibility and thereby establish networks of users. In this section, we will first systematize network coordination problems (3.1), discuss them using Straight Through Processing for securities transactions as an example (3.2) and then develop equilibrium and efficiency concepts for an economic network analysis. Coordination problems in business networks

Besides incomplete information about others' behaviour, the source of the coordination problem associated with network participation decisions is an asymmetry between the costs and benefits resulting from it. This constellation (together with the frequent underestimation of potential savings) is often regarded as being responsible for a very (s)low propensity to participate in networks (excess inertia, start-up problem, see aggressive awaiting in section 2.1). Depending on the institutional settings of the deciding agents, network participation decisions can, principally, be centrally coordinated or they can be decentralized as, for example, in sovereign business units with autonomous IT budgets (Weit04). While an idealistic optimal solution for a network (“centralized coordination”) might theoretically be deducted, the networks found in corporate reality are often less efficient (“decentralized coordination”).

366

Wolfgang König, Tim Weitzel

From a theoretical perspective, in contrast to neoclassical economic analysis, network effects imply multiple equilibria [Arth89]. As a result, network behaviour is difficult to predict, and planning, controlling and operating networks is a complex challenge. In practice, when trying to reap network benefits it is often difficult to implement a favourable network as a set of agents (individuals, business units) participating in the network of, for example, users of a certain EDI standard or on an electronic market place. The reason is that network participation is associated with certain costs and benefits. While benefits comprise cost savings due to a deeper integration with business partners and reduced friction costs, adapting to the network can induce costs of technology adoption to achieve compatibility (i.e. standardization). From an individual perspective, network participation is advantageous when benefits exceed costs. But quite often, advantageous solutions do not evolve amongst agents deciding decentrally about their network participation. Thus, agents are stuck in sub optima and substantial possible network gains remain unexploited. Straight Through Processing in securities business

A recent example of coordination challenges in cooperation networks is straight through processing (STP) for cross-border end-to-end integration in securities business, comparable to traditional electronic data interchange networks e.g. in the automotive industry (Fric+02). A lack of communications standards in particular as well as incompatibilities between the multitude of different internal and external clearing and settlement systems together with nationally diverse regulatory environment are substantial obstacles to swift, smooth and fully automated cross-border securities transactions [Swif00, 28]. By automating system-to-system communication and the associated avoidance of media discontinuities, network partners hope to accelerate processes, reduce error rates and decrease process costs [Emme93, 17-29]. There are many analogous integration problems discussed throughout the literature. Yet, in the banking industry they appear to be especially severe as the industry is characterized by: • very heterogeneous IT landscapes resulting from nationally diverse industry specific regulations and a high frequency of mergers, • many systems (30-40) supporting several products within a single bank [Ruh+01, 78], and • below average industrialization degree as indicated by a lack of standard

Towards the E-Enterprise

367

software and high proportions of proprietary software development (60%) [Shah02].

media discontinuities within and between stages status quo

smooth information flow allows fully automated process

Legend = media discontinuities

Figure 79: Concept of straight-through processing

As a consequence, frequent media discontinuities cause 15–20% of all transnational securities transactions to be faulty. The error rate of transactions with emerging markets is even around 33% [Gilk01]. According to SWIFT, the portion of costs for the correction of errors alone amounts to a third of the entire processing costs or €1 billion each year [Brow01]. Reuters estimates the savings potential through efficient processes to be $ 12 billion annually [Reut01]. But as the history of EDI shows, designing cooperation networks is a complex challenge. The notorious start-up problems are particularly apparent in the STP context: Because the anticipated (internal and external) STP utility requires high (internal) STP rates from all partners, there is an incentive to wait for other partners to invest into internal and external interfaces and system integration first and hence to avoid the risk of early – and possibly solitary and thus unprofitable – investment. Besides this start-up problem [KaSh85; Rohl74; OrSm81], there is a second phenomenon to be taken into consideration. Since real STP requires the participation of all process participants and can therefore frequently affect dozens of independent partners in different regions of the world [Emme93, 6–8], it is possible that while for a specific actor, i.e. for an individual, participation might not pay off, from the perspective of the entire value chain, or of a majority of part-

368

Wolfgang König, Tim Weitzel

ners, this actor should still join the network. Exactly this discrepancy exemplifies a fundamental E-Business problem, namely the general coordination challenge for networks: how to synchronize local and global efficiency? The significance of these two principal problems – the start-up problem and fundamental conflicts with respect to participation advantages – is their implication for the costs of a solution design for cooperation networks. In the first case, the start-up problem can in principle easily be overcome by simple coordination designs such as pure information intermediation. The reason is that network participation is advantageous from both the individual perspective of the local agent as well as the overall network perspective. In traditional neoclassical models, this is the default case. But the second problem, occasioned by the existence of network effects, is more difficult to address since it regularly requires expensive compensational mechanisms such as redistribution of costs. An example of this problem in the STP domain is the ‘wait and see’ attitude of many banks who are facing the decision of using the Transaction Flow Manager (TFM), a software solution of the ‘GSTPA Initiative’ for straight through processing primarily in the transnational securities business. The TFM is principally regarded as helpful if all actors participate. But due to substantial implementation costs, the cost-fairness ratio is evaluated quite diversely [Weit+03b]. Economic network analysis: network effects and efficiency

When discussing network efficiency, quite often the objective is an overall measure, such as duration of production processes throughout an entire value chain or aggregate (i.e. supply-chain-wide) cost efficiency (centralized solution). Still, in corporate reality such an overall solution, derived from an implicitly assumed collective utility function, does not describe the strategic investment decision of all the particular agents. Instead, they seek for an individually rather than collectively optimal decision (decentralized solution). This discrepancy is partly responsible for many network infrastructures to stay far behind their potential and is known as start-up problem in network effect theory (see above). In traditional neoclassical economics there is no difference between these settings if the validity of the fundamental theorems of welfare economics (HiKi76) can be proven. This is the case when certain premises are fulfilled as especially the absence of externalities. Unfortunately, network effects as a constituting particularity in networks are a form of externality, thus disturbing the automatic transmission from local to global efficiency (Weit+00).

Towards the E-Enterprise

369

In economics, an externality is considered to be present whenever the utility function Ui(.) of some economic agent i includes real variables whose values are chosen by another economic agent j without particular attention to the welfare effect on i’s utility. Generally speaking, in accordance with traditional literature on economics, a network externality exists if market participants fail to somehow internalise the impact of a new network actor on others; with positive network externalities the private value from another actor is smaller than the social value, leading to networks smaller than efficient. Thus the question arises how to internalise these effects, or in other words how to find coordination designs to build better networks. Equilibria in networks

If network effects imply multiple equilibria, it is crucial to be able to distinguish between them in terms of efficiency and achievability. If there is more than one equilibrium, which one will and should be achieved from the perspective of an individual and which one from the aggregate perspective of an entire network? Among the criteria discussed in the literature for distinguishing and evaluating different equilibria (NeMo67, 49-52), Pareto efficiency and Kaldor-Hicks (or Hicks-Kaldor) efficiency (Fees00, 54-57) are useful for network equilibrium analysis. A central – though not undisputed – concept in welfare economics, an equilibrium is called Pareto-efficient if no one can be made better off without at least someone being worse off. More formally, an allocation x is considered to be Pareto-optimal (or Pareto-efficient, respectively) if and only if no other allocation y exists which is weakly preferred over x by all individuals and strongly preferred by at least one individual. In neoclassical economics, markets move toward a Pareto optimum, since no individual would willingly trade goods or services in a manner that would harm his own welfare. The concept of Pareto efficiency conserves the ordinality of utility theory in avoiding interpersonal utility comparisons. Yet – as in the case of networks – there can be multiple Pareto-efficient equilibria. The Kaldor-Hicks criterion describes a preference order for various Pareto-efficient equilibria [Fees00, 56-67] [BöIl95, 259-260]. The basic idea is a possible compensation between players such that everyone is better off or at least not worse. A change (in policy or policy regime) from the present state should be undertaken if the gainers from the change could compensate the losers and still be better off. The criterion does not require that the compensation actu-

370

Wolfgang König, Tim Weitzel

ally be paid, which, if it did, would make this the same as the Pareto criterion. If the payoff matrix shows monetary values the Kaldor-Hicks criterion ranges the Pareto-efficient equilibria by aggregate values. That is why problems of fairness are often discussed in the context of Kaldor-Hicks optimality. In comparison, Kaldor-Hicks-superiority is a more lenient test than is Pareto superiority; the Kaldor-Hicks criterion is sometimes termed the “potential” Pareto criterion. A problem of Kaldor-Hicks as a yardstick for wealth maximization or evaluating legal rules, however, is that it implies the very interpersonal comparisons of utility that the Pareto criterion set out to avoid. The reason is that adding the losses to determine the compensation implies that a transfer of wealth reflects a transfer of utility in the same proportion. Table 5: Pay-off matrix for network participation game with two Pareto-efficient (P) and one Kaldor-Hicks-efficient (K) equilibrium

Player 2 Player 1

The game in table 1 has two Pareto-efficient equilibria in ((s11, s21); (s21, s22)). The Pareto criterion can identify the equilibria. But if there is more than one it does not provide any help to arrange them into a preference order. According to the Kaldor-Hicks criterion, (s12, s22) is the preferable equilibrium because there, player 1 is better off even considering side payments of 1 unit to player 2 to make up for her losses. Implications of the equilibrium concepts for cooperation networks

What network structures can we expect under centralized and under decentralized coordination and how efficient are they? For analyzing the efficiency of network equilibria, the two concepts of Pareto and Kaldor-Hicks efficiency as introduced above are used. Generally, positive network effects can be identified by the on-diagonal of the payoff matrix exceeding the offdiagonal as in the battle of the sexes coordination game of table 1 that got

Towards the E-Enterprise

371

its name from the strategic situation of a couple yearning for quality time together but disagreeing whether there is more quality in going to a boxing match or to the opera together. Since almost all models in the literature only consider positive network effects most standardization problems are found to have a battle of the sexes structure [FaSa88, 238], among them such famous cases as television standards (PAL, SECAM, NTSC) [Cran79] or AM stereo [BeJo86]. Existence and efficiency of resulting equilibria in networks strongly depend on the particularities of the players’ strategic decision situation as especially the information set available to them and the decision structure (static or singular vs. sequential vs. reversible decisions). One key finding from network equilibrium analysis is that in case of decentralized coordination there is no unique (pure strategy) equilibrium: either there is one equilibrium (i.e. no network participation) or there are two (both/no participate). This elementary finding is also valid for more complex games [Weit04]: there is a smaller propensity to join a decentralized network, constituting a frequent efficiency gap to centrally coordinated networks. This network (or efficiency) gap is depicted in figure 2 (section 4.3). When comparing the different equilibria with regard to the coordination form, it becomes obvious that whenever there are multiple equilibria, Kaldor-Hicks efficiency equals the centralized solution. Generally we find that most decentralized equilibria are Pareeto-efficient. But centralized solution quality is not always achieved. Especially the strategic situation of one player suffering individual (decentralized) losses from network participation but increasing the network value for many others (centralized), decentralized network efficiency can be improved. One possible solution are side payments. As a result there are two basic problems: when mutual network participation is desirable but there are multiple equilibria, network agents either need more information to close the part of the efficiency gap resulting from uncertainty about the participation behavior of the prospective network partners. Or there are situations where realizing the full network potential requires explicit coordination based on a redistribution of network participation costs and benefits. This differentiation will prove useful when comparing different decision scenarios as part of the simulation-based analysis in the following sections.

372

Wolfgang König, Tim Weitzel

Developing a network model After a brief description of the simulation model and the parameters used (4.1 and 4.2) we analyze individual network efficiency in centrally and decentrally coordinated networks. Basic model

The model is based on the works of [Weit04] [West+99] [Buxm+99]. For the use of simulation models see [Tesf02] [Vrie96] [Vrie99] [WeKö03]. A more detailed decentralized standardization model incorporating network topology, installed base effects, multiple technologies, etc. is provided in [Weit04], for an elaborate derivation of the centralized model see [Buxm96]. The benefits of network participation derive from improved information exchange between partners [Klme98, 63]. These improvements are associated with decreased information costs due to cheaper and faster communication and less converting and friction costs (media discontinuities) [BrWh85] [Thum95, 14-15] as well as more strategic benefits enabling a further savings potential like for instance just in time production (Pico+93). The costs of network participation include technical and organizational integration costs (hardware, software, training, etc.). To describe a network let Ki denote network participation costs of agent i and cij the benefits to agent i from j joining the network. To compare individual and aggregate network utility, two coordination regimes are modelled: a decentrally coordinated network models individual agents’ behaviour in the extreme case of no ex ante coordination or external force. In contrast and to benchmark the efficacy of eventual realistic solution strategies, a centrally coordinated network describes the maximum possible coordination quality, assuming that agency or coordination costs are resolved (at zero cost) and that there is a central manager who can determine and implement the optimal result network-wide. Thus the individual implications of network participation decisions are irrelevant from the centralized perspective. In a game theoretical equilibrium analysis it becomes clear that the centralized solution is a Kaldor-Hicks optimum determining the biggest savings yet to be (re)distributed [Fees00, 54-57] [BöIl95, 259-260]. The (decentralized) network participation condition of agent i is described by equation 1. Network benefits are modelled as deriving from cost savings due to the synchronization of processes with partners also participating in the network.

Towards the E-Enterprise

373

(1) However, given autonomous agents and the availability of a realistic information set, it is not clear that all partners actually join the network. Each agent i then needs to anticipate the behaviour of the other agents j . If it is assumed that all agents i know Kj and the network benefits directly associated with him (i.e. cij and cji) but no further data, like the network benefits between other agents, the problems is mainly anticipating the partners’ strategic network participation decisions. We assume that the agents are risk-neutral decision-makers and propose an approximation of their anticipation calculus according to equation 2 determining expected utility (ex ante). (2)

pij describes the probability with which agent i believes that j will join. If EXPECT [Ei] > 0 then agent i will join. Given the assumptions about the availability of data outlined above, pij can be heuristically computed as in equation 2 using all data presumed available to the deciding agent. Simulation design and parameters

Individual ex post benefits Ei to agent i (i ∈ {1,…,n}) from network participation are determined as the aggregate benefits for i derived from enhanced communication with all partners j also participating. In equation 3 the binary variable xj is one if j participates. (3) According to equation 4, CE (“coordination efficiency”) denotes aggregate network-wide savings resulting from standardization, i.e. the horizontal aggregation of all individuals' benefits. (4) The transformation in equation 4 enables deduction of CE from ex ante costs and ex post costs. yij takes a value of zero if both agents, i and j, have

374

Wolfgang König, Tim Weitzel

joined the network. The interpretation of aggregate savings, CE, is not unproblematic in a decentralized network, as will be elaborated. But it measures overall decision quality in decentrally coordinated networks (i.e. focusing on individual efficiency), especially in contrast to the potential coordination efficiency that is achievable in centrally coordinated networks (i.e. with no regard to individual participants but to the entire network) in the ideal situation of complete absence of coordination costs. The individual implications are the main focus of the rest of this paper. All simulation results were generated using JAVA 1.3 applications. For data analysis SPSS 10.0 was used. Overall, networks are randomly generated and all agents are assigned individual cost and benefit data associated with a potential network participation according to the distribution parameters. Once the network is generated, all agents decide according to their decision functions outlined above and reacting to the observable decisions of their network neighbours. More precisely, first a network is initialised assigning approximately normally distributed random values to Ki and cij (no cost values ≤ 0 are used). Having generated a network, the centralized solution is determined according to the centralized model described below. For this purpose, a linear program with all network data is formulated (equation 5.1-5) and solved using JAVA-packages lp.solve 2.0 by M. Berkelaar (http://siesta.cs.wustl.edu/~javagrp/help/LinearProgramming.html).

s.t.:

(5.1-5)

The entire simulation process is repeated 50 times before altering particular parameters such as mostly reducing µ(K) and then starting anew. Figure 2 thus consists of about 4,500 simulation runs. Simulation results

In figure 80, the intuitive discrepancy between centralized and decentralized networks is shown as the perpendicular distance between the central-

Towards the E-Enterprise

375

ized and the decentralized solution at different µ(K), quantifying the magnitude of the network potential accessible by adequate internalisation strategies (efficiency gap). Obviously, in decentralized networks there is a smaller propensity of network participation. Figure 80 shows box plots for individual savings Ei at different values of µ(K) with .(cen) indicating centralized and .(dec) decentralized coordination. The box borders denote the 0.25 and 0.75 quantiles and the bar in the box shows the median. The antennas show the whole range of resulting Ei (except few extreme outcasts). To clearly show changes of Ei over time, “true” individual savings in t=1 and stationary state T resulting from decentralized coordination are depicted. Interestingly, relatively few Ei are negative, especially in the case of centralized coordination. This means that cases of agents that are forced to participate against their will (or that would require compensations ex post in a decentralized context) are scarce. Obviously, not only the whole network but also the vast majority of individuals are better off getting the optimal solution from a central principal. The fraction of these individuals is of course highly responsive to deviations of costs and network effects and is analyzed in more detail in [Weit+03].

Figure 80: Individual decision quality Ei for different µ(K) resulting from centralized and decentralized network coordination

376

Wolfgang König, Tim Weitzel

These findings and especially the underlying systematic concepts are quite relevant to overall coordination efficiency considerations in network: In cases where there are no Ei(cen) < 0 the centralized solution is Paretosuperior to the decentralized solution. This implies that all agents are better off when joining the network and even if they were forced to do so there won’t be any regrets afterwards. These strategic situations, as will be discussed in section 4, are principally less intricate to address since mutual coordination gains are possible. This strategic situation is sometimes referred to as a win-win situation especially in contrast to constant sum games that imply, as partly in situations when Ei(cen) < 0, more difficult or even no existing (Pareto-efficient) solution strategies at all.

Implications of the findings

Approaches to solving the start-up problem

What are the implications of the simulation results for possible solution strategies and how can they help to identify advantageous designs to overcome the start-up problem depicted in figure 80? There are two principle cost/benefit situations when trying to achieve centralized solution quality in decentralized networks. Either the gap can be explained by agents wrongly anticipating their environments' actions or some agents that should join a network from a central perspective are individually worse off doing so. While in the second case in a decentralized network some form of redistribution needs to be established, in the first case reducing uncertainty is in principal sufficient, i.e. designs aimed at enhancing the “information quality” concerning the respective network decision behaviour (i.e. to solve the start-up problem). Generally, insurance solutions seem applicable to solving network problems as described: agents not only join a network but are also granted an exogenously enforceable promise to compensate for all costs exceeding the benefits. If the option issuer is convinced (and rightly so) that his chosen network of participants will pay off for everybody then this is an elegant way of overcoming the start-up problem. This, of course, assumes that the set of participants has already been determined. Additionally, there are the problems (and costs) of objectively determining cost and benefit data. For the “simple” equilibria of the first kind, “cheap talk” mechanism as means of information intermediation can close parts of the gap by coupling ex ante with ex post decisions. For example, a black board

Towards the E-Enterprise

377

or “round table” could start the desired starting a bandwagon process (FaSa86). There has been discussion suggesting that agents join the network if they expect to be better off than otherwise. But due to the information asymmetries in decentralized cases, some agents can be worse off ex post if others “make a mistake”. A general allocation procedure aimed at overcoming this problem is to offer each agent a choice between his status quo ante and his actual endowment at equilibrium state. The final allocation is only implemented if and only if all agents agree that they (weakly) prefer their current endowment and are therefore better off [Vari94]. If some agents object to the status quo, they could be excluded from a replay among the remaining agents. Moreover, if particular agents would suffer individually but contribute to aggregate efficiency (the second case described above) a compensation plan needs to be established since the option model would not per se solve the start-up problem by its pure existence but also some side payments would have to be made. But according to the simulations, these cases are quite rare. There are a lot of limitations of this simple model. Like most parts of network effect theory, only positive network effects are considered favoring win-win situations, i.e. the “simple” equilibria. Besides snob effects a loss of individuality could be a substantial negative effect especially in an EBusiness context [Thum95, 26]. Industry consortia as solution concepts

The findings of the network equilibrium analysis and simulation studies indicate that consortia structures appear to be an important coordination effort for networks. There are many mechanisms found in real life aimed at reducing informational asymmetries concerning mutual goals and to coordinate individual decisions in a way that makes the participants better off. In the literature, especially for standardization and integration problems between enterprises, one principal way of closing the coordination efficiency gap is agreeing on a common strategy in a committee. In fact, there is a cornucopia of organizations concerned with standards for networks; some of them are governmental, some are privately organized. Recently, industry consortia such as W3C or OASIS seem to have produced standards faster than governmental organizations. If certain industries especially in the high-tech area require faster (technology life cycle) and more (existence of network effects) coordination than more traditional branches, the question arises what coordination form might be favourable. The simulations

378

Wolfgang König, Tim Weitzel

clearly showed a coordination demand e.g. to overcome the start-up problem. Additionally, equilibrium analysis revealed in what cases coordination requires informational agreements rather than subsidies. This makes consortia an attractive solution principle. One hypothesis why industry consortia might be favourable to legislative or governmental committees is not only that they are thought to be faster but they are also less binding, as Farrell and Saloner found to be of advantage: Farrell/Saloner (1988) model three coordination processes for agreeing on a standard. 1. First, they model a committee: if agents can continuously meet to agree on a joint action this committee is faced with Battle of the Sexes situations each period (while the overall game is a war of attrition [FuTi86]. 2. Then the market solution (“bandwagon”) is modelled as a grab-the-dollar game (the player who grabs first wins, but both lose when grabbing simultaneously) with basically the same structure. For both cases the mixed-strategy equilibrium shows the effectiveness of committee negotiation and market action respectively. 3. Last, there is a hybrid between the former two coordination processes. In any period, committee members can also take market actions (e.g. adopt a standard on the market). It is shown that the committee dominates the market with the difference diminishing with the number of periods. “In a committee…nothing is likely to happen for a long time” while on the market early action is likely [FaSa88, 239]. But pure market coordination in unsponsored networks often produces inefficient oligopolies. The hybrid solution combines the advantages of both approaches.

Conclusions Looking at the individual consequences of network participation decisions it has been shown that the kinds of wrong decisions made by agents in decentralized networks are quite promising with respect to possible solution strategies. Notably, it appears that in many cases information intermediation can solve the start-up problem without the necessity of heavy redistribution of costs and/or benefits which is regularly much more difficult and expensive. After all, the existence of externalities makes one expect exactly this: the implications of one's decision on those of the others can-

Towards the E-Enterprise

379

not fully be considered. As in a neo-classical market with externalities, price has not enough dimensions to capture this information. The results presented in this paper can certainly only be small contributions to a better understanding of the complex dynamics in networks. Most notably, extending the modelling power in order to capture important determinants of network behaviour identified in other disciplines such as sociology or geography like network topology and density, agent size and installed base effects will allow a more sophisticated analysis. Requirements concerning the modelling power of an eventual interdisciplinary theory of network effects have been proposed in [Weit+00]. In the context of relevant solution strategies for corporate networks, using the classification of costs, benefits, and the resulting strategic situations it appears promising to work towards systematically developing and evaluating internalisation strategies based upon the findings from parts of the literature on game theory and managerial controlling to unite methodologies successfully applied for controlling resource dependencies but failing to capture network effects as in the controlling literature with network effect theory that has yet been unable to develop decentrally applicable solution strategies. One approach in this context could be using the model to analyse the influence of local coalitions in networks: agents could make binding agreements with their most important (e.g. biggest cij) one, two, or more partners imitating centralized decision behaviour within their individual clusters. Model extensions in [Weit+03] show how a “consortial ladder” can be used to coordinate real life cooperation networks by establishing consortia to address start-up problems among largely homogeneous partners (e.g. from the same supply-chain tier). These can then try to coordinate the interrelations between these more homogeneous clusters (e.g. entire supply-chain), therefore drastically reducing the number of necessary communication links (“Metcalfe's Law” (ShVa98, 184)). If co-ordination costs increase with the number of partners forming the coalition, we can expect to find an optimum number of internalisation partners per network tier and possibly hints at optimum network sizes. Empirical data concerning the cost development of the coordination costs compared to Ki, for example, could provide especially valuable assistance in evaluating coalitions as such. Application domains suitable for retrieving empirical data and for testing the results are, among others, EDI networks, corporate intranets and certainly the STP community.

380

Wolfgang König, Tim Weitzel

Bibliography [Agre98]Agre, P. (1998): A graduate seminar on the economic, legal, and strategic aspects of technical compatibility standards, http://dlis.gseis.ucla.edu/people/pagre/standards.html. [Arth83]Arthur, W.B. (1983): Competing technologies and lock-in by historical small events: the dynamics of allocation under increasing returns; International Institute for Applied Systems Analysis Paper WP-83-92, Laxenburg, Austria (Center for Economic Policy research, paper 43, Stanford). [Arth89]Arthur, W.B. (1989): Competing technologies, increasing returns, and lock-in by historical events, in: The Economic Journal, 99 (March 1989), 116131. [Arth96]Arthur, W.B. (1996): Increasing returns and the new World of business, in: Harvard Business Review, 74 (July-August 1996), 100-109. [BeFa94]Besen, S.M.; Farrell, J. (1994): Choosing How to Compete: Strategies and Tactics in Standardization, in: Journal of Economic Perspectives, vol. 8 (1994), no. 2, 117-131. [BeJo86]Besen, S.M.; Johnson, L.L. (1986): Compatibility Standards, Competition, and Innovation in the Broadcasting Industry, in: RAND Corporation R3453-NSF, November 1986. [BöIl95]Böventer, E.v.; Illing, G. (1995): Einführung in die Mikroökonomie, 8. Auflage, München; Wien 1995. [BrKa95]Branscomb, L.M.; Kahin, B. (1995): Standards processes and objectives for the National Information Infrastructure, in: Kahin, B./ Abbate, J. (eds.): Standards Policy for Information Infrastructure, Cambridge: MIT Press, 1995. [Brow01]Brown, J. (2001): Twists and turns on the straight-through route, in: EuroMoney (April 2001) 384, p. 112-118. [BrWh85]Braunstein, Y.M.; White, L.J. (1985): Setting technical compatibility standards: An economic analysis, in: Antitrust Bulletin, vol. 30 (1985), 337355. [Bund98]Bundesumweltministerium (1998): Bundesumweltamt (ed.): So geht's auch! Gasantrieb, Bonn 1998. [Buxm+99]Buxmann, P.; Weitzel, T.; König, W. (1999): Auswirkung alternativer Koordinationsmechanismen auf die Auswahl von Kommunikationsstandards, in: ZfB (Zeitschrift für Betriebswirtschaft), Ergänzungsheft 02/99 Innovation und Absatz, 133-151. [Buxm96]Buxmann, P. (1996): Standardisierung betrieblicher Informationssysteme, Wiesbaden 1996. [CeKa82]Ceci, S.J.; Kain, E.L. (1982): Jumping on the bandwagon: The impact of attitude polls on polling behaviour, Public Opinion Quarterly, 46, 228-242. [Cowa90]Cowan, R. (1990): Nuclear Power Reactors: A study in technological lock-in, in: The Journal of Economic History, vol. L, no. 3 (Sept. 1990), 541567.

Towards the E-Enterprise

381

[Cran79]Crane, R. (1979): The Politics of International Standards, Norwood, N.J., 1979. [DaGr90]David, P.A.; Greenstein, S. (1990): The economics of compatibility standards: An introduction to recent research, in: Economics of innovation and new technology, 1, 3-41, 1990. [Davi85]David, P.A. (1985): Clio and the economics of QWERTY, in: American Economic Review, Papers and Proceedings, vol. 75, 332-337. [EcHi95]Economides, N.; Himmelberg, C. (1995): Critical Mass and Network Size with Application to the US FAX Market, Discussion Paper EC-95-11, Stern School of Business, New York University. [Econ00]Economides, N. (2000): An interactive bibliography on the economics of networks and related subjects, http://www.stern.nyu.edu/networks/biblio.html. [Emme93]Emmelhainz, M.A. (1993): EDI: A Total Management Guide, 2nd edition, New York 1993. [FaSa85]Farrell, J.; Saloner, G. (1985): Standardization, Compatibility, and Innovation, in: Rand Journal of Economics 16, 1985, 70-83. [FaSa86]Farrell, J.; Saloner, G. (1986): Installed Base and Compatibility: Innovation, Product Preannouncements, and Predation, in: The American Economic Review, vol. 76, no. 5 (December 1986), 940-955. [FaSa87]Farrell, J.; Saloner, G. (1987): Competition, compatibility, and standards: The economics of horses, penguins, and lemmings, in: Gabel H. Landis (ed.): Product standardization and competitive strategy, Amsterdam 1987, 1-21. [FaSa88]Farrell, J.; Saloner, G. (1988): Coordination through committees and markets, in: RAND Journal of Economics, Vol. 19, No. 2 (Summer 1988), 235-252. [Fees00]Feess, E. (2000): Mikroökonomie – Eine spieltheoretisch- und anwendungsorientierte Einführung, 2. Auflage, Marburg 2000. [Fric+02]Fricke, M.; Weitzel. T.; König, W.; Lampe, R. (2002): EDI and Businessto-Business Systems: The Status Quo and the Future of Business Relations in the European Automotive Industry, in: Proceedings of the Sixth Pacific Asia Conference on Information Systems (PACIS-2002), http://muchmagic.wiwi.uni-frankfurt.de/~tweitzel/paper/EDI-automotive.pdf. [FuTi83]Fudenberg, D.; Tirole, J. (1983): Learning by doing and market performance, in: Bell Journal of economics 14(1983), pp. 522-530. [Gabe87]Gabel, H.L. (1987): Product Standardization and Competitive Strategy, Amsterdam 1987. [Gabe91]Gabel, H.L. (1991): Competitive Strategies for Product Standards, London 1991. [Gand94]Gandal, N. (1994): Hedonic price indexes for spreadsheets and empirical test for network-externalities, in: Rand Journal of Economics, vol. 25 (1994), no. 1, 160-170. [GeHö95]Gerwin, J.; Höcherl, I. (1995): Video 2000: Strategische Produktpolitik bei internationalem Wettbewerb, in: Brockhoff, Klaus (ed.): Management von

382

Wolfgang König, Tim Weitzel

Innovationen: Planung und Durchsetzung; Erfolge und Misserfolge, Wiesbaden 1995, 17-44, 217-244. [Gilk01]Gilks, D. (2001): The GSTPA Is Primed For Pilot. The European Union Banking & Finance News Network (EUBFN): 2001, http://www.eubfn.com/arts/gstpa2.htm [Gröh99]Gröhn, A. (1999): Netzeffekte und Wettbewerbspolitik. Eine ökonomische Analyse des Softwaremarktes, Kieler Studien 296, Tübingen. [Habe89]Habermeier, K.F. (1989): Competing technologies, the learning curve, and rational expectations, in: European Economic Review 33 (1989), 12931311. [HaTe90]Hartmann, R.S.; Teece, D.J. (1990): Product emulation strategies in the presence of reputation effects and network externalities: some evidence from the minicomputer industry, in: Economics of innovation and new technology, vol. 1-2, 157-182. [Heme75]Hemenway, D. (1975): Industry wide voluntary product standards, Massachusetts 1975. [HiKi76]Hildenbrand, W.; Kirman, A.P. (1976): Introduction to equilibrium analysis, North-Holland, Amsterdam, 1976. [KaSh85]Katz, M. L.; Shapiro, C. (1985): Network externalities, competition, and compatibility, in: The American Economic Review, vol. 75, no. 3 (June 1985), 424-440. [KaSh86]Katz, M. L.; Shapiro, C. (1986): Technology adoption in the presence of network externalities. in: Journal of Political Economy, vol. 94 (1986), no. 4, 822-841. [KaSh92]Katz, M. L.; Shapiro, C. (1992): Product Introduction with Network Externalities, in: Journal of Industrial Economics, 40 (1) (1992), 55-83. [KaSh94]Katz, M. L.; Shapiro, C. (1994): Systems Competition and Network Effects, Journal of Economic Perspectives, Spring 1994, 8, 93-115. [Kind83]Kindleberger, C. P. (1983): Standards as Public, Collective and Private Goods. Kyklos – International Review for Social Sciences, 36(3), 377-396. [Klei93]Kleinaltenkamp, M. (1993): Standardisierung und Marktprozeß – Entwicklungen und Auswirkungen im CIM-Bereich, Wiesbaden 1993. [Klme98]Kleinemeyer, J. (1998): Standardisierung zwischen Kooperation und Wettbewerb, Frankfurt 1998. [Leib50]Leibenstein, H. (1950): Bandwagon, snob, and Veblen effects in the theory of consumers demand, in: Quarterly Journal of Economics, 64 (2), 183207. [LiMa90]Liebowitz, S.J.; Margolis, S.E. (1990): The fable of the keys, in: Journal of Law and Economics, vol. 33, 1990, 1 – 25. [LiMa94]Liebowitz, S.J.; Margolis, S.E. (1994): Network Externality: An Uncommon Tragedy, in: The Journal of Economic Perspectives, Spring 1994, 133-150. [LiMa95a]Liebowitz, S.J.; Margolis, S.E. (1995a): Are Network Externalities A New Source of Market Failure?, in: Research in Law and Economics, 1995.

Towards the E-Enterprise

383

[LiMa95b]Liebowitz, S.J.; Margolis, S.E. (1995b): Path Dependence, Lock-In, and History, in: Journal of Law, Economics and Organization, April 1995, 11, 205-226. [Maub+00]Mauboussin, M.J.; Schay, A.; Kawaja, S.G. (2000): Network to Net Worth, The Rising Role of the Replacement Cycle, Credit Suisse First Boston, Equity Research, Frontiers of Finance, vol. 5 (May 11, 2000), http://www.capatcolumbia.com/Articles/FoStrategy/Ni1924.pdf. [Moch95]Moch, D. (1995): Ein hedonischer Preisindex für PC-Datenbanksoftware: Eine empirische Untersuchung, in: Harhoff, D./Müller, M. (eds.): Preismessung und technischer Fortschritt, Baden Baden. [NeMo67]Neumann, J.v.; Morgenstern, O. (1967): Spieltheorie und wirtschaftliches Verhalten (Theory of Games and Economic Behavior). Deutsche Übersetzung der 3. Auflage von M. Leppig, Würzburg 1967. [Nigg94]Niggl, J. (1994): Die Entstehung von Electronic Data Interchange Standards, Wiesbaden 1994. [OrSm81]Oren, S.S.; Smith, S.A. (1981): Critical Mass and Tariff Structure in Electronic Communications Markets, Bell Journal of Economics, Autumn 1981, 12, 467-87. [Pico+93]Picot, A.; Neuburger, R.; Niggl, J. (1993): Electronic Data Interchange (EDI) und Lean Management, in: Zeitschrift für Führung und Organisation, Nr. 1/1993, p. 20-25. [Plum37]Plummer, A. (1937): New British Industries in the Twentieth Century, London 1937. [Reut01]Reuters (2001): o.V., in: Venture Financial Systems Group: Straight Through Processing – Just in Time, 2001, www.venturefsg.com/article2.htm, service, in: Bell Journal of Economics 5(1), 1974, 16-37. [Rohl74]Rohlfs, J. (1974): A theory of interdependent demand for a communications [Ruh+01]Ruh, W.A.; Maginnis, F.; Brown, W. (2001): Enterprise Application Integration: a Wiley tech brief, New York 2001. [Schi00] Schinzer, H. (2000): Electronic Commerce, in: Mertens et al. (ed.): Lexikon der Wirtschaftsinformatik, 3., vollständig neu bearbeitete und erweiterte Auflage [Shah02]Shahrawat, D. (2002): EAI's Impact on Financial Institutions – An Interview With Dushyant Shahrawat. In: EAI Journal, Feb. 2002, http://www.eaijournal.com/PDF/Shahrawat.pdf. [ShVa98]Shapiro, C.; Varian, H. R. (1998): Information rules: A strategic guide to network economy, Boston, Massachusetts. [SWIF00]Swift Solutions – July 2000, Nr.7, Informationsbroschüre, www.swift.com/index.cfm?item_id=3488. [Tesf03]Tesfatsion, L. (2002): Agent- based computational economics, http://www.econ.iastate.edu/tesfatsi/ace.htm. [Thum95]Thum, M. (1995): Netzwerkeffekte, Standardisierung und staatlicher Regulierungsbedarf, Tübingen 1995.

384

Wolfgang König, Tim Weitzel

[Vari94]Varian, H. (1994): A Solution to the Problem of Externalities When Agents are Well-Informed, in: American Economic Review, December 1994, 1278-93, also: Microeconomics/Economics Working Paper Archive at WUSTL, RePEc:wpa:wuwpmi:9401003. [Vrie96]Vriend, N. (1996): Rational Behavior and Economic Theory, Journal of Economic Behavior and Organization 29 (1996), 263-285. [Vrie99]Vriend, N. (1999): Was Hayek an ACE?, Working Paper 403, Queen Mary and Westfield College, University of London, UK, May 1999. [WaWa83]Warren, M.E.; Warren, M. (1983): Baltimore – when she was what she used to be, 1850-1930, Baltimore: John Hopkins University Press. [Weit+00]Weitzel, T.; Wendt, O.; Westarp, F.v. (2000): Reconsidering Network Effect Theory, in: Proceedings of the 8th European Conference on Information Systems (ECIS 2000), http://www.wiwi.uni-frankfurt.de/~westarp/publ/recon/reconsidering.pdf. [Weit+01]Weitzel, T.; Harder, T.; Buxmann, P. (2001): Electronic Business und EDI mit XML, Heidelberg 2001. [Weit+03]Weitzel, T.; Beimborn, D.; König, W. (2003): Coordination In Networks: An Economic Equilibrium Analysis, forthcoming in: Information Systems and e-Business Management (ISeB), Springer 2003. [Weit+03b]Weitzel, T.; Martin, S.; König, W. (2003): Straight Through Processing auf XML-Basis im Wertpapiergeschäft, in: Wirtschaftsinformatik 4/2003. [Weit04]Weitzel, T. (2004): Economics of Standards in Information Networks, Springer/Physica, New York, Heidelberg 2004. [WeKö03]Weitzel, T.; König, W. (2003): Standardisierungstheorie und Netzeffekte: Agent-based Computational Economics als wirtschaftsinformatischer Beitrag zu einer interdisziplinären Netzwerktheorie, erscheint in: WIRTSCHAFTSINFORMATIK 5/2003. [WeSK01] Weitzel, T.; Son, S.; König, W. (2001): Infrastrukturentscheidungen in vernetzten Unternehmen: Eine Wirtschaftlichkeitsanalyse am Beispiel von X.500 Directory Services, in: WIRTSCHAFTSINFORMATIK 4/2001, 371381. [West+99]Westarp, F.v.; Weitzel, T.; Buxmann, P.; König, W. (1999): The Status Quo and the Future of EDI, in: Proceedings of the 1999 European Conference on Information Systems (ECIS'99). [West03]Westarp, F.v. (2003): Modeling Software Markets – Empirical Analysis, Network Simulations, and Marketing Implications, Physica, Heidelberg, 2003. [WeWe02]Weitzel, T.; Westarp, F.v. (2002): From QWERTY to nuclear power reactors: Historic battles for the standard, in: Geis; Koenig; Westarp (eds.): Networks – Standardization, Infrastructure, and Applications, Berlin, New York (Springer) 2002, 33-61. [Wies90]Wiese, H. (1990): Netzeffekte und Kompatibilität, Stuttgart 1990. [Yang97]Yang, Y. (1997): Essays on network effects, Dissertation, Department of Economics, Utah State University, Logan, Utah.

Andreas Pfeifer

The evolution to real-time architecture

The ability of a company to trade in real-time depends very much on a highly-developed IT infrastructure. However, the current landscape of IT infrastructure found in companies tends to obstruct the evolution to a realtime enterprise more than to support it. As a result, the IT infrastructure must be adapted to the requirements of a real-time enterprise in three phases in parallel to and in advance of the design of business processes and IT applications: 1.) Consolidation and standardization, 2.) Development into an “infrastructure on call”, 3.) Expansion to one of an external operatorrelated infrastructure. Only an IT infrastructure that quickly matures in accordance to these three phases will no longer hinder the progression of the company to a real-time enterprise, but clearly represents a supporting function of this development. This is exactly what distinguishes top companies.

Real-time capability leads to company success Fast reaction in the manner of a real-time enterprise distinguishes top companies. Only if a company quickly recognizes changes in its environment and their effects of these changes can it successfully position itself to ensure present and future value-added opportunities. The speed of a realtime enterprise allows it to rapidly implement the acquired perceptions into new differentiating work flows and to ensure a position as an innovator in the field of its own core competencies.1 2 If not only a company but an entire industry develops into a real-time industrial sector in which both communication and business processes between companies can take place without unnecessary time losses, this will form the basis for the outsourcing of company activities that do not belong to their core competencies and, as a 1 2

See Davenport et al. 2003, p. 3, Linder et al. 2003 p. 45. 2003 See also O’Mahony et al. 2003.

386

Andreas Pfeifer

result, to achieve or ensure the status of a top company. Hence, companies that can react more quickly are the winners. They can only react quicker, however, if their IT infrastructure also allows it. Hence this article puts forward the following propositions: Only a real-time capable IT infrastructure allows a company to develop into a real-time – and, consequently, a leading – enterprise.3 Today’s infrastructure has, however, developed in an organic way and cannot adequately support this development. To achieve a real-time capable IT infrastructure, the current IT-landscape must firstly be consolidated and standardized, then converted into an internal “infrastructure on call” and, finally, in the areas where its makes sense, no longer provided internally but by an external service provider. Only an infrastructure developed in this way allows the company to evolve into a real-time and, consequently, leading enterprise. This article discusses the development of the IT infrastructure in a realtime enterprise. The development of the strategy, processes and applications are not covered in detail. The first section is concerned with the requirements of a real-time enterprise regarding the infrastructure. The second section shows that most infrastructures found in today’s companies are in no way real-time capable. It then describes the evolutionary steps which appear to be necessary in order to achieve a real-time capable infrastructure. The last section covers the evolution of the infrastructure within the development of the entire company and examines the fact that only with a real-time capable infrastructure does a leading real-time enterprise come into being.

Requirements placed on the IT infrastructure of a realtime enterprise The Gartner market research institute defined a real-time enterprise as “an enterprise which gains a competitive advantage by reducing delays in the management and execution of critical processes using up-to-date information”.4 The status of the development to a real-time enterprise can be measured by the answer to the following question: “How long does it take to react to a significant event or to solve problems which arise?”5 For a further definition of a top company, see Brien 2003, p. 20f. See Drobik et al. 2002, p. 1 and Flint 2003. 5 See Raskino et al. 2002, p. 2 3 4

The evolution to real-time architecture

387

There are many substantial events and reaction cycles in a company, which take place across a complete organization. The number is considerably larger if not just one company is considered in isolation, but rather the entire value chain of an industry. IT can significantly reduce the time that a company requires to react to an event.6 Alexander Röder, CIO of O2, is informed each morning of the latest customer numbers by SMS.7 IT can initiate the processes of the transfer from one link of the chain to the next as well as the formulation of a question which supports the conclusion and communication of the answer or the final solution of the problem. Internet technology is used thereby to reduce communication costs and to accelerate the processes in a sustained way.8 However, this is only possible if the infrastructure permits the relevant form of communication. The IT infrastructure must enable communication between humans, and by humans with and between applications in real time from anywhere and at any time. If the infrastructure is divided into the areas of workplace, data center and network, the workplace must allow users everywhere to gain access to all systems and communication channels at any time, the data center must allow applications to be reachable at any time and to react quickly to requests, and the network must enable coordinated communication between humans, by humans with and between applications within companies and across company boundaries in order to prevent media discontinuities. The workplace – everywhere and at any time

A workplace that best supports a real-time enterprise must allow the users to carry out their work at any time and at any place. Users are becoming ever more mobile. They work in different offices, on customer premises, from home or en route, e.g. in the airplane. A workplace supports this and can be found wherever the worker is carrying out his allotted tasks. The IT infrastructure provides him with access at any time to the tools and communication routes he requires.9 See Roby 2003 See Ellermann 2003 8 See Siegel 2002, p. 32 9 See Mattes; Emmerson; 2003, p. 28: “The end user experiences a seamless, integrated interface which is created by software for language and data.” 6 7

388

Andreas Pfeifer

earlier …

… today

Figure 81: Development of the workplace

If the user works at different locations, it must be possible for the configuration of the workplace to follow him across the different locations (the roaming workplace). If the user works outside the office, the workplace must also allow him access to all tools and data relevant to him (the remote workplace); of course, this is also true if the work is carried out at home (the home workplace). In order to prevent any unnecessary delays in a realtime enterprise for contact partners who are on the road, it is particularly important that he can be reached and that, at the same time, he has access to the necessary tools and communication channels (the public workplace). Depending on the specific requirements of the user, the workplace must support this with corresponding terminals such as laptops, PDAs, Smart Phones and Thin Clients.10 The availability and functional capability of the terminals is ensured by adequate support (Help Desk, etc.). However, a suitable workplace architecture has far-reaching consequences over and above the choice of terminal. The extended possibilities of the workplace may necessitate the changeover and migration of users, data and applications. The different requirements of the user, e.g. the user with a variety of office workplaces, the user with a home workplace and the field service employee have far reaching effects on the design of the respective workplace. Effective management of the workplace is becoming increasingly challenging from a technical, logistical and financial perspective.11 New functions, See Curtis 2003, p. 13 11 See Schmalzl 2004, p. 111 ff. for details of the economics of workplace design 10

The evolution to real-time architecture

389

higher reliability, easier and more secure access and high quality of service are just as necessary as lower costs and shorter implementation times (the flexible or mobile workplace concept). Real-time data center – flexibility in computing and storage

Real-time companies require flexible computing capacity and, consequently, the computing capacity of external providers, which is charged according to the level of use. Even the hardware trend for servers is moving in the same direction. The capabilities of standard components is growing. Because of the rising performance of processors, high-performance computers can be offered at increasingly low prices. This trend is reinforced by the use of 64-bit processors. The consolidation of the chip market is leading to a standardization of the server environment and the progressive standardization of hardware is resulting in a further expansion of the outsourcing of hardware support. Computer performance has consequently become a commodity provided increasingly by external suppliers as required. The supply of computer performance is becoming more and more like a utility as represented today by the electricity and water service companies. In order to save costs, companies are developing management systems for high-volume data storage to an increasing degree. New storage management software will provide the basis of the fact that shared storage will prevail throughout the complete infrastructure. For this reason data storage has developed from a standard option in the configuration of server platforms into a strategic company facility. It’s easier to make data storage available in the network than to directly allocate it to individual devices, although storage environments also become more complex as a result. Growing data volumes are leading to storage management services, which are complicated by the interdependency of processes and devices. As a result, companies fall back on external resources. Also the trend for storage is leading to storage ‘on call’ or to the creation of virtual storage. Storage space can be made available by an external provider as the need arises. The network – seamless combination of data and voice

The realization of a real-time enterprise requires a network that can connect a multitude of the most varied groups of users to different locations with a wide variety of terminals. The network of the future must be oriented to changing requirements during the process of evolution of the company into

390

Andreas Pfeifer

a real-time enterprise. To enable real-time communication, it is not enough just to connect the employees of a company with one another. Customers and suppliers must also be included; subcontractors and representatives need to be connected not only with the communication network, but also to the data network.

Mobile customers

Production • Every employee, customer, supplier, subcontractor, representative,…

Zuhause Customers Offices Field service employees Outsourced applications (ASP)

Dial-up Small Office, Home Office

Network

Internet pages/applications

• All PCs, telephones, PBXs, servers, Internet pages, PDAs, vehicles, internal & external applications, Internet, messages, video, audio… • Everywhere office, at home, hotel, vehicle

Unified Messaging Video conference Public telecommunication network

Security, Directory

Suppliers

Vehicles

Figure 82: Data network of the future

Delays occur particularly often in communication-intensive processes if employees are working at different locations or while they are traveling. The logical result is: dialog partners cannot be reached, do not react to EMail or only answer them after a considerable time delay. Another reason for holdups in the business process are breaches in the communication flow, e.g. between two different media or between man and machine. Breaks in communication can only be prevented by an integrated architecture. Voice and data systems use the same network in an integrated architecture of a real-time enterprise. Thus, all voice services can be integrated into the IT landscape. In order to allow the mobile user to reach the required applications using a variety of terminals at any time, the network of a real-time enterprise must connect different systems (PCs, telephones, telephone sys-

The evolution to real-time architecture

391

tems, servers, websites, PDAs, vehicles, internal & external applications, Internet, ...) with one another. These devices are used at a great many locations such as offices, home workplaces, hotels, vehicles, etc. The network enables the communication and the data transfer in this case. All users will be connected with the whole range of devices at any location by the data and communication network of the future. Technological developments together with the strong demand of companies for cost-effective solutions form a major driving force for networks of the next generation. Mainframe and PC/LAN applications TDM, leased lines and shared LANs were largely used in the 80s. In the 90s, client servers and web applications that used Frame Relay, ATM and Routed Internetworks prevailed. Since 2000 there has been a trend towards IPbased, network-centered applications, and thus also to the development of IP-based VPNs, intelligent, context-sensitive networks. The capacity requirement of a network can vary enormously. If the data need to be restored, this results in capacity peaks. Even the action of a retailer can result in transient massive access to the Internet page, leading to capacity peaks in the network (e.g. Amazon.com). To avoid having to provide huge capacities which are barely used most of the time, a trend is developing where network service is based on an external provider who is paid for it’s use.

The constraints of today’s IT infrastructure landscape The landscape of IT infrastructure found in companies today can no longer meet the requirements of a real-time enterprise. The IT infrastructure has evolved organically. Often it was not designed within the framework of a company-wide standard strategy, but each business area developed its own infrastructure under its own direction. Company mergers have also frequently contributed to the diversity of the infrastructure. IT and TC infrastructures were established separately from one another and, on top of that, still lie mostly within different areas of responsibility. Data centers were set up as required. To some extent, each application has its own server. Most companies have a non-standardized desktop infrastructure. There is scant control of the software which is mostly still supported manually by the respective local IT support service or is installed directly by the user. In the majority of cases security guidelines are missing with the result that terminals have been mostly excessively customized to the user. Thus, both the costs for the upkeep of the desktop structure rise as well as the competencies nec-

392

Andreas Pfeifer

essary to support this heterogeneous landscape. Business-critical applications remain on old platforms whose maintenance is expensive. Current systems mostly do not support the mobility requirements to a sufficient degree. Even server landscapes are often non-standardized, with many different platforms in use. Because of a lack of automation and a vast number of manual processes and tools, the management and maintenance are mainly very personnel intensive. Since the old hardware often works imperfectly, this frequently leads to the breakdown of critical systems. Due to the different operating systems and other server software, the licensing costs are unnecessarily high. Maintenance is expensive and the know-how demanded from the service organization is extensive. The evolution of the network has mainly been organic in nature and, as a result, it is complex, heterogeneous, inefficient and rigid. There are different networks for data, voice, video and remote access. Because of its heterogeneity, the maintenance requires a high number of personnel. The networks have security loopholes and are costly. This complex, non-standardized infrastructure results in high costs throughout the entire lifecycle. Such a widely varied infrastructure landscape does not support the advancement to a real-time enterprise; in fact, it stands in the way.

Three phases for the infrastructure of a real-time enterprise How can a company develop its current infrastructure so that it fulfils the requirements of a real-time enterprise? This certainly is not a development that can take place at once. Achieving the ideal infrastructure for supporting a real-time enterprise from the current infrastructure landscape is much more a case of proceeding along an evolutionary path which results in part from the requirements of changing business processes, but is also in part clearly a case of the development of the business processes progressing in a timely way.12 Accenture has developed a three-phase model for advancing resolutely towards company objectives along the evolutionary path (see Figure 83). In the initial phase the diversity of the infrastructure is unified by consolidaSee O’Mahony et. al. 2003, p. 27: ”Innovative organizations consider new capabilities IT can provide when they are setting their business strategy, rather then treating IT simply as a facilitator of a business strategy that has already been decided.” 12

The evolution to real-time architecture

393

tion and standardization. During the second phase the infrastructure is converted into an “infrastructure on call” which can provide different capacities according to demand. This is a major step for positioning the company for an infrastructure of an external provider whose capacity can be dynamically adapted to the respective requirements. Everywhere & at any time 2008

2006

lace

Dynamic system resources (commodity)

Wor kp

2007

er

nt

ta

ce

Da

2005

rk

Netwo

Universal IP fixed line and mobile network for voice & data

2004 Phases:

1.Consolidation and standardization of the existing infrastructure

2.Conversion into an internal infrastructure on call

3.Development to dynamic infrastructure of an external supplier

Figure 83: Development phases of the infrastructure

In the last phase the infrastructure ultimately becomes a commodity. System resources can be used dynamically at any time and from any location. This is facilitated by a standardized mobile workplace, the use of a data processing center of an external provider whose capacity can react flexibly to changes in demand, and an IP network which seamlessly enables data and voice communication between human beings, machines and man & machines. Consolidation and standardization

In order to reduce the current complexity, the costs and the problems in sustaining the infrastructure, a consolidation and standardization of the present infrastructure is necessary. This simultaneously represents the initial preliminary step for using the “capacity on call” model of an external provider. To successfully design a consolidation and standardization of the workplaces, the requirements of the workplaces must be introduced early on. This requires the current workplace design to be considered in a first step. Experience from different consolidation and standardization projects shows

394

Andreas Pfeifer

that, mainly due to the organically evolved characteristics of workplaces, formal definition and documentation was neglected. An early survey of the current situation ensures that the hidden characteristics of existing workplaces are also assimilated in the design of better workplaces. In this case, current workplaces are characterized by the type and number of locations and configurations, servers, applications, peripherals, modems, ad-hoc networks and Internet connections as well as other factors. The next step is concerned with the recording of user profiles. The design of workplaces can only take place with knowledge of the number of users, their data, their application requirement and their usage behavior (time, location and priorities). This means that, among other things, the number of users, the locations of workplaces, the type of workplaces, the working hours, the need to work at several locations, the demand for remote access and the applications and data are recorded to which the user must have access. Extended security aspects can, in part, considerably constrain use, flexibility and efficiency. Hence, these subjects are best determined in detail early in the design of the workplaces. Amongst other things, these include: multiple logins, session time-outs, complexity requirements and maximum validity of passwords, barring requirements of workplaces, restrictions of Internet access, monitoring of Internet use and restrictions of dial-in. The success of workplaces is clearly measured – in the face of previously agreed service levels. For example, the availability, response time, throughput, disk use, mailbox size, time which the helpdesk needs for the answering and solving of problems, planned downtimes and the number and type of breakdowns can be stated. Applications must be consolidated in the area of the data center. The databases must be rationalized and consolidated because of data strategy and legal requirements. The servers and storage can be consolidated on the basis of subsequent requirements. This leads to the change of requirements for storage space and enables the locations of data centers to be rationalized and consolidated for this reason. Consequently, a new architecture and a new TCO model (Total Cost of Ownership) can be constructed and implemented on the basis of information regarding the locations, hardware, required capacity and the costs. Cost-intensive, private network connections are migrating to more economic flexible IP VPNs. Voice and data traffic is transported on WAN & remote access connections to reduce high connection costs and to enable an agreed communication of the different communication channels. Dial-in

The evolution to real-time architecture

395

users are converted to secure and reliable Internet-based services, which allow broadband access. Small locations are connected with cheaper Internet-based VPNs instead of with expensive, private networks. In order to enable real-time communication of critical processes, the communication traffic is prioritized according to user, application and location. The LAN infrastructure is set up so it also enables the transmission of voice and video as well as the transmission of data. Locations which still use traditional telephone systems, are equipped with a LAN-based IP telephone system. Voice over IP is used ultimately to reduce charges and costs. Mobile solutions are set up for companies that use Wireless LAN. Alternative connections such as satellites or mobile phones are possible and the applications are adapted accordingly for use via mobile connections. Answering machines, E-Mails, instant messaging and paging are integrated in one Unified Messaging application. The user is given access to the messaging systems from any terminal (e.g. access to E-Mail via mobile phone). Universal telephones are used (e.g. mobile telephones and fixed line telephones). Development to “infrastructure on call”

With a standardized environment companies are well positioned to set up utility-based data centers, full mobility and IP (voice and data) services in the second phase. The objective of an “infrastructure on call” is the enabling of dynamic scaling of applications through virtualization and utilization management (see Fig. 84). Hal Yarbrough, CIO of Lego, the fourth largest toy manufacturer in the world backs ‘IT on call’. Every year, in the period before Christmas, the logistic and merchandizing systems are buzzing at Lego. In contrast, throughout the rest of the year Yarbrough has managed overcapacity up to now. Yarbrough states: “In the seasonal toy business, an IT infrastructure which is adapted to the business processes is crucial for competitiveness”.13 Even Continental AG backs ‘IT on call’. The tire business is subjected to strong cyclical fluctuations. Seasonal “huge peaks in the utilization of IT” are experienced in spring, autumn or winter, according to Schwefer, CIO at Continental. “On Demand Computing” is always worth it if the demand for IT resources is not plannable and a high degree of flexibility is required”, says Paul Schwefer14. In particular, ‘IT on call’ offers an advantage to companies whose demand for computer performance is subject to regular peaks such as Lego 13 14

See Reppesgaard/ Lixenfeld 2003 See Reppesgaard/ Lixenfeld 2003

396

Andreas Pfeifer

or Continental AG, or are even subject to sudden unexpected peaks. These include, for example, large retailers who carry out periodical analysis of purchasing behavior (e.g. the Christmas trade), Internet providers with strongly fluctuating demand (e.g. Amazon.com) or Pay TV providers (e.g. for large pay per view television events)

MACHINE 1

MACHINE 2

Virtual partial area 2 Application C Scaling 'up'

Virtual partial area 2 Application B

Virtual partial area 1 Application B

Virtual partial area 1 Application A Application A Component 1

Application A Component 2

Application A Application A Component 3

Scaling 'out'

Figure 84: Paths to the realization of scalability through virtualization and advanced utilization management

‘IT on call’ is of interest to medium and small companies where complex and costly applications are used which are too expensive to purchase or which would go beyond the capacity of existing computer platforms (e.g. Collaborative CAD). Furthermore, ‘IT on demand’ is of interest to small and medium companies who conduct regular marketing campaigns using standard CRM software or who want to implement ERP functionalities over the Internet. ‘IT on call’ doesn’t make sense for all applications, but considerable costs can be saved and increases in speed achieved for relevant applications. As a result, before the introduction of ‘IT on call’, it is essential to carry out a detailed check of which areas come into question. This requires the evaluation of current demand and the identification of the level of use and the costs of resources, the pattern of demand and the quantification of costs that arise if service levels are not achieved. Even future demand must

The evolution to real-time architecture

397

be forecast with regard to the company strategy and market environment. In order to define the boundaries of future demand, often scenarios for the entry of the best case and worst case environment conditions are suitable. On the basis of current and future demand, the proper yardstick can be selected to determine the use. Appropriate parameters for this are, for example, the number of CPUs, the amount of storage or the number of users. The capacity on demand does need to include all IT resources. It is possible to start on a small scale and to increase this over time. An ‘infrastructure on call’ allows high utilization and, in particular, flexible reaction to unexpected new requirements. One aspect of the IT infrastructure from external suppliers

The third phase of the infrastructure transformation makes it possible to completely exploit the advantages of Utility Computing. Utility Computing permits the user to use computer performance, storage and network capacity on call just like today’s electricity and water utilities.15 This is facilitated by a virtualized, optimized, dynamically scalable, completely automated and commonly used infrastructure. The further development of the model of Utility Computing will include the IT applications and business processes as well as the pure infrastructure. Even if Ultility Computing has some characteristics of the ASP model, the subject matter of Utility Computing is much more comprehensive. Utility Computing can be used in both in-house and outsourcing scenarios. Major IT infrastructure providers such as IBM and HP, for example, integrate hardware and software with demand-oriented services. These include provision, installation, configuration, maintenance and support, emergency repair and security. This approach can be seen as the answer to the commoditization of hardware and, consequently, lower margins on hardware sales. In addition to the data centers, demand-oriented offers also cover networks, desktops, laptops and PDAs. Companies that use Utility Computing can realize considerable advantages from the point of view of business structure and from the organizational and financial perspective. Utility Computing will provide the entire company with higher flexibility, which makes it possible to react more quickly to suddenly arising potential and to detect transitory opportunities or market chances. Utility Computing will support companies in greatly reducing the product introduction time for new IT applications and business processes. 15

See Hamm 2003.

398

Andreas Pfeifer

2004 Voice

2005

2006

2007

2008

Mobile Phone

Thin client/ wireless devices

Fixed Line Phone PDA

Workplace

Laptop

Data

Desktop

Compute on request Consolidated & standardized

Data center

LAN

COPPER /Fibre

WAN

Point to point

Utility data centre

Utility computing

Wireless

Network MPLS LVI

All IP Enterprise on demand(Voice& Data)

Figure 85: Chronological development of the infrastructure

Utility Computing will improve the efficiency of IT departments. Automation means that less time and costs are required for demand planning. Also fewer specialized resources are required because the complexity of the technology, which the IT department itself must support, is reduced. Utility Computing allows cost reductions, cost prevention and cost control. Costs for unused resources are avioded. Because the costs can be consolidated and converted into variable and sales oriented costs, the level of fixed costs can be reduced. The costs are therefore more transparent, the capacity planning is improved, the transparency of the use of resources increases and the capacities become flexible. In this way, companies which use Utility Computing are ideally equipped to meet the requirements a real-time enterprise places on its infrastructure. Figure 85 again provides an overview on the chronological development of the infrastructure.

The evolution to real-time architecture

399

Development of the IT infrastructure in the context of the whole company 2004 Strategy, Processes & Applications

2005

2006

2007

2008

Business & IT Strategy Processing of the Business Processes ‘On Request’ Application Portfolio

Real-Time Company Dynamically Provisioned Infrastructure ‘On Request’ Infrastructure Consolidation & Standardization

Infrastructure

Figure 86: The evolution of the IT infrastructure as part of the development of the whole company

Of necessity the development of the IT infrastructure must be viewed in the context of the development of the whole company to a real-time enterprise. While, on the one hand, the company strategy and the processes are adapted in order for the company to become a real-time enterprise, on the other hand the IT infrastructure is further developed to enable a realization of the strategy and real-time processes (see Fig. 86). Only simultaneous development of the strategy, the processes, the applications and the infrastructure sanctions successful evolution of the company of today into a functioning real-time enterprise of the future.16

16

See Ferguson et al. 2002.

400

Andreas Pfeifer

Outlook Today, the development of companies into “Real Time Enterprises” is no longer in question. This development represents a complex and long-term change as we have clearly seen in the previous chapters. However, no company can successfully proceed on the path to a real-time enterprise if this is not enabled by the corresponding development of the IT infrastructure. The management must therefore place the focus clearly on the evolution of the infrastructure. The three-phase model developed by Accenture facilitates this evolutionary path in a structured and objective-oriented way. If the infrastructure continues its proactive development, it will no longer represent such an obstacle to the evolution of the company into a real-time enterprise as it does today, but would adopt a clear supporting function.

Bibliography Brien, T (2003): Redefining high performance, in: Accenture Outlook Journal, 3, p.19–23. (www.accenture.com/Outlook) Curtis, G.A.; Page, S.D.; Kaltenmark, J.K. (2003), Information Technology: Thinking Bigger, in: Accenture Outlook Journal, May 2003, p. 9–15. Davenport, T.H.; Prusak, L.; Wilson, H.J. (2003): Who's Bringing You Hot Ideas (And How Are You Responding)?, in: Harvard Business Review, Feb 1st. Drobik, A.; Raskino, M.; Flint, D.; Austin, T.; MacDonald, N.; McGee, K. (2002): The Gartner Definition of Real-Time Enterprise, Gartner Research Note, October 1st, 2002. Ellermann, H. (2003): Schwerpunkt Data Warehousing: Mobile Intelligence bei O2, in: CIO, 6. Ferguson, G. T.; Gershman, A.V.; Hughes, L.P.; Taylor, S.J. (2002): Reality Online – The Business Impact of a Virtual World, in: Accenture Outlook Special Edition, September 2002. Flint, D. (2003): The Real-Time Enterprise: Key Issues for 2002, in Gartner Research Note, January 9th, 2003. Hamm, S. (2003): Tech Wave 1: Utility Computing, in: Business Week Online, August 25th, 2003. Linder, J. C.; Jarvenpaa, S.; Davenport T.H. (2003): Toward an Innovation Sourcing Strategy, in: Sloan Management Review, 44, 4, p. 43–49. Mattes, A.; Emmerson, B. (2003): 21st Century Communications – Was Manager über Kommunikationslösungen wissen müssen, Hoffmann und Campe Verlag, Hamburg.

The evolution to real-time architecture

401

O'Mahony, R.; Padmore, L.; Suh, B. (2003): The Innovator's Advantage – using innovation and technology to improve business performance, Accenture Study Raskino, M.; Kyte, A.; Flint, D.; Drobik, A. (2002): The RTE ’Cyclones’ Model Changes the View, Garnter Research Note, October 1st, 2002. Reppesgaard, L.; Lixenfeld, C. (2003): Flexibilisierung von Anwendungen und Ressourcen – IT on demand, in: CIO, 11. Roby, T. (2003): Web Services: Universal Integration Power Seamless, Long Sought After Business Services, Accenture Outlook Point of View, April 2003. Schmalzl, B. (2004), Arbeit und elektronische Kommunikation der Zukunft: Methoden und Fallstudien zur Optimierung der Arbeitsplatzgestaltung, SpringerVerlag, Berlin. Siegel, L. (2002): The Real Time Enterprise, in: The Economist, Jan 31st, p. 21–43.

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

Distributed mini-factory networks as a form of real-time enterprise: concept, flexibility potential and case studies*

Competition in the real-time economy The dynamics of market require manufacturers to be flexible. Currently, however, more flexibility can only be achieved with a considerable tradeoff in resource utilization. Under-utilization of resources in many companies presently lies at between 20–50%, resulting in massive cost disadvantages. This particularly applies to the field of machine and plant engineering. The main reason for poor resource utilization is that today’s value chains are geared toward product and unit quantities. To a large extent, flexibility and efficiency have only been optimized at the level of the individual enterprise. In part, this is due to product complexity and the shortening of delivery times in recent years by almost 50% (Eggers/Kinkel 2002). Since the potential within companies for rationalization and flexibility has largely been exhausted, the most promising avenue for responding to the need for further dynamics in the creation of value today is considered to be the reorganization of the entire value chain spanning company boundaries. At the center of attention is production that takes place in dynamic, quickly adaptable networks. Complementary and partially overlapping production competencies in a value-added network permit the metamorphosis from rigid value chains to highly dynamic networks (Zahn/Foschiani 2002). Interaction within dynamic production networks stabilizes and improves * This article draws largely on the initial results of the Collaborative Research Center 582, “Production of individualized products close to the market” (http://www.sfb582.de/), funded by the Deutsche Forschungsgemeinschaft (DFG – German Research Foundation) in Bonn.

404

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

utilization of existing production capacities. As a direct consequence, this permits a reduction in investments in the plant and equipment. The buzz word “real-time economy” describes management and planning practices of the company as a direct response to customer wishes, market demands and to the circumstances of external value partners (Reichwald/Piller 2003, p. 515f; The Economist 2002). Traditionally, supply and demand have been decoupled, an approach that leads to efficiency benefits from an internal point of view, but that from the perspective of value chains causes mounting adaptation costs due to the increased dynamics of demand. Thus, instead of decoupling supply and demand, output generation is to be triggered within a network in direct response to a specific market demand. An organizational form that is able to ideally adapt to a “real-time economy” is the “real-time enterprise”. An important feature of the real-time enterprise is its ability to flexibly link output generation to existing customer wishes. Specifically, this means: Large parts of the value chain are not activated until after a customer order has actually come in. Until now, lead times have at best been unaffected in an order handling network, although such networks typically tend to extend lead times because of the additional interfaces. The market, of course, is unwilling to accept the associated delivery delays and demands “real time”. Therefore, a further goal is implementing competency network structures that would make it possible to substantially reduce lead times to as little as one tenth of today’s values. This in turn would result in a dramatic reduction in the level of non-fixed assets in the network (see figure).

Initiation Traditional production

Weeks

Real-time enterprise

Days

Preparation

Execution

Days

Hours

Hours

Real-time

Figure 87: Goals of the real-time enterprise (source: WZL, RWTH Aachen)

Distributed mini-factory networks as a form of real-time enterprise

405

This type of modern management gives rise to new cost structures and cost reduction potential generated by a customer-specific and customerintegrated value creation (Piller/Möslein 2002; Reichwald/Piller 2002a). Where up to now discussion has primarily centered around procurementrelated approaches to supply chain management or “efficient consumer response”, which are likewise based on the potential offered by an intensive, informative and up-to-date network of interdependent value partners in real-time, this article will focus on the inclusion of customers in “realtime”. The basis for efficient and effective management in real-time is the improvement of the knowledge base, both from a procurement-related and customer-related perspective. In this regard, intensive discussions under the heading of “Customer Knowledge Management” have recently dealt with the broadening of information and knowledge management to include knowledge about customers and knowledge contributed by customers. While “knowledge about customers” tends to aim at the familiar approaches of market research and the evaluation of customer data by means of CRM activities (purchase histories, customer account cards, scanner data, etc.), “knowledge contributed by customers” refers to new techniques that actively include the customer in knowledge acquisition (see Piller/Stotko 2003; 2004). The approach is to directly involve customers in corporate output generation. The customer becomes an integral part of the dynamic value-added network. Mass customization is a comprehensive concept for the efficient integration of customers in flexible, inter-company value creation (for basic information, see Piller 2003; Pine 1993; Pine 1997; Victor et al. 1996; for current developments, see, e. g. Reichwald/Piller 2002b; Reichwald et al. 2003 and in particular the articles in Tseng/Piller 2003). The goal of mass customization is to create customized products and services at an efficiency equal to that of mass production. While scientific discussion and practical implementations of the concept initially focused on enhancing production flexibility by introducing modular product structures, the emphasis has now shifted to accessing customer knowledge through customer integration. Customer interaction, which automatically becomes necessary when customizing products or services (during the configuration phase), is a very valuable source of information on the customer and from the customer. Thus, this information is used not only in the creation of the respective customized product or service, but is also a contributing factor in the long-term improvement of output potential. In this regard, the necessity for and increased significance of customer

406

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

knowledge have led to a fundamental discussion of existing organizational structures for output generation (Reichwald/Piller 2003). Access to customer knowledge demands close customer relations. Dealing in real-time demands new organizational structures. Therefore, the focus of this article will be on a new organizational approach, namely that the prevailing division of tasks between centralized production and distributed sales is to be replaced by distributed mini-factory structures in the form of integrated sales, innovation, customer interaction and production units that are in close proximity to the customer. The aim is to achieve a more efficient customized production (efficient flexibility) in addition to improving access to customer knowledge. The mini-factory is where customer integration occurs, and it is here that real-time management takes place. Thus, the mini-factory network represents the ideal model of a real-time enterprise. In fact, futurologists have already projected the implementation of this organizational form. “Instead of new factories overseas, there will be an increase in independent mini-factories in an urban context thanks to innovative production engineering such as real-time manufacturing,” was the forecast that appeared in a current article for the year 2020 made by Burrmeister et al. (2003, p. 119). In the course of the present article we will show that mini-factories are not merely an idea for the future, but are already coming into existence today and have begun to develop their potential. After a short description of mass customization, we will elaborate on the concept of the mini-factory as a specific form of the new organizational structure required for mass customization. Then we will examine the advantages of such a structure with regard to increased flexibility and efficiency in the context of a real-time enterprise. Eventually, we will present several case studies that clearly illustrate the theory underlying the concepts and demonstrate that they are founded in reality.

Mass customization as a response to the real-time economy The term “mass customization” was coined by Davis and combines the inherently contradictory concepts of “mass production” and “customization”. Its objective is: “[...] the same large number of customers can be reached as in mass markets of the industrial economy, and simultaneously they can be treated individually as in the customized markets of pre-industrial economies” (Davis 1987, p. 169). To this end, the production of prod-

Distributed mini-factory networks as a form of real-time enterprise

407

ucts and services must meet the following conditions (Pine 1993, 1997; Piller 2003; Tseng/Piller 2003): • Customized products on offer answer to the differing needs of each individual customer. However, from the perspective of the product policy, the selection options are limited (unless the digitization of individual components or production processes permits a virtually unlimited adaptation without efficiency losses). The goal of production is to use stable processes to produce customized products. • A (relatively) large market is targeted that corresponds to the market segments that would have been targeted with traditional mass-produced goods. • The price for the customized goods is not so high as to lead to a change in the market segment, i. e. the same customers that previously would have purchased a comparable mass-produced article would now decide for the individualized mass customization article. There is no shift to a higher market segment as is usually the case for customized production. • The information that is acquired in the course of the customization process is used to build a lasting individual relationship with each and every customer. Until now, mass customization has usually been implemented centrally on the basis of mass production processes. This is in accordance with recommendations stemming from research and practical experience from the time when the production of customized products first became an issue at the beginning of the 1990’s. “Pine 1993 and Kotha 1995 both explore standard product producers integrating mass customization into their product lines. Pine states that mass production is the counter point for mass customization and that standardization of products is a starting point for the development of mass customization. Kotha explores the learning relationship between mass customized and standard product production at the same plant” (Duray 2002, p. 319). However, an examination of failed pioneers of mass customization has led to a rethinking of the concept of a centralized production of customized products (Piller/Ihl 2002). Traditionally, organizational changes by mass customization have been seen on the level of singular activities of the value chain, with the inter-functional division of tasks largely untouched. This meant, for example, that manufacturing would implement new technologies, production planning approaches and conveyor systems; sales would change a salesperson’s image to that of a consultant who responds to cus-

408

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

tomer needs using highly complex configuration systems. The division of tasks between production and sales, however, remained traditional. The only advancement was to be the use of new information systems to improve the interfaces for the purpose of meeting increased information-related demands. With all its advantages, mass customization nevertheless is subject to a number of problems (Piller/Ihl 2002) that, in our opinion, can at least in part be ascribed to existing organizational structures. For example, the abundance of information that must be exchanged between the customer and the provider for the purpose of customization often brings on uneasiness and frustration in the customer, a situation that could end in the discontinuation of the customer relationship (Franke/Piller 2003; Huffmann/Kahn 1998). Another weak spot of many concepts has turned out to be the poor market proximity exhibited by many companies. For too many goods, customization is based on cultural and regional distinctions or on certain sector structures. Many manufacturers of mass-produced goods long ago relinquished the notion of being globally active with a single product program. However, the same applies to modular product systems and selection options offered by mass customization companies, who generally need to become more market-oriented than they now are. Other points are excessive delivery times, high logistics costs and inadequate information for customers on order status. This analysis prompted the idea of abandoning the prevailing concept of a centralized production model and replacing it with a distributed model. The outcome is a vision of a market-oriented, efficient production of customized products in mini-factory structures. These structures are the focus of the Collaborative Research Center SFB 582, advanced by the Deutsche Forschungsgemeinschaft (DFG – German Research Foundation) at the Technische Universität München (see www.sfb582.de), on whose initial results this article is based. Below follows a discussion of the economic potential of the real-time production of customized products and services in mini-factory structures. As a comprehensive example, we have selected the customization of a cleaning robot, a choice that is also the development scenario at SFB 582. The robot can be customized either with regard to its physical characteristics (design, functionality, quality, etc.) or with respect to the accompanying services (e. g. cleaning services). For example, customers can fashion the cleaning robot according to their personal design preferences (e. g. “to match the color of the furnishings”), to the available space (e. g. “device is to be stored in a narrow space beneath the stairs”) or, above all, to func-

Distributed mini-factory networks as a form of real-time enterprise

409

tional requirements (e. g. type of floor, climate, cleaning program). The device can also be customized by selecting specific services, such as a cleaning crew responsible for the infrequent and strenuous task of cleaning difficult to reach areas (e. g. windows). In this way, a comprehensive cleaning solution (product service system) can be created by combining customized hardware and additional services.

Mini-factories as a real-time enterprise

Definition

In contrast to large factories with a “Tayloristic” style, mini-factories are able to respond flexibly to sales and production tasks because of the aggregation of a number of functions “under one roof ”. Mini-factories perform independently on the market and must be able to manage customer interaction both before and after the purchase while also producing customized solutions. To be able to accomplish these core tasks, a mini-factory is designed as a scalable, modular, geographically distributed unit that is networked with other units of this type. A mini-factory performs the entire scope of activities necessary to effectively serve the customer. This particularly includes designing the product with heavy customer involvement, supplying the product and providing customer service after the purchase. This means: mini-factories must be located in their customers’ vicinity and, in addition to sales activities, must also carry out repair and maintenance work on the purchased products and offer supplementary services (e. g. complementary cleaning services). In the operative area, i. e. when involving customers in the value-added process, a mini-factory must handle the following process steps (Figure 88): • Recording of customer specifications by sales personnel using tool kits (configurators), • Translation of customer specifications to customized product features (this may include a complete redesign of the product), • Customized production of the product requested by the customer and • Delivery of the product, customer service over the product life and initiation of a repeat sale.

410

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

Mini-factory Customer interaction

Customer request

Delivery and customer service

Production

Support of/by other mini-factories in the network as needed

Figure 88: Operative processes handled by a mini-factory

However, the range of tasks handled by a mini-factory goes still further: In addition to the operative activities involving the customer, the mini-factory is responsible for structuring potentials, an area that is unconnected with the customer (see Figure 89). “Structuring of potentials” refers to the designing of company workflows that are not related to customer requests (e. g. the design of modeling and analysis techniques used for customer integration). By contrast, “operative activities” refers to the execution of processes for the purpose of meeting individual customer requirements. A mini-factory has the capability of quickly reacting to customer feedback by adjusting strategic, tactical and operative potentials. In this way, it is able to modify the process design of production and interaction systems and the architecture of product and service design in order to gain a strong position in its particular market. In this regard, mass customization in mini-factory structures is primarily intended to improve access to customer knowledge and to use it purposefully, both for a current order and for improving overall company potential in the long-term. Strategic planning of potentials

Tactical planning of potentials

Operative planning of potentials

What products and services should be offered to the customer?

What number of variants should be offered to the customer?

How should customer interaction be structured to reach the strategic goals?

Process design of production and interaction systems Architecture of product and service design

Customer feedback

Figure 89: Planning the company potential in a mini-factory

Distributed mini-factory networks as a form of real-time enterprise

411

Differentiation from similar organizational forms

Unlike businesses in the skilled trades, which also feature many of the above-mentioned characteristics, the mini-factory is part of an enterprise that comprises of a network of mini-factories. The individual mini-factories within this network interact intensively with one another. The network between the mini-factories serves the purpose of exchanging knowledge, a process that is just as important for configuring customized products and services as it is for marketing and production. The exchanged information relates to the feasibility of the customer request, the ensuing costs, technical details, general experience and customer profiles (Reichwald et al. 2003, p. 56). If a customer requests an unusual specification, an inquiry within the mini-factory network for similar specifications can significantly increase solution efficiency. A conceivable situation would be a customer who contacts a mini-factory in Germany to design a cleaning robot for hot and humid environments (e. g. for cleaning saunas). An inquiry in the minifactory network might reveal that a similar order had once been filed in a Swedish mini-factory. Instead of “reinventing the wheel”, the mini-factory designer can present the German customer with the Swedish solution and quickly fulfill the customer’s request. Production specifications for the solution could likewise be passed along. Ideally, the Swedish mini-factory would already have gathered suggestions for improvement in the course of customer care following the sale of the product, and these suggestions could flow directly into the German product. Mini-factories also differ from traditional businesses in the skilled trades in that they have a central support unit. It supplies all mini-factories with standard components, basic product developments and employee training. Our mini-factory concept also differs from applications discussed in literature that come fairly close to our concept and are based on similar principles, usually at the production level (for an overview, see the table below. The first three concepts listed there, MIRS, PLUTO and Küche Direkt, will be described in detail in the last section of this article). The U. S. military’s concept of the mobile parts hospital (MPH) shares with our concept the principle of an efficient production of small batches and the transferability of solutions between different “markets”. The MPH aims to accelerate the supply of replacement parts in the field by manufacturing parts on location in a mini-factory: “The overall approach is to develop a mini-factory that can be deployed to the field to manufacture replacement parts so that military equipment can be quickly returned to fully operational combat ready status” (no author, 2003b).

412

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

This approach differs from our mini-factory primarily in that it doesn’t require involved customer interaction for product specification. All data required for replacement parts are transmitted to the mini-factory in the field via satellite from a database that contains the technical specifications for all parts in the deployed military apparatus. Thus, this version of a minifactory plainly shows the flexibility of this kind of production unit. Almost any replacement part can be promptly produced nearly anywhere in the world. This type of flexibility is also of great significance to replacement part businesses in the civil sector: it is precisely the quick supply of rarely needed replacement parts (e. g. ventilation fans for 20-year-old car models) that is the strength of mass customization in a mini-factory (Suomala et al. 2002). However, this production-related aspect will not be discussed further in this article since it deals with an order-based production of standard and not customized parts. Table 6: Defining characteristics of the SFB 582 mini-factory Concept of...

SFB 582

MIRS

PLUTO

Küche Direkt

MPH

IPA

...tools with which customers can construct their own designs

Customers can design a customized product from the ground up using CAD tool kits.

Not a main focus, but individual volume specifications can be made, such as replacement part requirements.

Not a main focus, but the entire range of variants can be produced.

Selection of a customized kitchen with the support of a salesperson.

Not a main focus

Not a main focus

...production facilities that permit the efficient production of batches of size 1

Development of several techniques (e.g. computersupported drive, droplet pressure, etc.).

Robot production. Complete break with existing production procedures.

Completely new procedure that eliminates the requirement for uninterrupted production.

Comprehensive implementation of customized production.

Agile manufacturing on location.

Not a main focus, possible for some applications.

...processes that allow the use of mass customization as a CRM instrument

The mini-factory provides aftersales customer care (economies of relationship).

Not a main focus

Not a main focus

Not a main focus

Not a main focus

Not a main focus

...structures for easy transfer to new markets and scales.

Modular structure at the process level permits scaling and installation at other locations.

Modular structure permits scaling and installation at other locations.

Not a main focus, but possible in principle.

Modular structure permits scaling and installation at other locations.

Main focus on mobile implementation, possible in principle.

Not a main focus, possible for some applications.

Another example of a mini-factory is that designed by the Fraunhofer Institute for Manufacturing Engineering and Automation (IPA), the “Advanced Modular Micro-Production System” (AMMS), which meets the requirements for system components with a modular and miniaturized design. Using standardized interfaces, small handling and process modules are mounted on a desktop platform in plug-and-play mode and combined to form a manufacturing system. From the range of modules available, the

Distributed mini-factory networks as a form of real-time enterprise

413

user can select the modules required for his or her specific application and assemble an AMMS production structure. The module interfaces have been simply designed so as to permit a rapid alteration of the manufacturing concept (Dobler/Malthan, no year). However, this type of mini-factory does not allow for the changed economic situation described above, but addresses the challenges in manufacturing that arise from the miniaturization of components. The increased miniaturization of mechanical and electronic components demands suitable manufacturing technologies characterized by the efficient manufacture of small lot sizes in a large number of variations, high changeover flexibility, low investment costs and small space and infrastructure requirements. However, this has no effect on customer interaction, an aspect that is essential to our concept of a mini-factory. Nevertheless, there is one feature that may prove interesting and could be implemented as a building block in the mini-factory for mass customization. The ability of the AMMS to generate a modular production design could be used to create a structure that enables the transfer of the potential of the mini-factory into new markets by scaling and copying mini-factory modules. Mini-factory structures as we envision them have already been realized at Schott Zwiesel, Pirelli and Küche Direkt (see also chapter “Mini-factories for market-oriented production”). These case studies are evidence of the efficiency advantages that can be gained from the production of customized products in distributed locations. For example, offering customized goods is more costly on account of the greater flexibility. This can be counterbalanced by the improved efficiency brought on by customer interaction and production. The MPH and AMMS examples described above will not be discussed in further detail, since they do not generate specific efficiency benefits in the production of customized products. They were included here merely to demonstrate the existence of smaller production plants in the sense of a mini-factory.

Economic potential of a mini-factory

Applying the mini-factory concept

A network of mini-factories is only one method of combining efficiency and flexibility in a real-time enterprise. There will continue to be tasks that are more suited to a traditional system in which there is a delay between

414

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

production and the customer request. Using a contingency approach, we will now identify the situations in which a mini-factory structure is advantageous over a centralized production. As situation variables, we will use the mini-factory tasks shown in Figure 90. The task characteristics will be the same as those selected in Picot et al. (2003): variability, specificity, frequency, structuring and similarity. Plotting the degree of these task characteristics against the functions involved in processing a mass customization order results in the structure shown in Figure 90. Centralized production

Mini-factory Variability

Specificity

Frequency

Structuring

Similarity

Customer interaction

Customer request

Production requirements

High variability and high specificity of customer interaction, customer requests, production requirements and customer care indicate that a mini-factory could be economically advantageous.

The frequent occurrence of structured and similar orders indicate that centralized production could be economically advantageous.

Customer care

Figure 90: Dimensions of customized production

This analysis demonstrates that a mini-factory is suitable for creating customized products and services if a company’s situation is characterized by a high degree of variability and specificity. This would cause the company to be highly insecure about being able to fulfill customer requests and its products would undergo significant depreciation if rejected by the customer. These are important requirements that arise in the environment of a real-time economy. When producing customized products, this situation usually arises when a company offers an unrestricted customization of its products that goes beyond a simple combination of modules. In the example of the customized cleaning robot, this would be the case if the housing style were to be designed entirely according to customer request without any module restrictions. In contrast, a strictly modular offer of customized products on the basis of fixed components would tend to favor production in a centralized factory.

Distributed mini-factory networks as a form of real-time enterprise

415

We will elaborate on this argumentation in the following two paragraphs, in which we discuss the specific potential benefits offered by outsourcing production into mini-factories. These advantages must balance out the additional cost of mass customization resulting from increased flexibility if one is to meet the objective defined for the concept (i. e. a level of efficiency comparable to that of mass production). Figure 91 provides an overview of the following argumentation. Interaction

Production

• Access to “sticky information”: cost reduction potential from customer care • Increase of acquisition potential: influence on customer preferences through customization • Repeat purchases: influence on the repeat purchase rate through close customer contact

• Flexibilization: cost reduction potential by “postponing” production • “Size advantage” for small businesses: specific advantages of mini-factories over large companies that result in cost reduction potentials (e. g. low investment requirements).

Advantages of a mini-factory with regard to fulfill customized orders with high variability and specificity

Figure 91: Advantages of a mini-factory with regard to interaction and production

Customer interaction in the real-time enterprise

Access to “sticky information”: cost reduction potential from customer care through “customer interaction”

The starting point of our argumentation is the necessity for better access to knowledge about customers and to knowledge contributed by customers as the basis for running a company in real-time. In this context, the proximity of a mini-factory to the market offers a cost reduction potential resulting from better access to customer knowledge. Customer interaction in a minifactory generates better access to “sticky information”. The stickiness of information is defined as follows: “We define the stickiness of a given unit of information in a given instance as the incremental expenditure required to transfer that unit of information to a specified locus in a form usable by

416

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

a given information seeker. When this cost is low, information stickiness is low; when it is high, stickiness is high” (von Hippel 1994, p. 430). One possibility of receiving information on what the customer really wants is to employ the instruments of market research. However, especially in the area of customized products, these methods are usually inadequate. “Need information is very complex, and conventional market research techniques only skim the surface” (von Hippel 2001, p. 247). Traditional market research methods commonly only assess the current situation and often do not contribute to the correct assessment of future customer requirements. The resulting strategic gap between built-up output potential and requested output and the ad hoc measures needed for closing this gap are a significant cost driver in companies that simply are not able to collect the right customer information in real-time. Customer interaction in a mini-factory offers a new approach to understanding customers and recognizing their future needs. Access to this information during order handling lowers access costs to “sticky information” and thus contributes to efficiency improvements in market research. This can lead to a reduction in transaction costs in a mini-factory in the areas of initiation, negotiation and handling of subsequent orders. What’s more, costs arising from the multiple iteration loops necessary to accurately record customer wishes and translate them into customized product features are lowered through the greater accessibility to “sticky information” in a mini-factory. “Sticky information” is used at the various levels at which the potential benefits of mass customization are planned, as introduced above in Figure 91. In our view, the cost of access to “sticky information” is easier to reduce in a mini-factory than in a centralized production with distributed sales. The reason for this lies in the proximity of sales (e. g. show rooms) and production, which makes it easier to reduce the number of iteration loops mentioned above by clearing up questions of feasibility directly with production, which is also located within the mini-factory. In addition, customers will be more likely to come up with additional requests if they have a better understanding of the mini-factory’s output potential (see “Repeat purchases”). Increasing the acquisitive potential: influence on customer preferences through customization

A fundamental advantage of mass customization is that, in the eyes of the customer, customized products are always of better quality than compara-

Distributed mini-factory networks as a form of real-time enterprise

417

ble made-to-stock products. This leads to greater willingness to pay a higher price, an advantage the suppliers can skim off of customized solutions (Chamberlin 1962). A company that is in the position of offering customized products is in a “near monopolistic” (Weigand/Lehmann 1997) state since the offered products, at least on the surface, cannot be compared with products from the competition. The customer is willing to reimburse this uniqueness by paying a higher price, and in the extreme case, the supplier is able to skim off this benefit from each individual customer (see e. g. Skiera 2003). However, to be able to take advantage of this potential, any insecurities remaining with the customer must first be overcome. From the customer’s point of view, mass customization is completely different from purchasing a mass product. Mass customization is complex, obscure and risky (Huffmann/Kahn 1998; Zipkin 2001). Many customers do not have sufficient knowledge to define a product specification that meets their requirements. The result is not only a considerable amount of time spent by the provider in defining the specification, but also increased insecurity on the part of the customer. The situation is more pronounced the newer and the more customized the product is. This puts companies to the difficult task of offering their customers a large variety of options while at the same time taking suitable measures to offer assistance in selecting the correct customized product. Only then will it be likely that the offer of customized products will lead to a profitable business model. “If customers become frustrated or dissatisfied with the complexity, a [...] customization strategy obviously would not be a competitive advantage [...]” (Huffmann/Kahn 1998, p. 492). We argue that a mini-factory is better suited to overcoming customer insecurity than comparable sales concepts from a supplier with a centralized production. First, there is the emotional attachment that customers feel toward a mini-factory in their own surroundings. Secondly, a mini-factory makes it possible to assist customers with the configuration procedure to the extent that each customer requires. But the tour through the configuration system not only provides technical support in finding an appropriate specification, it also represents a special shopping experience. Empirical studies have shown that the perceived product satisfaction in mass customization is strongly correlated to the satisfaction experienced during the purchasing process (Franke/Piller 2003). For many customers, their participation in designing a customized product is a special experience. This experience is further enhanced in a mini-factory through its direct access to the customer and the customer’s active participation in product development and production. This can markedly increase customer identification

418

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

and involvement with the final product. Consequently, it is the main sales task of the mini-factory to convey features that generate enthusiasm, thereby helping to fully tap the acquisition potential in customized products. In this way, the mini-factory differs greatly from a dealer, for example, who only sells customized products. Even a professionally organized “event shopping concept” cannot, in our view, generate the enthusiasm experienced by customers through active on-the-spot participation in the valueadded chain. Finally, a mini-factory makes it possible to exercise influence on customer preferences. For example, customized products can be supplemented with customized services such as maintenance and repair work or training sessions. Repeat purchases: influence on the repeat purchase rate through close customer care

Another method by which the mini-factory takes advantage of customer interaction to reduce costs is the use of mass customization as an instrument of relationship marketing. Many approaches of relationship marketing fail on account of the unwillingness of customers to allow themselves to be “milked” for information without receiving an adequate service in return. “Relationship Marketing as practiced today has not brought companies any closer to their customers. On the contrary, the gap has only widened” (Fournier et al. 1998, p. 108). Offering customized products to customers in mini-factories is an effective alternative, since the customized product represents a tangible motivation for the customer to participate in the measures used to promote customer loyalty. In this way, mass customization comes close to meeting the requirement for a value-generating reciprocal exchange. After all, by creating a customized design through mutual interaction (communication), relationship marketing aims to “integrate buyers into an exchange that is value-generating and lasting for both sides” (Wehrli/Krick 1998, p. 63). Using the knowledge gained during the customer relationship and the willingness of the customer to interact with the mini-factory even after the purchase has been completed (e. g. by giving feedback on whether the product meets expectations), a mini-factory is able to capitalize on the customer relationship to sell several products to the customer in repeat purchases. In the end, then, mini-factories meet the following frequent requirement: to abandon the one-sided focus on gaining new customers (and losing regular customers in the meantime) in favor of fostering a stronger loyalty of existing customers to the company (Diller/Müllner 1998, p. 1220). Collection

Distributed mini-factory networks as a form of real-time enterprise

419

and comparison of information on individual customers increase the information density mini-factories have on their market and promote targeted and successful market development (see also Peppers/Rogers 1997). New customers can be given better and more efficient assistance. For example, customers may be recommended a customized product variant that other customers with a similar profile had purchased in the past (“profiling”). In a mini-factory network, profiling can span across many mini-factories (see also the example of the cleaning robot for a sauna), thus raising its success rate. This close proximity to customers can in the end help to realize the cost reduction potentials that we refer to as “economies of relationship”. Economies of relationship are the cost reduction potentials that lie in the additional opportunities that a mini-factory offers for building customer loyalty. Greater customer loyalty leads to increased sales to existing customers and to cost reductions in meeting these customers’ wishes. For example, new sales can be generated by offering existing customers other products for purchase or by bridging the time between purchases with “intermediate sales” (e. g. repair, maintenance, replacement parts, etc....). The recommendations made to the customer are based on profile information stemming from customer interaction during previous mass customization purchases. This generates cost reduction potential because acquisition, configuration and customer retention costs for that customer are lower when repeat purchases are made (Piller/Stotko 2003, p. 215). Overall, this is how a real-time enterprise in the form of a mini-factory network can compensate the additional costs that result from greater flexibility and more intensive customer interaction (see figure 92).

• Additional costs • Increased complexity

Figure 92: Economic potential of a mini-factory

• Additional earnings • Cost reduction potential

420

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

Production in the real-time enterprise

Flexibilization: cost reduction potential by “postponing” production

The ability to postpone the adaptation of products to specific customer requirements to the latest possible stage in the value chain is considered to be one of the most important prerequisites for the successful implementation of mass customization. “The key to mass-customizing effectively is postponing the task of differentiating a product for a specific customer until the latest possible point in the supply network (a company's supply, manufacturing, and distribution chain)” (Feitzinger/Lee 1997, p. 116). The postponement of the final specifications can be with regard to aspects of design, time or location. In a mini-factory structure, postponement occurs by delaying the final product design. “Form [Design] postponement means that companies delay production, assembly, or even design until after customer orders have been received, which increases the ability to fine tune products to specific customer wishes.” (Hoeck van et al. 1998, p. 33). Postponement with regard to time and location concerns the movement of finished products in the distribution chain or of supplier parts in the supply chain. These two aspects of delay are not well-suited to increasing a company’s ability to offer customized products since they primarily relate to functions of distributors (bridging distance and time). Therefore, they will not be included in the discussion below. A mini-factory, however, provides a good opportunity for putting the design delay approach into practice. Particularly useful is the local aggregation of development, sales and production units, since cooperation of these departments is essential for the success of a mass customization strategy. “Customization involves an intimate connection between product design and manufacture [...]” (Spring/Dalrymple 2000, p. 445). To be precise, even product development is postponed in a mini-factory until the customer has asked for a product. With the mini-factory’s expertise in defining product concepts (e. g. the cleaning robot should be able to climb stairs) and implementing them in concrete product features (e. g. the cleaning robot will be equipped with a chain drive), product development can be delayed until the customer walks into the mini-factory, sits down with the construction engineer and begins designing the product. With respect to development expertise, the only thing that will already have been defined in a mini-factory before the customer is involved is the “solution space”, i. e. the scope within which customers may realize their individual

Distributed mini-factory networks as a form of real-time enterprise

421

preferences (von Hippel 2001, p. 251f). From this perspective, a mini-factory offers the best possible flexibility for responding to customer requests, in some cases even adapting the output potential of the mini-factory to fulfill customer wishes. In other words, if a “solution space” turns out to be insufficient, a mini-factory has the capability of adjusting it to match actual customer requests. This clearly distinguishes a mini-factory from a centralized production model whose output potential is difficult to adapt to individual customer requirements. Because centralized production is oriented toward frequent orders that are highly similar and highly structured (see Figure 90), adjusting workflows to the requirements of individual customers it more difficult than it is in a mini-factory. In our view, cost efficiency in the completion of orders with high variability and specificity is therefore easier to achieve in a mini-factory than via centralized production. Specific advantages of mini-factories over large companies that result in cost reduction potentials

In addition to the advantages arising from the implementation of postponement, a mini-factory offers benefits that are a direct result of its small size: All business is local. And that must include production. This is the only way that individualization can be managed on the many world markets. (Reuther 2000). This statement is based on the (dramatic) changes that general production conditions have undergone in the last years. They have resulted in a liberalized world trade that gives manufacturers the opportunity of building up proximity to the market and customers in terms of location, time and social aspects, leading to further competitive advantages that Porter summarizes as follows: “Paradoxically, the enduring competitive advantages in a global economy lie increasingly in local things – knowledge, relationships, and motivation that distant rivals cannot match” (Porter 1998, p. 77). In striving for market proximity, smaller businesses can capitalize on specific advantages they have over large companies. One important means of realizing these potentials is modern information and communication technology. By employing such technologies, small companies can flexibly generate customer benefits across the globe. Traditionally, the implementation of new communications technologies, especially of high-speed networks with high data volumes, was limited to large enterprises. Only these large enterprises were capable of carrying the fixed installation and training costs – costs that were prohibitively high for small businesses with relatively small capital stock.

422

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

However, recent technical advancements that are reflected in the rise of the World Wide Web (WWW), compatible standards for exchanging (CAD) data (e. g. STEP, XML, etc. ...) and sharp declines in hardware prices now also make it economically feasible for smaller businesses to employ electronic media technologies for worldwide collaboration. The improved communication technologies permit a physical-geographic relocation of flexible production cells from the central production location into the vicinity of the customer. This effectively lowers the entry barriers into new domestic and international markets faced by small businesses in the form of a minifactory, since customized goods can now be economically produced in small batch sizes outside of niche markets. It essentially heightens the competitive capacity of small businesses in a market environment that requires proximity to the customer and distributed decision-making structures. “The adaptation of information technology tends to decentralize the economy and to reduce the average firm size, even if the information technology lowers both internal and external coordination costs” (Jonscher 1994, p. 38). It is precisely this demand that a networked mini-factory structure can meet.

Examples of mini-factories for market-oriented production in practice Below, the concept of a mini-factory will be described in further detail using several real-life examples. The case studies presented here are based on on-site investigations by the authors in each of the companies, on interviews with management and on the evaluation of literature and other sources. Due to the very small number of real examples, these cases were selected on the basis of being able to access the necessary information. Schott-Zwiesel AG: Pluto (Production Logistics Under Target Group Optimization aspects)

A case study that comes very close to our vision of a mini-factory is the PLUTO project of Schott Zwiesel AG. Schott Zwiesel is a world leader in crystal drinking glasses for the gourmet dining establishment, bulk quantity and lifestyle markets. The motivation for Schott Zwiesel to begin its mini-factory project arose due to highly fluctuating sales figures among its wide product spectrum, which encompasses some 2,500 models. By implementing a network of mini-factories, the response to specific customer demands was moved as far downstream as possible.

Distributed mini-factory networks as a form of real-time enterprise

423

While worldwide sales of water, wine and champagne glasses as well as burgundy and bordeaux goblets are fairly high at 100,000 units per year, sales of “exotic glasses” such as liqueur and sherry glasses lie at less than 10,000 units per year (Kreiß 2001, p. 128). In particular, these wide variations in sales figures for individual models prohibits the economical production of less popular models by mass production, which can only take advantage of economies of scale at a minimum volume of approx 30 to 50 thousand pieces. The PLUTO project is intended to improve this situation by shifting production to customer-oriented “satellites”. Central

• Transportable semi-finished products • Mobile machines and systems

*) Called a “satellite” at Schott

Distributed

Distributed mini-factory*)

Individuallymanufactured customized product

Figure 93: Overall concept at Schott-Zwiesel-Pluto (source: Schott-Zwiesel)

At the root of the PLUTO project is a fundamental rethinking of the premises of the production line as it exists today. A glass production line consists of several consecutive process steps that build on each other to produce the final glass product and that add value to the glass melt without interruption until it the final product is completed (Figure 94). The finished glasses are transported to storage, sometimes in the vicinity of the customer, and remain there until an order from a distributor triggers shipping of the required glasses.

424

Ralf Reichwald, Christof M. Stotko, Frank T. Piller Production line today Site in Zwiesel

Raw materials

Melt

Hot production

Cooling conveyor

Finishing

Refining

Transport

Storage

Shipping

Customer

Shipping

Customer

PLUTO production line Site in Zwiesel

Melt

Production of semi-finished product

Storage of semi-finished product

Mini-factory close to customer

Reheating

Blowing

Pressing

Cooling conveyor

Figure 94: Production line today and as envisioned by PLUTO (source: Kreiß 2001, p. 128f)

The production model envisioned by the PLUTO project breaks with the traditional concepts of glass production. Thanks to new manufacturing technologies and innovations in the glass melt, the requirement for uninterrupted processing through to completion of the final product no longer applies. Instead, production at the central location is limited to the manufacture of semi-finished products (small disks the size of a hockey puck), which are then reheated in a mini-factory close to the customer and further processed to create the final product. This means, first, that production can respond flexibly to customer requests as they arise (sherry goblet or bordeaux glass). Second, transport costs are reduced significantly. Since shipping finished glasses to a warehouse close to the customer literally means “shipping air”, cargo space can be much more effectively used by shipping semi-finished products. The cost reduction potentials this brings about are easy to appreciate. At the same cost, significantly more semi-finished products can be shipped to distributed mini-factories than finished glasses can be shipped to distributed warehouses. Also, the cost of storing semi-finished products lies well below that of storing glasses. In addition, the value of semi-finished products lies well below that of finished glasses. This means that the opportunity costs for stock of finished products will be lower due to lost interest income. This aspect is of particular importance when serving overseas markets. Besides the positive aspect of reduced transportation costs, there is the added advantage of lower initial investments in a mini-factory. This ultimately permits an economical development of markets that previously lay outside of the company’s own business radius.

Distributed mini-factory networks as a form of real-time enterprise

Capacity on a Euro pallet with a stack height of 100 cm [pcs.] 9600

Cost of pallet storage space (1 m3) per finished glass and per semi-finished product [cents] 6

+ 1820 %

-95%

0,3

500 Finished glasses

425

Semi-finished products

Finished glasses

Semi-finished products

Figure 95: Comparison of warehousing and transport of finished glasses and semifinished products (source: Kreiß 2001, p. 129)

The vision of Schott Zwiesel also permits a postponement in the design, which earlier in this article was described as an important condition for the implementation of a mass customization strategy. This case study deviates from the vision of a mini-factory in only a few points. Probably the most important one is that customers cannot (yet) design the glass themselves, since the PLUTO mini-factory needs exactly the same tools for each glass model as are used in the central production plant. Small production volumes could, for example, become feasible by employing wooden tools. Generative procedures in tooling making, as are being researched in SFB 582, may be able to close this gap. As for customer participation in designing the glass, this is easy to achieve using suitable tool kits for calculating volumes and stability. Pirelli: MIRS (Modular Integrated Robotized System)

While the PLUTO project is still in the conceptual phase and has not yet been introduced on the market, the following concept for a mini-factory has already proved its worth. Pirelli is a globally-active, international enterprise involved in the business areas of tires, energy and telecommunication cables. Pirelli concentrates on these key markets and is a leader in the industry and a prime innovator. It is one of the largest global suppliers of tires to major car manufacturers. The MIRS system used by Pirelli makes the production of a single tire economical. A robot-supported factory of this type can be accommodated in a space of approx. 350 m2. For Pirelli, the necessity of implementing a strategy of this nature arose from the fact that the prices that customers were willing to pay for tires in

426

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

the after market had fallen considerably over the last several years (original equipment manufacturers had never been very willing to pay). This low willingness to pay is accompanied by the markedly increased lifetime for today’s tires, which results in still another loss in turnover. While the average tire performance in the 1980’s was approx. 45,000 km, today’s tires reach almost 70,000 km. In this type of market environment, a cost-minimizing production is needed to remain competitive. That is exactly what Pirelli managed to accomplish with the MIRS program. Often, the cornerstone for the implementation of a mini-factory is innovative manufacturing technology that breaks with the old familiar premises and makes customized production economically viable. This, too, Pirelli has accomplished. At a heretofore unknown speed, MIRS robots perform the entire production cycle of a tire. In contrast to traditional tire production, the robots accomplish this without interruption and without the intermediate transportation of semi-finished products and tires. Integrated software controls every production phase, such as robot motion, automatic material loading, selection of the tire size, selection of a ready-made drum, manufacturing of the tire, vulcanization of the tire and transport of the finished product. Thanks to this technological support of production, tires of greatly differing sizes can be manufactured in any sequence. This method of tire production is more efficient than in a comparable mass production plant, since a number of steps that add no value to the final product can be eliminated. Thus, for example, preparation time is well below that for mass production, as is the changeover time before manufacturing a different tire size. In addition, this system achieves entirely new levels of product quality, since circumstances that are detrimental to quality, such as process interruptions, human intervention and temperature fluctuations of the semi-finished products during transport and storage can be avoided, otherwise often leading to vulcanization defects due to an inhomogeneous temperature distribution. MIRS is also suitable for implementing a mass customization strategy because tire dimensions of any size can be manufactured consecutively. Thus, tire manufacture is no longer bound to the planning of large batches in the millions, but can now produce tires according to demand, with sizes ranging from those for small vehicles (e. g. Fiat Punto) to sports utility vehicles (e. g. Ford F-series). Like the mobile parts hospital described above, the Pirelli system is suitable for the production of replacement parts. Instead of manufacturing tire models that are seldom required in large batches and warehousing them, these slow-moving items can be economi-

Distributed mini-factory networks as a form of real-time enterprise

427

cally manufactured in the MIRS system after a customer order comes in. Required area**)[m2]

Preparation time [h] 48

5,6 -80%

-98%

1,1 1,2 Traditional*) MIRS

Minimum economical batch size [pc.] 3200

Traditional*) MIRS

Changeover time [min] 375

-88%

-95%

375 Traditional*) MIRS

20 Traditional*) MIRS

*) Basis of comparison: manufacturing unit for the annual production of 1 million tires of type V/Z **)From preparation to finish

Figure 96: Characteristic values of the MIRS system (no author, 2003a)

The market for replacement tires is enormous. In Germany, some 40 million tires are replaced every year (average tire lifetime of 4.5 years at 45 million passenger cars and wagons (Federal Motoring Authority of Germany, status 1.1.03)). The replacement tire market extends to literally thousands of tire sizes, as a result of the specifications of car manufacturers, who require new tires with special characteristics (with respect to the speed and load index) for almost every new vehicle model. All this requires is a customer interaction facility that is suited for querying tire data. In the extreme case, this could even take place on an automatic interface into which the customer enters the vehicle data found on the registration document and can then view a selection of suitable, legally and technically approved tires. A click on the required tire followed by the entry of the number of tires required, and the order is placed and quickly filled in the MIRS module (the manufacture of one tire takes three minutes.) The MIRS system at Pirelli exhibits the features of a mini-factory as we have defined it, since it is possible to scale the facilities at the existing location and copy them to another site. For example, in Breuberg, Hessia, Pirelli has a facility for manufacturing the tires mounted on the BMW Mini Cooper S. Since construction of this MIRS facility began in December 2000, 12 of the 13 planned modules have been put into operation. With this equipment, the plant in Breuberg will be able to produce 2 million tires annual-

428

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

ly. Breuberg is a copy of the first MIRS facility that Pirelli put into operation in the Bicocca plant close to Milan in July 2000 (Pirola 2001, p. 15). Küche Direkt

Another example for the implementation of the mini-factory concept is provided by Küche Direkt (Kornacher/Suwelack 2003). With respect to customer interaction and mass customization, this example comes closest to the mini-factory concept we have presented in this article. It takes advantage of direct customer interaction to configure customized kitchens, and makes use of innovative manufacturing methods that permit the production of customized kitchens at costs comparable to those of mass production. The business goal of Küche Direkt, a joint project of the Rudolf Systemmöbel furniture company, IMA engineering company and Suwelack business consultancy, is to offer customized kitchen furniture on the basis of a fully parameterized furniture program at a bulk price (Ikea price). Together with a trained consultant, customers design their own dream kitchen. To ensure that kitchens would be available quickly while still being affordable, a completely new process chain was developed. The kitchen is manufactured within just a few days and assembled on the customer’s site. Küche Direkt was designed as a “production franchise system” and encompasses the production and sales systems. The intention is for the franchisee to be able to concentrate on his or her strengths – planning, selling, producing and assembly – without having to have actually system know-how. Data maintenance, continued development of organization software, comprehensive marketing, procurement, product development, etc., are all carried out and advanced at the system headquarters. This division of tasks corresponds to the basic mini-factory structure presented in this article. A central unit performs the core support functions, while the mini-factory handles customers and customized production on location. The approach used by Küche Direkt makes use of the following principles of mass customization: • The kitchen fulfills customer wishes with regard to its appearance. In addition, the height of the counter can be adjusted to suit the customer and the width and depth of the standard cabinets can be fit to the space available. Also, compared to the classical range of options, there is now a much larger selection of colors, handles and styles at no extra charge. • Because the product architecture is fully parameterized, parts lists and production information can be generated at the touch of a button notwith-

Distributed mini-factory networks as a form of real-time enterprise

429

standing the large number of variants. • Currently, plastic-coated particleboard still dominates production. This has the advantage that it does not require additional surface treatment. The production methods used (see below) are suitable for all other materials as well (solid wood, three-layered plywood, veneered board, etc.). In standard models, the exterior design is also found in the cabinet interiors. Küche Direkt also follows the mini-factory approach described above when it comes to customer interaction: • A configurator is not only used to plan the kitchen but also to visualize the desired models. • Information flow from the laptop of the salesperson to the flexible production system is fully linked. Thus the manufacturer has direct access to the customer, i. e. the usual interface between retailer and manufacturer is eliminated. This has the advantage that there is no duplicate handling of orders, permitting a fast response time. In the production, Küche Direkt also follows the principles recommended for a mini-factory: • Production is designed as a scalable mini-factory that encompasses the entire scope of production engineering from unfinished board to finished kitchen in only a few integrated workstations. Production costs are lowered considerably by using a highly automated, flexible production system and a consistently simple and clearly structured basic product architecture. Moreover, the absence of warehousing and elimination of the risk of not being able to sell the product avoid additional cost drivers that are usual for mass production (for warehousing). Only two employees control and carry out the entire production (from initial cutting to final assembly). • Paperless production begins after the customer requirements have been defined and does away with intermediate storage. Lead times are therefore extremely short. The finished product can be delivered within 2 days (without appliances) if the customer selects a decor for which raw material is already present. • The “upgrading” of materials leads to a significant reduction in complexity. The furniture design was tailored to suit the machine concept with a view to making production as efficient as possible. For example, all parts are constructed from the same 19 mm particleboard – even for

430

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

back panels. This increases material costs but greatly reduces the complexity of production planning and cutting. In part, scraps can be used for production of the next order. • A customized production that begins with the raw materials results in enormous savings in logistics costs. Vendors are integrated in the customized manufacture of the product since they deliver the kitchen counter made to measure. The price of the individually produced kitchen is therefore not higher than that of an average kitchen sold by large furniture stores. What has not yet been fully implemented and up to now has only been implicitly used is the opportunity of gaining customer knowledge and the systematic use of such knowledge in the overall network of mini-factories. Missing are systems suitable for accomplishing this and organization guidelines. Thus, essential potential benefits are not being utilized, especially since the participating companies and franchisees often continue to offer “mass-like” series products. An information transfer of aggregated knowledge from the customized orders to the series products could provide those products with important information relevant to gaining a competitive advantage.

Summary The mini-factory model presented here is an example for the concept of a real-time enterprise. It opens up interesting perspectives for the efficient and flexible production of customized products. Due to their small investment requirements, mini-factories bear fewer risks when entering new markets or experimenting with new organizational forms. Entering and exiting markets is easier to accomplish than with a large, centralized production structure. This is because the processes within a mini-factory – from order acceptance to order processing – are all modularly structured. Since such processes are scalable, the capacities of existing mini-factories are easier to adapt to market requirements in “real-time”. Thus, the mini-factory is in accordance with the view that a factory should be “living and breathing”. As well, modular processes make it possible to transfer (copy) proven concepts to other markets and to thereby gain access to new markets with low initial investments. The mini-factory is particularly well suited for the production and sale of customized products, since the sales and production units in a mini-factory

Distributed mini-factory networks as a form of real-time enterprise

431

are all joined together “under one roof ”. This serves to raise customer confidence and enthusiasm, and to meet in real-time even those individual requests that require a complete reconstruction of the product. The minifactory concept is made possible by technical innovations in the area of production that permit a break with the batch-based manufacturing technologies employed until now. The case studies of Schott Zwiesel, Pirelli and Küche Direkt demonstrate that the overall vision of the mini-factory is a viable alternative even today under the right economic conditions. The future of the real-time enterprise has already begun.

Bibliography Blackmore, J.; Kendrick, T. (2002): Ten things you really need to know about your customers, in: Management Services, 46, 10, p. 14–16. Burrmeister, K.; Neef, A.; Schulz-Montag, B.; Steinmüller, K. (2003): Deutschland und Europa 2020. Ein Zukunfts-Szenario, in: Volume One, 5, 10, p. 108–121. Chamberlin, E. H. (1962): The Theory of monopolistic competition : a re-orientation of the theory of value, Chambridge, Mass. Davis, S. (1987): Future Perfect, New York. Diller, H.; Müllner, M. (1998): Kundenbindungsmanagement, in: Meyer, A. (Eds.): Handbuch Dienstleistungs-Marketing, Stuttgart, p. 1219–1240. Dobler, H.; Malthan, D. (o. J.): Minaturisierte Produktionssysteme für Laborprozesse und -produkte (AMMS), published by Schraft, Dieter; Westkämper, Engelbert. Fraunhofer Institut Produktionstechnik und Automatisierung (IPA, Fraunhofer Institute for Manufacturing Engineering and Automation). Duray, R. (2002): Mass Customization origins: mass or custom manufacturing?, in: International Journal of Operations & Production Management, 22, 3, p. 314–328. Eggers, T.; Kinkel, S. (2002): Produktionsnetzwerke im Verarbeitenden Gewerbe: Erfolgversprechend aber selten genutzt, in: Zeitschrift für wirtschaftliche Fertigung (ZWF), 97, 3, p. 435–440. Feitzinger, E.; Lee, H. L. (1997): Mass Customization at Hewlett-Packard: the power of postponement, in: Harvard Business Review, 75, 1, p. 116–121. Fournier, S.; Dobscha, S.; Mick, D. G. (1998): Beziehungsmarketing: Des Guten zuviel für die Stammkäufer, in: Harvard Business Manager, 20, 3, p. 101–108. Franke, N.; Piller, F. T. (2003): Key research Issues in User Interaction with Configuration Toolkits, in: International Journal of Technology Management (IJTM), 26, 5/6, p. 578–599. Hippel, E. v. (1994): Sticky information and the locus of problem solving, in: Management Science, 40, 4, p. 429–439. Hippel, E. v. (2001): Perspective: User toolkits for innovation, in: Journal of Product Innovation Management, 18, 4, p. 247–257.

432

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

Hoeck van, R. I.; Commandeur, H. R.; Vos, B. (1998): Reconfiguring logistics systems through postponement strategies, in: Journal of Business Logistics, 19, 1, p. 33–54. Huffmann, C.; Kahn, B. E. (1998): Variety for sale – Mass Customization or Mass Confusion?, in: Journal of Retailing, 74, 4, p. 491 – 513. Jonscher, C. (1994): An Economic Study of the Information Technology Revolution, in: Allen, T. J.; Morton, M. S. S. (Eds.): Information Technology and the Corporation of the 1990's, New York; Oxford, p. 5–42. Kornacher, N.; Suwelack, C. (2003): Die individuelle Küchenproduktion zum Massenpreis, in: Piller, F. T.; Stotko, C. M. (Eds.): Mass Customization und Kundenintegration. Neue Wege zum innovativen Produkt, Düsseldorf (modularization chapter on the book in the Internet: www.symposion.de/masscustom). Kotha, S. (1995): Mass customization: implementing the emerging paradigm for competitive advantage, in: Strategic Management Journal, 16, Summer 1995 (special issue), p. 21–42. Kreiß, E. (2001): Von einem Businessmodell zu einer Demonstrationsanlage, in: Konferenzband Dritte deutsche Tagung zur Mass Customization, published by Frank Piller; Michael Schenk; Ralph Seelmann-Eggebert, 8. – 9.11.2001, Frankfurt a. M. Mills, P. K. (1985): The Control Mechanisms of Employees at the Encounter of Service Organizations, in: Czepiel, J. A.; Solomon, M. R.; Suprenant, C. F. (Eds.): The Service Encounter. Managing Employee/Customer Interaction in Service Businesses, Cambridge, MA, p. 163–177. No author (2003a): MIRS innovation, Pirelli, http://www.pirelli.com/ en_42/this_is_pirelli/ innovation/tyres/mirs_innovation.jhtml, edition: 10.4.03. No author (2003b): Mobile Parts Hospital, U. S. Army Tank-automotive and Armamants Command (TACOM), http://www.mobilepartshospital.com/welcome/, edition: 9.5.03. Peppers, D.; Rogers, M. (1997): Enterprise one to one, New York et. al. Picot, A.; Reichwald, R.; Wiegand, R. T. (2003): Die grenzenlose Unternehmung. Information, Organisation und Management, 5th edition., Wiesbaden. Piller, F. T. (2003): Mass Customization, 3th edition, Wiesbaden. Piller, F. T.; Ihl, J. C. (2002): Mass Customization ohne Mythos: Warum viele Unternehmen trotz der Nutzenpotentiale kundenindividueller Massenproduktion an der Umsetzung scheitern, in: io new management, 71, 10, p. 16–30. Piller, F. T.; Möslein, K. (2002): From Economies of scale towards economies of customer integration. Value creation in mass customization based electronic commerce, in: Konferenzband eReality: Constructing the eEconomy. Fifteenth Bled Electronic Commerce Conference, published by Claudia Loebbecke et al., 17.–19.6.2002, Bled, Slovenia. Piller, F. T.; Stotko, C. M. (2004): Mass Customization und Kundenintegration. Neue Wege zum innovativen Produkt (Spezialausgabe), in: Albers, S.; Haßmann, V.; Tomczak, T. (Eds.): Verkauf: Kundenmanagement, Vertriebs-

Distributed mini-factory networks as a form of real-time enterprise

433

steuerung, E-Commerce (loose-leaf collection), Düsseldorf. Piller, F. T.; Stotko, C. M. (Eds.) (2003): Mass Customization und Kundenintegration. Neue Wege zum innovativen Produkt, Düsseldorf. Pine, B. J. (1993): Mass Customization. The New Frontier in Business Competition, Boston. Pine, B. J. (1997): Mass Customization – the new imperative in business, foreword to : Anderson, D.: Agile product development for mass customization, Chicago, p. 3–24. Pirola, D. (2001): MIRS is running fast. E-Manufacturing Engages the Fifth Gear, in: World. The quarterly magazine for Pirelli's management throughout the world, 26, p. 15–16. Porter, M. E. (1980): Competitive strategy : techniques for analyzing industries and competitors, New York. Porter, M. E. (1998): Clusters and the new economics of competition, in: Harvard Business Review, 76, 6, p. 77–90. Reichwald, R.; Piller, F. T. (2002a): Der Kunde als Wertschöpfungspartner: Formen und Prinzipien, in: Albach, H. et al. (Eds.): Wertschöpfungsmanagement als Kernkompetenz, Wiesbaden, p. 27–52. Reichwald, R.; Piller, F. T. (2002b): Mass Customization-Konzepte im Electronic Business, in: Weiber, R. (Eds.): Handbuch Electronic Business, 2nd edition, Wiesbaden, p. 469–494. Reichwald, R.; Piller, F. T. (2003): Von Massenproduktion zu Co-Produktion, in: Wirtschaftsinformatik, 45, 5, p. 515–519. Reichwald, R.; Piller, F. T.; Jäger, S.; Zanner, S. (2003): Economic Evaluation of Mini-Plants for Mass Customization, in: Tseng, M.; Piller, F. T. (Eds.): The Customer Centric Enterprise. Advances in Mass Customization and Personalization, Berlin, Heidelberg, New York, p. 51–69. Reichwald, R.; Piller, F. T.; Stotko, C. M.; Ihl, J. C. (2002): Leveraging customer interaction to improve the bottom line of mass customization concepts in the field of complex consumer goods, in: conference volume of the Third International NAISO Symposium on Engineering of Intelligent Systems (24. – 27.9.2002), ICSC-NAISO Academic Press, Malaga. Reuther, E. (2000): Maschinen- und Anlagenbau – Kern der Produktionsgesellschaft, in: Konferenzband Karlsruher Arbeitsgespräche, published by FZK, 2. – 3.3.2000, Karlsruhe. Riemer, K. (2003): Grundzüge der internetbasierten Personalisierung, in: Piller, F. T.; Stotko, C. M. (Eds.): Mass Customization und Kundenintegration: Neue Wege zum innovativen Produkt, Düsseldorf (modularization chapter on the book in the Internet: www.symposion.de/masscustom). Skiera, B. (2003): Individuelle Preisbildung bei individualisierten Produkten, in: Piller, F. T.; Stotko, C. M. (Eds.): Mass Customization und Kundenintegration. Neue Wege zum innovativen Produkt, Düsseldorf (modularization chapter on the book in the Internet: www.symposion.de/masscustom). Spring, M.; Dalrymple, J. F. (2000): Product customisation and manufacturing

434

Ralf Reichwald, Christof M. Stotko, Frank T. Piller

strategy, in: International Journal of Operations & Production Management, 20, 4, p. 441–467. Suomala, P.; Sievänen, M.; Paranko, J. (2002): The effects of customization on spare part business: A case study in the metal industry, in: International Journal of Production Economics, 79, 1, p. 57–66. The Economist (2002): The now economy, 264, 4 (2.2.2002). Tseng, M.; Piller, F. T. (Eds.) (2003): The Customer Centric Enterprise: Advances in Mass Customization and Personalization, New York; Berlin. Victor, B.; Pine, B. J. I.; Boynton, A. C. (1996): Aligning IT with new competitive strategies, in: Luftman, J. (Eds.): Competing in the information age, New York, Oxford, p. 73–95. Wehrli, H. P.; Krick, M. (1998): Mit strategischen Netzwerken Kundennähe realisieren, in: Absatzwirtschaft, 41, 1, p. 62–68. Weigand, J.; Lehmann, E. (1997): Produktdifferenzierung, in: Wirtschaftswissenschaftliches Studium (WiSt), 26, 9, p. 477–480. Woodruff, R. B. (1997): Customer value: The next source for competitive advantage, in: Academy of Marketing Science. Journal, 25, 2, p. 139–153. Zahn, E.; Foschiani, S. (2002): Wertgenerierung in Netzwerken, in: Albach, H.; Kaluza, B.; Kersten, W. (eds.): Wertschöpfungsmanagement als Kernkompetenz, Wiesbaden, p. 265–275. Zipkin, P. (2001): The limits of mass customization, in: Sloan Management Review, 42, 3, p. 81–87.

Frank E. Gillett, Thomas Mendel

Organic IT: cut IT costs, speed up business The architecture behind grid, utility, on demand, and adaptive technologies

Organic IT can deliver big IT cost savings — and business gains Corporate computing is at the dawn of a new revolution that will transform how IT infrastructure is built, what it costs, and how it operates. Forrester believes that Organic IT will be the third major revolution in data center architecture, after the mainframe and client/server (see endnote 1). Over the last two years, eight major vendors — CA, Dell, EMC, HP, IBM, Microsoft, Sun and VERITAS — have announced initiatives under various names that aim at Organic IT and more, along with countless smaller vendors. Why should firms care? Because the Organic IT revolution in data center architecture addresses IT's three fundamental problems — wasteful technology, laborious processes, and rigid business capabilities — by (see Table 7): • Retrofitting and evolving existing technology. Organic IT is not a rip and replace revolution that requires ignoring or discarding existing assets. The technologies are extensions of today's distributed systems and often can be retrofitted to leading products already in the data center. • Offering immediate big savings — with more to come over the rest of the decade. In Forrester's experience, most large firms find that they can achieve seven-figure savings within a year or two with emerging Organic IT technologies. Some firms find the savings in software and servers, others in storage, networks, or systems management. These are real cash savings from avoiding new technology buys or staff hires, not

436

Frank E. Gillett, Thomas Mendel

ephemeral savings from faster results. And there's more coming: New Organic IT technologies and improvements will drive increased savings through at least 2010. • Helping all firms, whether they keep IT in-house or outsource. Organic IT data center architecture will benefit firms that outsource, too. It will enable outsourcers to offer lower costs and faster response times to customers, especially those that demand an early share of the benefits from this emerging technology. • Enabling firms to move to a leaner, faster business strategy: Organic Business. Organic IT also helps IT accelerate the response to new business needs because it makes it much easier to update software and reconfigure data center assets. At first, this will result in faster company mergers and quicker responses to competitive threats. But Forrester believes that ultimately, Organic IT will open the door for a new business strategy — Organic Business — that helps firms to interconnect their business processes with suppliers, partners, and customers.(see endnote 2)

But industry ideas and offerings are confusing Executives are skeptical of this revolution in data center architecture — when offered a range of choices to describe how they think data center technology will evolve over the next two years, only 41% of execs chose radical or big improvement and 58% chose gradual improvement, or worse (see Figure 97). (see endnote 3) It's no surprise that many executives have low expectations because: • Public debate mixes three separate issues: architecture, outsourcing, and pricing. Firms really have three different choices to make: 1) how to architect the future data center; 2) whether to outsource part or all of IT; and 3) how to pay for technology — buy, rent, or pay per use.(see endnote 4) These three choices should really be made independently, with separate criteria. This report focuses on the first question — what's the right architecture for the future data center? • There's no common name or definition for this new technology — or related ideas. Vendors, press, and analysts are using many different terms to describe this new data center architecture, sometimes embedding it in bigger ideas that include outsourcing or business transformation, like HP's Adaptive Enterprise and IBM's On Demand (see Table 8). More narrow concepts — like virtualization or grid computing for

Organic IT

437

Table 7: Today, IT executives must endure three fundamental technology flaws

Problem

Wasteful

Examples

Consequence

Incompatible software interfaces in Windows, Java, and ISV apps like SAP, PeopleSoft, and Oracle Dedicated servers with one app and low utilization rates of 20% or less

Expensive EAI products and skills

Direct-attached disks and isolated disk arrays that are over-allocated and filled with obsolete files Fractured networks that can’t share gear and bandwidth across the LAN, WAN, and Internet. Manual recabling for changes.

80% of storage capacity wasted

Lengthy and uncoordinated software projects with lots of skilled engineers

High labor costs -- or no custom apps

Server builds and upgrades require CD shuffling and lots of command line work

Delays and cost -- or insecure servers

Storage management is split across many different administrative groups and tools Network reconfigurations require recabling and command line work Reactive management software that sends scads of alerts for each event but can’t pinpoint root cause and implement fixes

Hours of work and app downtime Long hours -- or insecure apps Lengthy troubleshooting sessions

Implementing a new pricing scheme takes months of coding Creating a new product requires extensive software integration Connecting to partners is delayed by WAN links and proprietary data formats

Competitor gains market share Product offering is too narrow Few or no links to partners

Finding technology problems in a business process takes hours Adding capacity takes weeks because of complex servers, storage, OS’, and apps

Customers buy from competitors Frustrated sta. and customers

Automating business processes requires implementing new systems and software, which takes too long

High business labor costs

80% of server capacity wasted

Higher service and gear costs

Laborious

Rigid

438

Frank E. Gillett, Thomas Mendel

numerically intensive problems — have lost their specific meaning as they have been applied to the broader concept of data center architecture.(see endnote 5) And it doesn't get any easier when firms look at the technology that vendors are offering. Firms find the solutions offered by vendors confusing because: • The market lacks standards. Other than Web services, standards for the new data center architecture are just getting started.(see endnote 6) The Storage Networking Industry Association's (SNIA) SMI-S standard for storage management is getting traction but server management standards are just getting started in Globus' Open Grid Services Architecture (OGSA) effort, the Distributed Management Task Force (DMTF), and the Organization for the Advancement of Structured Information Standards (OASIS) standards bodies. • Vendors are still assembling their offerings. Small vendors like Acopia Networks and Sychron offer targeted products, if firms will buy from small players. But big vendors initially pitched very expensive offerings like HP's UDC, services like IBM's Global Services, or features like Sun's Dynamic System Domains. Only now are big vendors coming out with entry-level offerings of interest to many firms, like IBM Tivoli Intelligent ThinkDynamic Orchestrator and VMware's ESX. Bottom line? There is no easy way for firms to compare products and overall offerings.

Figure 97: IT executives expect only gradual infrastructure technology improvements

Organic IT

439

Table 8: there are many names and meanings for new technology strategies

Firms Should Buy Three Capabilities. . . Despite the confusion of issues, terms, standards, and offerings, the goal of next-generation architecture is now clear. It is Organic IT, which Forrester defines as: IT infrastructure that automatically shares and manages reliable virtualized software, processors, storage, and networks across all applications and

440

Frank E. Gillett, Thomas Mendel

business services. To get to Organic IT, firms must buy and implement data center technology that enables: • Virtualization for technology sharing and maximum utilization. Virtualization enables firms to uncouple logical units of usage, such as an operating system or a storage volume, from physical units of operation, such as a server or a disk. This allows firms to maximize utilization — and gain great flexibility in moving and managing assets. • Automation for higher labor efficiency. Tasks no longer require laborintensive CD shuffling and lots of cryptic command line instructions. Need to build another PeopleSoft server? Click a single button to: 1) deploy a copy of the server-based image; 2) configure the app; 3) attach to the right network virtual LAN (VLAN); and 4) mount the correct storage volumes. • Self-management for business flexibility and speed. The data center, from individual assets like servers to the data center itself, must be able to respond to changes without human intervention — or to ask for help from a human. For example, preset priorities will guide the allocation of assets, keeping the Web-based configuration engine running by repurposing idle accounting capacity. . . . And Implement Three Best Practices But firms can't just buy their way to Organic IT by retrofitting existing technology and upgrading to Organic-ready products — today's data centers are too complex and rigid. Firms must also apply these three best practices to technology selection and implementation: • Standardize and simplify to cut complexity, variety, and integration. People often think their business or process is unique — but it usually isn't. Use industry practices, off-the-shelf apps, products from leading vendors, and common standards. • Abstract to cut dependencies and hide complexity. Use abstraction to hide complexity behind simple interfaces for humans and other technologies. Shoot for an IT architecture that uses modularity and loose coupling as much as possible to get resiliency and flexibility. • Integrate to link isolated processes. Many applications and business processes can only connect through human intervention via phone, fax, and email. Organic IT requires products that have integration built in —

Organic IT

441

or that enable firms to easily build custom integration among apps, products, and processes.

Firms must upgrade five technologies In the original Organic IT report, Forrester focused only on the foundational innovations in four infrastructure categories that made Organic IT possible. Over the last two years, we've added a fifth infrastructure category, listed multiple innovations across the five categories, and identified key prerequisites necessary for Organic IT. To capture available benefits and move toward Organic IT over the next three years, firms must investigate and implement a combination of new and existing technologies to build new capabilities in five infrastructure categories (see Figure 98):

Figure 98: Organic IT requires five integrated sets of infrastructure capabilities

1. Software: Web services and composite applications enable business services. 2. Servers: Virtualized processing shares and maximizes computing capacity. 3. Storage: Automated management and virtualized disks optimize storage.

442

Frank E. Gillett, Thomas Mendel

4. Networks: Automated, virtualized LANs and WANs simplify and speed networks. 5. Management: The Fabric Operating System powers Organic Management.

Software: web services and composite apps enable business services Today's applications delay or prevent business changes with incompatible technologies and proprietary, expensive integration technologies. To get past these issues, firms will use standardized Internet-based technologies for building and connecting applications, processes, and businesses. The key software technologies for Organic IT are: • Web services to abstract and standardize software interfaces. Web services are standardized software interfaces developed since 2001 by the software industry, driven by the cooperation of IBM and Microsoft. They virtualize or hide proprietary interfaces behind simple, selfdescribing interfaces that can use several different popular Internet transport protocols.(see endnote 7) The result? Firms can connect applications together quickly and cheaply, even between companies, by using industry standard business formats like ACORD for insurance and FpML for financial derivatives. • Service-oriented architecture (SOA) for shared services. Web services are great — but each one needs capabilities for security, performance management, and high availability. Rather than hard-wire these capabilities into each Web service, firms should build a service-oriented architecture, which offers the capabilities as shared services across the environment.(see endnote 8) • Web services management for automation and self-management. Even with shared services in an SOA, firms will find it difficult to manage the modular Web services that connect across and outside the firm. Firms can look to startups like Systinet and AmberPoint, but systems management vendors are working hard to catch up.(see endnote 9) Firms must also work to add identity management capabilities to the SOA as part of the security infrastructure for Web services.(see endnote 10) • Composite applications to virtualize existing apps into new forms. The SOA, plus emerging Web services interfaces on packaged apps, will enable a new layer of applications that work across transactions, content,

Organic IT

443

and business analytics, which Forrester calls composite applications.(see endnote 11) Composite apps enable firms to automate processes that rely on human interaction and span traditional apps, especially those focused on strategy, planning, and optimization.

Servers: virtualized processing shares and maximizes computing capacity Today's "one app, one box" syndrome wastes 80% or more of today's server capacity. And server management is way too labor intensive, involving lots of CD shuffling and manual command line work. To gain big efficiencies and speed business response times, firms must implement: • Virtual servers. Today, firms can isolate applications by virtualizing servers in hardware or software. HP, IBM, and Sun offer hardware-based virtualization like HP's nPars, while Intel's Vanderpool project plans for chip-level virtualization.(see endnote 12) EMC/VMware and Microsoft's forthcoming Virtual Server offer software virtualization for Intel, while the Unix server vendors have offerings like HP's vPars and Sun's Solaris 10 N1 Grid Containers.(see endnote 13) • Server hardware management. The hardware vendors' low-level server management software — such as IBM Director, HP Systems Insight Manager, and Dell OpenManage — is great for remote control during low-level maintenance chores. But don't use them for automation scripts — instead treat them as a proprietary API or translator for the heterogeneous management automation discussed below. • Workload management — self-management within a server. For applications that can co-exist within the same OS instance, workload management (WLM) software can prioritize compute cycles for some apps over others. Today's best-known examples from HP and IBM are built for their proprietary Unix operating systems, but Forrester believes that WLM will evolve to work across the data center. (see endnote 14) • Processor grids — segmenting of compute problems onto multiple servers. Though "grid" is now casually used to refer to many forms of IT resource sharing, it does have a specific meaning. Some numerically intensive compute problems, such as financial modeling, scientific analysis, and engineering simulation, can be split across many servers. Though not applicable to most business apps, when appropriate it can speed results or cut costs dramatically.

444

Frank E. Gillett, Thomas Mendel

Storage: automated management and virtualized disks optimize storage Storage, like servers, wastes a lot of resources. Disk space is isolated, poorly utilized, not segmented into classes of services, filled with duplicates and obsolete data, and managed with multiple proprietary interfaces. And to make storage configuration changes, IT must stop business apps to remap apps to new storage locations. To make storage, Organic firms must: • Network all storage with NAS and SAN to share storage. Most firms continue to buy direct-attached storage because it's cheap and convenient — but it requires manual maintenance by server administrators and can't be shared to gain storage efficiencies. Networking all storage is a prerequisite — but not sufficient — for making storage Organic and enabling Organic Management. • Move toward storage virtualization. To simplify app and storage management, firms must move away from the current practice of accessing data by physical addresses.(see endnote 15) Firms should monitor storage virtualization technologies and plan to implement when ready. Heterogeneous storage virtualization, in hardware or software, can be bought from vendors like Cisco, FalconStor, IBM, and VERITAS. • Develop a strategy to optimize storage with information life cycle management. As storage infrastructures become increasingly networked and virtualized, firms should use information life cycle management (ILM) tools to further reduce costs and reduce legal risks. Storage veterans like EMC and IBM are developing ILM products that will take advantage of modern storage networks to purge obsolete files and move aging files off premium storage into secure archives.(see endnote 16) • Migrate to heterogeneous storage management software. Firms must stop depending on proprietary storage management that only manages one model or vendor's storage. Firms need SMI-S-compliant storage software that can do routine storage tasks across all storage gear, such as products from vendors like AppIQ, CreekPath, and Softek Storage Solutions.(see endnote 17)

Organic IT

445

Networks: automated, virtualized LANs and WANs simplify and speed networks To reallocate gear and setup new apps, firms are constantly rewiring the data center. Furthermore, to securely connect to other firms, IT must coordinate with network service providers, implement unique technology like Frame Relay, and pay service fees much higher than Internet rates. Both issues delay business improvements. To speed network changes, firms must implement: • LAN virtualization and automation. Firms must implement VLAN technology across the data center to get software control over network wiring — a prerequisite for enabling rapid network changes and Organic Management. As firms move to Organic IT, they'll need to be able to quickly reconfigure network resources like firewalls, VPNs, and load balancers. Vendors like Cisco, Inkra Networks, Nauticus Networks (recently acquired by Sun), and F5 Networks are working to automate network configuration by consolidating all this infrastructure into one intelligent, software-configurable box. • WAN virtualization with RAIL. WANs serve as private network links that guarantee security and performance, unlike the open Internet — but with much greater cost and rigidity. Forrester believes that some data centers can replace WAN links with a redundant array of Internet links (RAIL) that combine VPNs for security with route optimization across multiple Internet links to guarantee performance.(see endnote 18) This approach delivers good-enough performance at significantly lower costs than private WANs.

Management: the fabric operating system powers organic management Today's management software isn't ready for Organic IT — it can't deal with virtualized resources, rapid resource allocation, or coordination of resources across an entire app.(see endnote 19) IT is stuck managing manually and reacting to events, rather than preventing them. Firms must begin to plan for and implement the two elements of Organic Management (see Figure):

446

Frank E. Gillett, Thomas Mendel

Figure 99: Organic management requires a fabric OS and an orchestration engine

• The Fabric OS manages in seconds to minutes. The foundation of the Fabric OS starts with automated configuration management and provisioning software — firms are already implementing technology from vendors like BladeLogic, Opsware, and VIEO.(see endnote 20) These and other technologies will manage split-second resource allocation and workload management to ensure that the right apps get priority use of available resources. • The orchestration engine manages in minutes to months. The Organic orchestration engine watches resource utilization trends and plans

Organic IT

447

changes to resource allocation, such as queuing up extra capacity for the World Cup. But it isn't just operational — it must also support business service management, Forrester's vision for proactive service management.(see endnote 21) As firms gain comfort with this technology, they'll move to automated policy-based provisioning, like that in IBM Tivoli's Orchestrator.(see endnote 22)

Recommendations Firms Must update Standard practices • Rethink consolidation projects — bigger is not always better. Centralization helps Organic IT — but consolidating virtual machines onto big servers and storage arrays doesn't necessarily help. In fact, virtualization software combined with Fabric OS technologies for self-management will make racks of four-way servers competitive for consolidation in comparison with partitioned 16-plus CPU servers running lots of virtual machines. • Standardize procurement of new technology. Firms must stop buying technology project by project and start specifying standard configurations for any application. This means taking infrastructure selection out of the hands of project managers and putting it in the hands of IT architects.

What it means IT and business Units will ReOrganize • Firms must reorganize IT departments to get Organic IT success. Implementing Fabric OS automation won't work when management tasks are scattered across groups responsible for Unix, Windows, servers, storage, and networks. With servers and network links being changed constantly, firms must bring together infrastructure management while dispersing business process management to the business units. • Organic Business will drive business unit overhauls. As businesses move to embed their business services and processes into other firms, implementing Organic Business, they will find today's separation of corporate IT and business unit untenable. To succeed, business units will

448

Frank E. Gillett, Thomas Mendel

take on development, management, and operation of business services, but rely on corporate IT or outsourcers to operate the infrastructure that supports their business services.

Endnotes 1. The Organic IT idea is built on four innovations in abstraction of software, servers, storage, and networks that will fuse into a new data center architecture that will eventually replace client/server. Flexible modular service architecture is a similar idea developed independently that envisions the combination of Web services with automated IT infrastructure. See the April 22, 2002, Report "Organic IT," and see the November 26, 2002, Planning Assumption "Flexible Modular Service Architecture: Hardware and Software Converging to Flexible Service Orientation." 2. The real payoff of Organic IT will come when firms harness their technology horsepower and new Internet standards to bring their business services online in a strategy that Forrester calls Organic Business. An Organic Business will embed its business services in customers' and suppliers' operations under direct business control. The result? Sticky customers, efficient suppliers, and the rise of a business services Internet that overcomes location barriers to enable deeper business process outsourcing. See the March 18, 2004, Forrester Big Idea "Organic Business." 3. Of the 91 executives that Forrester surveyed in July 2003, only 24% believed that they had high awareness of, credibility for, and interest in leading Organic IT vendor initiatives. See the August 27, 2003, Brief "IT Execs Embrace Organic IT Initiatives." 4. Decision-making becomes much simpler when firms separate these three choices. And recent survey research by Forrester shows that firms expect to move toward pay-per-use pricing over the next few years. See the May 9, 2003, Brief "Simplifying Organic IT Decision-Making," and see the March 19, 2004, Trends "The Future Of Software Pricing." 5. Thirty-seven percent of firms are piloting, implementing, or rolling out some form of grid technologies. But 37% believe that grid means clustered computing, 20% think it is a confusing term with multiple meanings, and 15% don't know what it means. See the May 18, 2004, Trends "Grid Gets Big, But The Term Is Confusing." 6. Though standards for storage are emerging in SNIA, server automation standards are nascent. EDS and Opsware launched DCML (Data Center Markup Language), but Forrester believes that key standards will gel in the DMTF (Distributed Management Task Force), rather than DMCL.org or in the Globus Alliance's OGSA (Open Grid Standards Association) efforts. See the December 30, 2003, Brief "Organic IT Standards Gel In 2004 — In The DMTF." 7. Forrester believes that Web services will gain widespread adoption when firms

Organic IT

8.

9.

10.

11.

12.

13.

14.

15.

449

deliver Web services that solve clear business problems, make customers more productive, and can be used without modification by many customers. See the December 17, 2003, Brief "Ten Tips For Killer Web Services." If firms build Web services individually, they'll have to embed capabilities like security and manageability in each Web services, duplicating effort. Instead, firms should build a service-oriented architecture that offers security, manageability, and other standard infrastructure capabilities as a shared service for all Web services in the firm. See the December 12, 2002, Report "Road To A Service-Based Architecture." Widespread use of Web services will require new automated management capabilities for managing, monitoring, and securing Web services. See the June 1, 2003, Report "Mastering The Web Services Tier," and see the September 8, 2003, Brief "HP Kills Two Birds With Talking Blocks." As Web services embed transactions, firms must add capabilities to check and authenticate the identity of the entity behind the Web service request. See the September 24, 2003, Brief "Identity Management Splits, Users Gain Clarity." Packaged applications and custom apps will be combined with custom code via Web services and portals to create new applications and services without building new standalone applications. See the January 22, 2004, Planning Assumption "Packaged Composite Applications Emerge — Slowly." RISC servers have had virtualization built in for a long time — and Intel plans to build it in. See the October 30, 2002, Planning Assumption "Defining Server Virtualization," and see the September 19, 2003, Brief "Intel's Vanderpool Puts Virtual Machines In Silicon," and see the January 15, 2004, IdeaByte "Demystifying Virtualization Alternatives — HP's Solutions Are Best Example." VMware proved that software vendors and users would support and adopt software--based Intel server virtualization. To compete with VMware, Microsoft bought Connectix to get the foundation of its forthcoming Virtual Server product. See the June 19, 2003, Brief "Virtual Server Technology's Time Has Come" and see the August 20, 2003, IdeaByte "VMware — Extending the Virtualization Metaphor," and see the June 16, 2003, Planning Assumption "Recentralizing Server Sprawl Through VMware: From Best Strategies to Cost Savings," and see the March 29, 2004, Quick Take "Virtual Machines: Integration Matters." Workload management for individual servers must evolve to link to the emerging Fabric Operating System, so that management software can perform WLM across the data center and not just within each isolated server. See the December 22, 2003, Report "The Future Of Workload Management." Back to Text Today's storage technologies often link the application to the physical storage device. This approach requires application changes every time file locations change. In addition, server administrators must use each vendor's proprietary management software to administer each type of storage. These practices cause great rigidity and manual labor. Storage virtualization will help solve

450

Frank E. Gillett, Thomas Mendel

both problems. See the May 24, 2002, Report "Making Storage Organic." Back to Text 16. Information life-cycle management is software, and sometimes hardware solutions, that automatically implements firm policy on the storage class of service that each piece of data deserves through time. So files are automatically moved to the appropriate type of storage as they age and deleted at the end of their retention period. See the October 14, 2003, Brief "EMC Buys Documentum: Customers Gain Integration," and see the October 15, 2003, IdeaByte "EMC Acquires Documentum, Ushers in New Era in Enterprise Content Management," and see the December 19, 2003, Planning Assumption "Positioning Storage to Support Information Life-Cycle Management in a Regulated World." 17. Standards are just emerging now that will allow firms to move from proprietary storage management software and middleware to heterogeneous storage management based on standards. But firms will have to endure a migration period of software that mixes standards with support for proprietary storage APIs. See the March 15, 2004, Quick Take "Storage Management: Small Is Good" and see the March 23, 2004, Market Overview "Storage Management Gets Modular." 18. Route optimization technology is marketed as a means to accelerate outbound Web sites. But it can also be used to constantly choose the cheapest path for network traffic. See the October 23, 2002, Report "Saving Bandwidth Bucks With Route Control." 19. To build management software that is ready for Organic IT, vendors must build software that is: 1) federated, not hierarchical; 2) proactive, not reactive; and 3) integrated, not siloed. But it's not just up to the management software vendors: Hardware vendors must build support into the hardware to complement the middleware and orchestration engine of Organic Management software. See the December 31, 2002 Report "Managing Organic IT Infrastructure," and see the June 11, 2003 Report "Picking An Organic IT Management Vendor." 20. The Fabric OS consists of a service expediter, a workload manager, and a provisioning engine. See the September 24, 2003, Report "The Fabric Operating System." 21. Delivering business service management (BSM) requires the capabilities for service verification, rapid problem verification, failure prediction, and business impact assessment of infrastructure changes. Today's manual solutions for BSM must evolve to include automated asset discovery and dependency analysis and links to the Organic orchestration engine. See the September 15, 2003, Planning Assumption "Best Practices for Business Service Management," and see the March 25, 2004, Best Practices "Best Practices For Infrastructure Change Management." 22. IBM's acquisition of Think Dynamics was a brilliant move to add needed capabilities and gain an early start on policy-based automation. See the May 16, 2003, Brief "Small Buy, Big Impact: IBM Buys Think Dynamics," and see the

Organic IT

451

November 26, 2003, Planning Assumption "IBM Orchestrator — Changing The Dynamics of On-Demand Computing," and see the December 29, 2003, Brief "IBM's Orchestrator Pumps Up Organic IT."

Jörg Luther

Living and working in a global network

Wired life Networks and Internet-based communication and information access today are a crucial and indispensable part of working life. But in the private sphere, too, networked end-user equipment and the “network of networks” are playing an increasingly important role in everyday living. According to a study conducted by TNS Emnid (see TNS (N)ONLINER Atlas) in June 2004, almost 34 million Germans, or 53% of the population, already use the Internet privately, with a further 6.6% planning to do so “in the near future.” Thirty percent of the nation’s Net users stated that they had broadband access through technologies like DSL, and 10% used wireless LANs to network equipment. This means that the wired population has ceased to be a mere vision of the future and is now an everyday fact of life. In the next five years, technologies that already exist today, along with new access methods, and portable, easy-to-operate end-user equipment, will close up all those gaps in life where we are still offline. This report offers a brief backgrounder on the technical and economic aspects surrounding the most important technologies and highlights these technologies’ benefits and risks.

Corporate networks The changes will be least pronounced in those areas where networking is already a highly pervasive part of everyday life – in the workplace, for instance. A number of new aspects will emerge, however, as the continuing cycle of acceleration takes desktop networking up to the next level – from Fast Ethernet to Gigabit Ethernet (GBE). The technical prerequisites are already in place: By and large, every computer system shipping today –

454

Jörg Luther

desktop or laptop – is already equipped with a 100/1000 Mbps network interface. However, only a handful of companies are using this capability to deploy gigabit-to-the-desk architectures. The reason why businesses are holding back is that the requisite infrastructure components are still costly. The cost of a GBE switch interface may be just marginally higher than that of a Fast Ethernet interface, but the 10-gigabit Ethernet ports (10GE) needed to move data on backbone connections still cost around US$5,000 apiece at the start of 2004. However, 10GE per-port prices are already dropping rapidly: In early 2003, they were priced at US$80,000. So one can predict with more than a degree of certainty that the price per 10GE interface will dip below the magic US$500 threshold during the course of 2005, giving the green light for countless corporations to begin upgrading their networks to gigabit-to-the-desk technology. The increase in data throughput introduced by these faster corporate networks will open the door to new, more network-intensive applications that would normally overstretch the already tight capacity on today’s GBE backbones. With the advent of higher-capacity enterprise networks, a number of such applications already being used to a greater or lesser degree will find broader takeup. This category of application includes LAN-based employee training (e-learning), document sharing (collaboration), and virtual meetings (conferencing). Faster enterprise LANs will also be a crucial enabler for one much-vaunted but, as yet, rarely implemented technology, namely telephony over IP data networks, or voice over IP.

Voice over IP The term voice over IP (VoIP for short) actually subsumes a whole series of methods for transmitting voice traffic over IP-based networks, including in-company applications and IP telephony over the public Internet. Overall, though, VoIP is just one facet of a wider trend away from dedicated data and telephony networks and toward a unified network structure. The process of merging the technologies used to transport voice, multimedia, and data traffic is generally referred to as convergence. This has been a major issue for incumbent network carriers for some time now: In the latter half of the 1990s, their circuit-switched (telephony) networks began carrying an growing volume of packet-switched (data) traffic, but these networks lacked the kind of flexibility that this traffic needed. Whereas in 1990 the ratio of telephony to data was 80:20, the situation

Living and working in a global network

455

now is completely reversed. This has prompted carriers to begin gradually switching their backbones over to convergent voice and data services. The corporate sector, by contrast, still has a lot of catching up to do in this area in spite of the fact that switching to IP telephony affords numerous benefits and, particularly for companies with multiple locations, substantial cost advantages. In spite of contending with chronically congested enterprise LANs, companies continue to hold back, frequently for “political” reasons. These include long-term contracts binding businesses to telephony providers, fears among managers responsible for phone systems (rarely the same people as those in charge of networks) that they could lose authority, and users’ reservations about having telephony computers on their desks. But even long-term contracts eventually expire, mounting cost pressures can render turf wars irrelevant, and vendors can encourage user migration by marketing terminals cleverly disguised as office phones. This means that the network phone is finding its way onto an increasing number of desks in corporations, and in 2004 sales of voice over IP equipment will likely break through the US$1 billion barrier. But this is only the tip of the iceberg: By 2008, this figure will soar fivefold, and more than 260 million users worldwide will be placing and taking calls over a VoIP connection at work (see The Radicati Group). Whereas major companies generally seek to protect their investment by choosing hybrid solutions that incorporate classic, circuit-switched technology, small and midsize companies are tending to migrate fully to VoIP. Given that small and midsize enterprises (SMEs) are the biggest employers in Germany, the majority of us will be telephoning over IP by the end of the decade. As VoIP technology becomes increasingly widespread, telephony terminals will change substantially. First-generation terminals are still trying to court user acceptance by offering hybrid ISDN and VoIP functionality and by providing user interfaces that reflect classic phone front ends. However, there is no reason why, by the end of this decade, a VoIP terminal should continue to resemble a 1980s pushbutton phone; in fact, there are several compelling reasons why it should not. Of these, the most important is that confining the front end to a classic style of telephone interface unnecessarily restricts a terminal’s possible functionality. Today’s average IP phone features a large display but generally does not use it to show any information other than the kind a regular system phone would provide. By contrast, so-called soft phones – softwarebased VoIP applications running on PCs – offer a real functionality gain.

456

Jörg Luther

These are capable of efficiently melding phones’ and computers’ capabilities through computer telephony integration (CTI). When a soft phone receives an incoming call, it can retrieve data relevant to that call from a CRM application or from an internal directory service and display that data on screen. For example, when an outside call comes in, the customer contact data associated with the caller ID is displayed along with, say, recent transaction information retrieved from the inventory management system. The benefits of this kind of solution in terms of efficiency and customer service are largely self-evident.

Figure 100: Advanced functionality: Softphones are an ideal complement to a desktop device and they can present the same information in different ways to suit the individual needs of the user

Mobile telephony: VoWLAN, UMTS, and more There is no reason why VoIP should stop at the desktop. It can also be combined ideally with another form of mobile communication technology commonly already in place in the majority of businesses, namely wireless Ethernet, aka wireless LAN, or WLAN for short. Wireless LANs are nothing new to laptop users: Almost every notebook PC sold today comes already equipped with a wireless interface; Intel’s Centrino mobile chipset, for example, has built-in support for wireless networking. Two current WLAN standards, IEEE 801.11a and 11g, support bandwidths up to 54 Mbps – sufficient to handle not just any standard desk-

Living and working in a global network

457

top application but also multimedia and IP telephony. Increasing WLAN data rates – slated to rise to around 150 Mbps by 2006 – and emerging quality of service standards (IEEE 802.11e) will make it increasingly simple to integrate multiple services on WLAN connections. One such service is wireless telephony. Here, too, VoIP – or, more accurately, voice over WLAN (VoWLAN) – is a viable option. Just as with wireline VoIP, the benefits here are enhanced functionality and lower costs. Equipped with a WLAN mobile phone, users in companies can always be reached under the same phone number, regardless of whether they happen to be at their desks. Off campus, too, they can place calls through wireless hotspots – in train stations, airports, hotels, and even beer gardens – through an IP telephony provider much more affordably than through a classic mobile phone operator’s service. Forthcoming cell-phone chipsets like those currently being developed by Agere will also incorporate WLAN capability alongside regular mobile communications technology like GSM/GPRS and UMTS. Handsets and terminals based on these chips will enable users not just to communicate practically anywhere, they will also allow them to choose the least expensive communication channel available. Here, too, the Internet protocol is the key: Much to the chagrin of many a mobile phone provider, VoIP allows calls to be placed inexpensively over UMTS connections. As a study conducted by Mummert Consulting in June 2004 reveals, calls made using VoIP on UMTS are roughly cheaper by two-thirds than calls placed directly over UMTS, because this mobile service (which operates at speeds up to 384 Kbps) offers better value for money when carrying business data rather than voice. The growing number of mobile applications in the next few years will drive a change in mobile end-user equipment – away from “dumb” mobile phones in favor of “smart phones.” Besides serving as handsets for simple phone calls, these terminals will incorporate a personal information manager, a web browser, and Java support to help mobile users remain permanently connected to a full range of information sources in the enterprise and to reconcile and replicate data through web applications. But even use in an enterprise context is not sufficient to push these smart phones to the limits of their capabilities. They are also suitable as remotecontrol units in an environment where networking will likely not take hold fully for perhaps another decade or so: the “smart” home.

458

Jörg Luther

Figure 101: The shape of things to come: Tomorrow’s UMTS-based smart phones could look something like this Nokia design study

Home networks – smart living The Futurelife Project (see Futurelife Project) in Cham, Switzerland, reveals how a networked home can be about much more than, say, Internet access, interconnected consumer electronics, and the much-touted Internetenabled refrigerator. The four members of the Steiner family live in a single-family house kitted out with the very latest technology as part of an initiative designed to put the possibilities of wired living to the test and to gauge the limits of this kind of environment. The heart of their networked home is a central server rack, located in the basement, to which network-enabled components and household appliances are connected over a standard European Installation Bus (EIB) system. The Steiner family can monitor the status of most of these components over an Internet connection; many of the devices can also be controlled over the Web. This means that the occupants can control the heating and ventilation, the lights, and all of the power sockets in the home remotely – effectively putting paid to the eternal worry that the iron might still be plugged in or the washing machine running. Using the same means, they can switch on the coffee machine or the kitchen oven while out of the house. The family can also control and operate their home’s security and access control systems over an Internet, intranet, or WAP connection.

Living and working in a global network

459

Besides the powerful, standards-based but difficult to upgrade EIB, the Futurelife house uses network technology designed to be retrofitted to buildings neither originally intended for smart living, nor fitted with the requisite components. This technology goes by the name of Powerline Communications (PLC) and allows equipment to be networked via standard power outlets. Siemens offers a complete range of networkable household appliances to match: The serve@Home series (see Siemens serve@Home) includes everything from cookers and ovens to washing machines, dryers, and network-enabled fridge-freezer combinations. These all connect to a central gateway that doubles as an Internet access point. When a technical problem occurs, the appliances can notify customer service automatically, if required. Using the diagnostic data provided, service engineers can then select the right spares before coming out to complete the repair.

230 volt power grid (Powerline)

Tablet PC

Today Gateway Future

Mobile phone

Heating / air conditioning

Security

Light

Shutter control

Figure 102: Home network architecture: Smart home appliances are interconnected via powerlines and controlled from a gateway. Other building technology can be added as necessary and integrated via the gateway

Developments and solutions like these, which aim to ease aspects of everyday life in private households, are potentially some of the key technologies of the decade. According to a study by Electrolux, around 20 mil-

460

Jörg Luther

lion U.S. households will spend a total of US$15 billion on home appliances with Internet control capabilities in 2005. And a forecast by UK market researchers Datamonitor points to a similar number of households in Europe that could be equipped with systems and appliances like this by the end of the decade.

Internet access in automobiles The realm of private networking would not be complete were it not to include the Germans’ number one passion, the automobile. One of the most comprehensive and best-conceived solutions for networking on four wheels for some time now has been ConnectedDrive (see ConnectedDrive), a system fitted by BMW to its 7 series sedans. ConnectedDrive has a display mounted in a separate panel next to the instrument cluster, and is operated with a joystick-type control wheel in the center console. Using a GSM car phone, the system connects to the BMW Online Portal to retrieve information on cultural events and current affairs, plus the latest political, business and stock market news. Through their own BMW Online account, drivers can receive e-mail and even compile replies from a selection of ready-made texts. As one of the organizations involved in the Futurelife Project, BMW has shown that controlling intelligent building systems from on the road poses no serious challenge to an in-car system like ConnectedDrive.

Figure 103: Functional: The ConnectedDrive system’s main controls are located on the center console. Almost all of the system’s functions can be selected using the joystick-type control wheel

Living and working in a global network

461

ConnectedDrive’s more advanced and valuable capabilities include a variety of location-based services delivered with the aid of the vehicle’s onboard GPS navigation system. For example, the system can check BMW’s Online Portal to find the next multistory car park with a free space and then guide the driver straight to it. It can also do much the same thing with restaurants and hotel accommodation. And if the car has a technical problem or breaks down, the system can locate the nearest BMW dealership and navigate to it or send online notification that a repair job is needed. The BMW Online Portal also offers quick and simple access to a wide variety of other information – including some 35 million business and private addresses preprogrammed into the vehicle’s navigation system, plus current flight information, and even details of the nearest open pharmacy.

Internet access in the air Anyone not traveling in their own car but by train, aircraft or ship has had to do without network access until now. However, this situation will change rapidly in the next few years. U.S. aircraft maker Boeing made an initial thrust to change this state of affairs in 2004 with the launch of its Connexion by Boeing service (see Connexion by Boeing), which offers passengers in-flight Internet access. Travelers flying with ANA, China Airlines, JAL, KAL (from 2005), SAS, Singapore Airlines and, last but not least, Lufthansa can now surf the Web, send and receive e-mail, and, if need be, connect to a corporate VPN while in the air (on selected routes, at least). Using satellites in geostationary orbit, Connexion by Boeing supports up to four downstream channels, each with a data rate of 5 Gbps. The maximum speed upstream is 1 Mbps, again split across four channels, each operating at 256 Kbps. This may not sound like all that much, but the capacity is roughly equivalent to a typical business-grade DSL connection in a small company. The service’s hardware and software requirements are modest: According to Boeing, all users require is a notebook computer running Windows (Windows 95 or higher) or MacOS 9/X with Open Transport 2.6, and an IEEE 802.11b-compatible network interface. Users’ browser software must support JavaScript and SSL, which means that Microsoft Internet Explorer Version 5 and higher or Netscape Navigator Version 4.7 and higher are sufficient. To use this airborne Internet connection, passengers must first create a

462

Jörg Luther

new wireless profile and set the WLAN SSID to ‘Connexion1.’ Using a web browser, they then sign up for the service through the URL cbb.by.boeing, where they receive their login data. Users pay for the service by credit card and are required to supply the relevant billing details during the signup process. Connexion by Boeing offers users a choice of charge models, including a flat fee for unlimited access which varies depending on the flight’s duration. Alternatively they can choose to be charged by the minute on top of a basic access fee that again varies depending on the duration of the flight. After initial tests in 2003 showed that Boeing’s Connexion service not only worked well but also represented an important selling point, numerous airlines began kitting out their fleets with the requisite systems in 2004. Lufthansa, for instance, wants to equip some 80 long-haul jets – Boeing 747-700, A 330, and A 340 aircraft – with Connexion (see Lufthansa FlyNet). Boeing’s satellite-based Internet access system has also gained a foothold in the shipping sector. Connexion inked its first agreement covering maritime trials with Teekay Shipping Corporation in June 2004.

Figure 104: Complex connectivity: Connexion by Boeing is a satellite-based system that relies on an extensive infrastructure on the ground. As a result, access charges are relatively high

Living and working in a global network

463

Besides access to Internet and intranet content, ships’ operators and passengers will also be able to retrieve e-mails – even messages with large attachments. The maritime system uses the same satellites and ground infrastructure as the airlines and will likely operate at 2.5 Mbps downstream and at 256 Kbps upstream, making it substantially faster than the Inmarsat narrowband system currently in service at sea.

Internet access on trains Once air travelers have grown accustomed to the convenience of in-flight Internet access, rail carriers will soon feel compelled to launch a similar service of their own. Carriers like Deutsche Bahn, eager to entice business travelers out of planes and into trains, will soon have no choice but to offer travelers network access. To date, Deutsche Bahn has confined itself to setting up WLAN hotspots at train stations in major cities (see Deutsche Bahn rail&mail). In 2003, the company trialed information services in collaboration with Microsoft on the intercity express service operating between Nuremberg and Hanover, albeit on a purely offline basis, because rail routes “lacked the requisite infrastructure” for an online service, it was claimed. Closer scrutiny, however, reveals that this was likely a case of the former national rail carrier (not known for its propensity for innovation) merely making excuses. After all, people can place mobile phone calls in trains, and the same technology can be used to offer Internet access on rail services. Marburger IPmotion GmbH (see IPmotion GmbH), for example, already began operating a similar, GPRS-based system in 2002. The system, a modular multichannel GPRS router that goes by the name of CAR-A-VAN.rail, is already in service on Stuttgart’s commuter rail system and has also been deployed by Edelaraudtee, Estonia’s rail company. In fact, the latter offers passengers traveling first class free Internet access on express services, and has figures to prove that it generate a return on the investment within the space of one year if just one additional passenger on every trip purchases a first-class ticket on the strength of the free Internet service.

464

Jörg Luther

Brave new world? As the examples presented here reveal, the technology itself has long since ceased to be a stumbling block in the way of achieving comprehensive networking in both work and home life. In a society in which information has long since advanced to become one of the most important resources, people really will be completely wired by 2010. But once people are online around the clock – at home, in the office, on land, sea and in the air – they will also face new annoyances and dangers. The least of these are the computer viruses, worms, and Trojan horses. Making software and communication protocols secure and providing users with regular, preferably transparent, updates is primarily the responsibility of the manufacturers of end-user equipment. Keeping browsers up-to-date on laptop computers is something that users can be expected to do themselves, but regular provision of updated software for washing machines, home control systems, and automobiles’ online systems has to take place automatically as far as possible so as to ensure ease of use and operational safety. One snowballing problem is the influx of unwanted information. Unsolicited e-mail (aka “spam”) is a major nuisance that now accounts for more than half of all the messages received in Germany. According to estimates by e-mail providers like MessageLabs, roughly eight in every ten messages in the U.S. at the end of 2004 were spam. And this phenomenon is by no means restricted to electronic mail. It is becoming increasingly common for SMS spam to be sent to mobile phones. And “bluejacking” (see www.bluejakq.com) – the manipulation of Bluetooth-enabled mobile phones – underscores how those in the know enjoy exploiting every possible vulnerability in order to hack network-connected equipment. Spam can be blocked with the aid of secured communication protocols and filtering software, but this merely targets the symptoms rather than addressing the root of the problem. In this globally networked world, what is missing are internationally coordinated statutory regulations that prohibit the sending of spam of any kind and enable cross-border criminal prosecution. Much the same applies in essence to the most serious problem facing the networked individual, namely privacy. People who are constantly online – around the clock and wherever they go – leave behind them a traceable trail of logged information. Anyone who connects to a network with a mobile device can be located based on their current radio cell and can be followed as they move from one cell to the next. Every order placed online by a networked fridge and every query sent to a company server leaves traces of

Living and working in a global network

465

data on the network which, even with today’s technology, can be mapped precisely to individual users. Without internationally coordinated data protection regulations, the wired individual will inevitably end up being a transparent citizen – the beginning, perhaps, of an Orwellian nightmare.

Sources and links: [1] TNS Emnid (N)ONLINER Atlas http://www.nonliner-atlas.de/ [2] The Radicati Group, “Corporate VoIP Market, 2004–2008” http://www.radicati.com/pubs/news/04IP%20Telephony.pdf [3] Futurelife Project http://www.futurelife.ch [4] Siemens serve@Home http://www.serve-home.de [5] BMW ConnectedDrive http://www.bmw.com/generic/com/en/fascination/technology/connecteddrive/ [6] Connexion by Boeing http://www.connexionbyboeing.com/ [7] Lufthansa FlyNet http://www.flynet.lufthansa.com/ [8] Deutsche Bahn rail&mail http://www.bahn.de/pv/view/home/aktion/rail_and_mail.shtml [9] IPmotion GmbH http://www.ipmotion.de/

Andy Mattes

Delivering the promise. Making it pervasive

The RTE is an exciting, wide-ranging concept that is predicated on the need to act in real time. Keep that simple statement in mind; it’s the promise that has to be delivered throughout the enterprise. The real-time enterprise has been an IT goal for a decade or more. Search the Internet and phrases such as “moving mission-critical data through intranets at near real time will turn up”. You will also find catchy phrases like “what time is real time” and “the right time being what really matters”. And if you start reading the content you may conclude that the RTE is too complex and too expensive to take on board. It appears to be complex because the RTE is an acronym that encompasses just about everything. It’s IM, it’s business process integration, it’s Web services and so on. In other words, different agendas drive the way the RTE is marketed, which is unfortunate. The apparent breadth of this concept allows journalists and analysts to approach it from diverse directions, which only adds to the confusion. Instead of catchy phrases we need authoritative comment and analysis, which this book seeks to provide. Confusion must be replaced by clarity, because a confused market does not buy, it waits. Let’s start by repeating the opening statement that the RTE is predicated on the need of enterprises to act in real time. That is the primary goal; everything else is secondary. Technology can minimize the lag time between detecting, reporting and responding to an event, but acting on information and responding to events and issues in real time involves people. Computers can flag issues, but computers can only make decisions when everything runs smoothly. They stop and wait for instructions when something goes wrong. The event/issue may be something relatively simple like a parts shortage in a manufacturing process or it could be complex. For example, an important customer has called with a problem involving a set of issues that can only be addressed by a team, e.g. a senior manager, the account manager and high-level techies.

468

Andy Mattes

In each case the intrinsic functionality that needs to be in place is the same: (1) find the right people in minutes not hours; (2) supply the relevant information to all parties; (3) put them in contact via a conference call. Steps two and three are a no-brainer for HiPath OpenScape, a suite of realtime, presence-aware software. A skill-based directory best handles the first step, i.e. enter a few key words in order to find the relevant names along with data on their presence and availability. Once contact is established people take over: the issue is discussed, decisions are taken, and (hopefully) the process proceeds. Seamless linking between skill-based directories and HiPath OpenScape is therefore an important RTE enabler. Step 2 might also involve automated access to information known to be relevant to the event/issue. This facility is provided by HiPath software and the relevant files are displayed on the desktops of all parties participating in the conference call. Right now the ICT landscape is littered with island solutions (aka stovepipes) and in many cases the tasks for which they were originally designed have moved on. Legacy voicemail systems, for example, are of little value to today’s increasingly mobile workforce. Dialing in only to be told that you have no new messages is an expensive waste of time and money. The economy has picked up and the brakes are being gently released on CAPEX budgets. This is enabling corporate management to make topdown appraisals of their communications and collaboration requirements. It is clear that these will vary from concern to concern, but island solutions are definitely out and unification is very much in. Convergence at the network level — transporting voice, data and video over the same infrastructure — is a given. The next step is the seamless integration of real-time communications with business processes. However, there are early signs that different vendor agendas will drag the RTE in different directions. Recall the earlier two events. Parts shortages are an ERP issue. The hypothetical customer who has a set of problems might come in via the contact center, which makes it a CRM issue. Alternatively, he/she might simply call the account manager. This indicates that a meaningful RTE solution must be holistic: it must be available to anybody who needs to act in real time to any event-driven situation. Solutions (so called) that are based on a particular technology such as IM or an application such as data warehousing are just another island. That may sound like an obvious statement, but an Internet search throws up many examples of island thinking. For example, the real-time enterprise

Delivering the promise. Making it pervasive

469

is not about the need to respond to emails in minutes instead of hours, days, or not at all. That’s an important issue and it’s a problem that should be fixed, but it has little to do with the RTE. At best it plays a secondary role. There are solutions that interpret email content and send an instant message to the relevant party in order to facilitate a speedy response. An email sent to ‘info@’ might send an IM to sales personnel who are on line. A more sophisticated solution might use key words in order to scan the content of the email so that the most relevant individual is contacted. There is much to commend here, but it is wrong to hype email-to-IM conversion and call it an RTE solution. However, if these IMs are linked to presence and availability services, then the functionality can form an integral and important part of a solution. Thus, while it is a useful application in its own right, it is not sitting on an island. HiPath OpenScape provides generic presence and availability functionality, i.e. the benefits apply to individuals and workgroups as well as processes. The optimization of those processes starts with a top-down procedure that specifies how the RTE solution should function. Implementation, however, is bottom-up. The software’s intrinsic functionality is mapped to the specific requirements of the various processes, ERP, CPM, etc. The start of this article indicated the need to deliver the RTE promise throughout the enterprise. And we have touched on the need for solutions to be holistic. That may sound like a daunting task; in fact, the RTE concept might sound like one of those high-tech IT projects that goes over budget and underperforms. Let us be honest with ourselves, the industry’s track record has been less than impressive in recent years. With the benefit of hindsight it is easy to see what went wrong. All too often the cost and complexity of implementing high-flying solutions exceeded the long-term benefits, so there was no return on investment. The ICT industry also became infatuated with technology at times and along the way lost sight of the fact that technology was only a means to an end. Those days are over, but a degree of caution remains and that is positive. There has to be a compelling business case for the RTE and it has to be made upfront. Gartner talks about “squeezing lag time or ‘information float’ out of core operational and managerial processes.” They also bring in the ‘time is money’ cliché and indicate that speed is a competitive advantage. These are nice words, but haven’t we been here before? Million dollar/euro investments are not going to be made on consultancy speak. And recent high-tech history also tells us that the value of the RTE is not

470

Andy Mattes

the ability to process in real time; it’s the ability to act on events as they happen and thereby gain a competitive advantage. It’s that simple and there’s nothing new about resolving issues quickly. Before the economy went global and workforces went mobile you could cross the corridor and consult a colleague. The difference now is that we need high-tech software and smart wireless devices to enable yesterday’s high-touch way of working, which brings us neatly back to real-time, presence-aware applications and the need to make the benefits pervasive. Let’s start with presence, which is a very basic parameter. Fred is in his office across the hall, he’s ‘present’ and if he’s not busy then he’s ‘available’ and a colleague can phone him or walk over and talk. Digital presence refers to the network status of devices such as PCs, PDAs and IP phones. If a PC is ‘on-line’ then it is ‘present’ and a presence-aware application can detect that fact and display it to authorized parties who are also on-line at the time. Instant Messaging (IM) is the obvious example of a presenceaware communications application. Availability refers to the status of the user of the device. A PC may be on-line but the user is absent or he/she does not want to be disturbed. IM programs allow users to display availability information alongside the presence icon. Thus, availability can be managed, although the functionality in the case of popular public programs is limited to a handful of standard messages such as busy, on the phone, out to lunch, etc. An IP phone is a data device that connects to a local area network and applications such as HiPath OpenScape have been developed, see figure 105, that display status in a similar way to that of PCs. In presence-aware telephony the attribute is an icon that shows if the phone is on-hook (free to take a call) or off-hook (the user is talking). In this case availability is managed via personal profiles having more functionality than that of popular IM applications. Unlike IM, which is an established application used on both the Internet and intranets, telephony presence is a relatively new concept, the implications of which are far-reaching. The Presence and Availability Management (PAM) Forum defines Availability as a property of an entity denoting its ability and willingness to communicate with another entity based on the preferences and policies that are associated with the entity. The reference to preferences and policies indicates that availability is managed at both the individual and corporate levels.

Delivering the promise. Making it pervasive

471

Figure 105: Icons display presence and availability. For example, Bruce Walker has blocked his phone (red icon); yellow indicates that the phone is busy, blue that it is not. The grey icon indicates that the party does not want his/her telephony status to be displayed

Managed availability solves the ‘who, what, when and where’ of communications. Calls and messages are received or rejected on the basis of: who is allowed to reach the called party; what device has been defined as being the most convenient at that time; and the times when the party can be reached. HiPath OpenScape uses a combination of presence and managed availability to minimize telephone tag, thereby boosting personal productivity, yet at the same time it reduces stress. The same application allows one colleague to talk to another immediately when there is a need to react to an event or an urgent issue. Contacts are the enterprise equivalent to IM buddies and as shown in figure 106, the ability to list them maps to the team/group way of working that is prevalent in today’s organizations. Many people work in different groups simultaneously; they may be in a marketing team but have some specific project assignments. Thus, they operate in a communications/collaboration matrix and their ideas and actions are organized along these lines. This means that there is a need for tools that facilitate and optimize that paradigm.

472

Andy Mattes

Figure 106: ‘My Workgroups’ appear in a second window, alongside ‘My Contacts’. This screen shot also indicates how users manage their baseline availability

Imagine that a manager is responsible for several projects. If an issue arises he/she can select the relevant portal, which displays the names of all members, their presence status for IM and telephony, as well as the file names of the relevant documents. The issue can be addressed immediately by clicking on the relevant names, i.e. a conference call is established and all participants have immediate access to the same set of documents. Adding content is easy; users simply drag documents into a new Window. They can also be linked to the session, so that they are available immediately when a new session for that particular workgroup starts up. Thus, once again we have an application that is not only valuable in its own right, but also a key RTE enabler. Replace ‘project’ with ‘issues’, e.g. the sourcing of parts for a manufacturing process and the ability to react in real time to an event such as a shortage of a particular widget is obvious. The presence and availability of all relevant parties is displayed and a single mouse click will initiate a conference call. If somebody is talking the application can send an IM to let him/her know that his/her presence (no pun intended) is needed. If there is a need to refer to documents, and this issue has come up before, then the relevant files will be displayed automatically in the collaboration window. Both IM and telephony presence are set to become core components of mainstream applications. When there is a glitch in a workflow application the intrinsic stickiness of presence can be used to make intelligent decisions about workflow routing, i.e. information on the glitch is sent to a list of people who are available and who have the requisite knowledge to handle it.

Delivering the promise. Making it pervasive

473

This function may be bundled within the solution or supplied as an additional ‘alert’ facility. IM is an excellent medium for use in RTE solutions since it is a real-time data medium and can therefore be integrated with business processes. Several mainstream and specialist vendors are marketing solutions that feature process integration. FaceTime, for example, has a network-independent application for customer interaction and internal agent/expert interaction. Basically it allows telephone agents to leverage the presence and availability features of IM networks from AOL, Microsoft and Yahoo. Ikimbo’s solution executes when a defined exceptional event occurs and after resolution it passes the necessary information back into the system that posted the event. What Siemens enables is the next real-time step. The HiPath OpenScape suite of presence-aware software links IM (real-time data) to real-time IP telephony. It also adds powerful collaboration features and puts everything together in an attractive graphical interface that can be customized to match individual communications and collaboration needs. So far so good. We’ve traveled quite a way down the road towards the RTE, but so far we have only considered the kind of internal communications functionality that is needed to realize this concept. In recent years many companies have focused their resources on core competences and generic activities such as manufacturing have been outsourced. Ecosystems, which are visualized in figure 107, represent complex webs of inter-company transactions between customers, suppliers and other authorized third parties are the natural extension of this trend. It is therefore clear that the ecosystem has to be brought into the RTE equation. Recall the earlier hypothetical issue of a widget shortage. This would normally come from a supplier, so the company would need to be contacted as soon as possible in order to resolve the problem. Extending the benefits of presence and the ‘click to communicate’ paradigm to authorized parties in the ecosystem is therefore a logical development. Moreover, once the benefits of enhanced communication and collaboration have been realized internally it makes sense to employ the same functionality throughout the ecosystem. In other words, for everyday business contact and not just emergencies.

474

Andy Mattes

Figure 107: Once the benefits of enhanced communication and collaboration have been realized inside the enterprise, the next step is to extend them to authorized parties in the ecosystem, both for the RTE as well as everyday contact

In theory any authorized third party who has an Internet connection and an IP communications device, phone or PC running softphone software, can communicate with employees of the ‘owner’ of the ecosystem, i.e. everybody sees each other’s presence and availability. The task is virtually identical to that of bringing teleworkers into the new communications loop, but the management of third parties is a barrier. However, a visionary concept that Siemens calls LifeWorks dissolves this and other communications barriers. LifeWorks is a carrier-class implementation of the company’s real-time communications applications, the most important of which is HiPath OpenScape. Authorized ecosystem partners that employ presence and availability as a managed service have the same presence and availability functionality as the employees of the ‘owner’ of the ecosystem. Extending the communications and collaboration paradigm in this way results in more efficient business processes, e.g. issues involving ecosystem partners can be resolved in real time. However, a less obvious benefit is the significant reduction in transaction costs. Outsourcing the manufacture of products drives prices down but ecosystems require many more transactions than in-house production. Research indicates that more than half the world’s GDP is based on transaction costs, so it does not require a great cognitive leap to realize that a similar percentage figure applies to the typical enterprise. It will vary from

Delivering the promise. Making it pervasive

475

industry to industry and company to company, but it does represent a substantial part of the cost of doing business. It is also clear that productivity is the key issue: it is the area where the biggest gains can be realized. Thus, reductions in this area will have a much bigger impact on prices than those in manufacture and distribution, where only relatively minor improvements can normally be realized. Time to step back for a moment and make some interim conclusions since we’ve covered a lot of ground. It is clear that we need smarter ways of communicating and collaborating in order to be able to act in real time. We need to be able to embed real-time communications within business processes. Solutions also require functionality that allows the right people to be located, regardless of location, network and current device, and then to enable synchronization with the information needed to make informed decisions. The communications and collaboration functionality must map to the RTE requirements. And ideally, we need a single instance of the technology that underpins this type of holistic solution. We need all this good stuff anyway. It is not driven by an RTE agenda but Siemens’ considered opinion is that this it the optimum way to reach that objective.

Figure 108: This conceptual view of a unified environment illustrates interoperation with other mainstream platforms as well as all mainstream client devices

476

Andy Mattes

A single instance of a productivity-enhancing engine enables the creation of unified communications and information environments. And as illustrated in figure 108, this development: enables communications via any real-time communications client; uses all mainstream media types; employs real-time communications applications; and allows knowledge workers to obtain information from back office systems. Nice words, but can it be done. The short answer is Yes: Siemens has brought its HiPath OpenScape concept to the marketplace, won various awards and realized a number of major wins. The core functionality, which is based on personal productivity and workgroup collaboration, works in conjunction with middleware in order to allow applications to interoperate with communications platforms, IT data platforms, and information databases, all of which operate in real-time. In addition, high-level Application Programming Interfaces (APIs) are provided so that third party developers can develop their own applications and customize those provided by Siemens. You can therefore think of the new communications and collaboration functionality as the engine that is driving the future of IP Communications. It sits at the center of the holistic solution and is customized to match specific needs, e.g. those of call centers, manufacturing processes, logistical processes and last but not least, the RTE. Two developments underpin our business and personal lives: the Internet and wireless telephony. The latter enabled today’s mobility model and in many companies over half the workforce are out of the office most of the time. In fact, the combination of the Net and the cell phone has changed the intrinsic concept of ‘the office’. No longer is it a place to which one goes to work. Today’s office is a mobile office. Smartphones and PDAs have powerful computing and communications resources and while there is room for improvement on the access side, we do have a complementary mix of highspeed “hot spots” and low-speed cellular networks. Thus, access is ubiquitous and services are getting better by the month. The need to bring mobile professionals into the RTE equation is clear. You can’t make the concept work by leaving out half the workforce, particularly since the people who are mobile most of the time tend to be the ones needed to address critical events and issues. The applications and services employed on mobile devices are basically the same as those of the regular office environment. However, many companies are adding wireless LAN extensions and doing so to such an extent that wireless access might overtake wireline in a few year’s time. This is a

Delivering the promise. Making it pervasive

477

huge subject in its own right, so let’s focus on a key issue: mobile presence. The presence and availability of mobile users has to become a seamless part of the unified environment. It has to anyway, and it is also mandatory for a meaningful real-time enterprise way of working. This is an evolution that is coming to wireless networks; in the case of push-to-talk telephony it has arrived, but wireless remains the domain of network operators. There are various solutions that will enable seamless interoperability, but there are a number of business issues. Mobile presence is another big subject in its own right, but we can cut to the chase here and fast forward to a development that, like IP and the Internet, changes everything: SIP (Session Initiation Protocol). It’s hard to under-hype SIP. Back in 1996 Vint Cerf, widely regarded as the father of the Internet said that SIP would be “the third great Internet protocol innovation, after TCP/IP and the web.” What SIP enables eight years on is almost unbelievable. SIP is an end-to-end, rendezvous protocol that establishes communications links over different IP networks and between disparate devices. SIPcompliant devices communicate with each other in a way that is totally transparent to the user. It is the only protocol needed to support voice, chat, instant messaging, presence, multimedia messaging and video and it has been selected as the call control standard for 3G wireless networks. Thus, mobile presence will arrive courtesy of this groundbreaking development. SIP can also interact with other communications protocols via gateways, and tie together multiple features into more advanced services, e.g. clickto-talk. And it works across different platforms — hardware phones and desktop soft clients plus enterprise- and carrier-class servers. This is a core technology with uses in many applications beyond VoIP. These include presence, IM, push-to-talk, presence & location-based services, etc. On the communications front, most everything that is SIP-compliant and IP-based will interoperate: everything encompasses applications, devices and networks. Unlike IP and the Web, this is a high-touch protocol, one that is already turning communications and collaboration processes on their heads. In a nutshell, SIP dissolves the boundaries between networks: the end user experience will be that of a unified domain. And SIP is the key technology that underpins the LifeWorks concept we outlined earlier. Things that sound too good to be true usually are, but SIP is the exception that proves the rule.

478

Andy Mattes

Conclusions

There’s nothing new about the RTE, but now we have the tools and the technology to make it happen. The concept is based on the need to act in real time and for that we need presence-aware, real-time communications applications. And in order to act in a timely manner people need to be synchronized with each other and the information needed to arrive at an informed decision. Right now we have a plethora of island solutions: an information and communications archipelago. They are being replaced by holistic solutions that address fundamental issues and meet intrinsic needs. These needs and issues are not confined to the RTE: they cover most everything on the communications and collaborations fronts. From the need to eliminate telephone tag through to minimizing transactions and on to the place where this article started: the Real-Time Enterprise. Q.E.D.

IX. The visions

Living and working in a real-time, globally networked world will bring dramatic changes in both the professional and private sides of our life. In fact, this development has already started. Even a conservative extrapolation of the possibilities of today’s technologies indicates that lifestyles regarded as science fiction today will become an everyday reality in the not too distant future.

Claus Weyrich

Knowledge-based companies – objectives and requirements

The information age and knowledge-based society Communication and information technology has not only revolutionized electrical engineering; it has pervaded practically all branches of industry and is increasingly shaping our lives. We have undoubtedly entered a new era – an age of information and knowledge, in which human beings are at the center of rapidly growing networks, not just in the classical sense of telecommunications, but in terms of rational, knowledge-based decisionsupport systems. These systems are becoming a foundation for workingtogether and living-together.

Percent 90 Agricultural age

80

Industrial age

Information age

70 60 Information

50 40 30

Services

20

“Classic” production

10 Agriculture 0 1800

1850

1900

1950

2000

Figure 109: Development of employment from 1800 to 2000

Year

482

Claus Weyrich

The transformation we are now witnessing is so profound as to defy comparison in human history. Its closest parallel is the emergence of the industrial revolution and the concurrent decline of agricultural society. The industrial revolution, which began roughly 200 years ago, profoundly altered the labor market – a process that accelerated during the 19th century as farming increasingly lost ground to industrial production. Triggering much of this were three crucial inventions: the steam engine, the electric motor and the internal-combustion engine. These basic innovations led to a rapid rise in productivity and, hence, to growth and to a substantial increase in employment. Today, similarly dramatic changes are taking shape. But the difference is that rather than merely multiplying our physical powers, machines are now multiplying our mental powers as well. This is driving a process of renewal in the economics of production and in society as information and knowledge become increasingly accessible. Workplaces that are a part of this process are gaining in importance relative to classical production sectors, which are losing relevance in terms of employment. And as we move toward an information- and knowledge-based society, the service sector is gaining in importance, primarily in the area of higher value and so-called value-added services. This is not to say, however, that the agricultural and industrial production sectors are unimportant. On the contrary, they are conspicuous features of our industrial and cultural landscape, and help to ensure that we do not move toward and information “monoculture.” As in nature, societies thrive on diversity and are weakened by a excessive homogeneity. Furthermore, the knowledge-based workplace depends on the existence of a highly developed system of industrial production.

Microelectronics and software: driving change The answer to the question of which basic innovation brought on the information age is evident: it was microelectronics. This crucial area of development permanently revolutionized computer and communication technologies. The advent of microelectronics can be traced to the invention of the transistor in 1947 and that of the integrated circuit in 1958 – incidentally, both achievements that were honored with Nobel prizes. To understand what breakthroughs in microelectronics have meant, consider the following comparison: The first fully electronic computer, ENIAC (Electrical Numerical Integrator and Computer), which was built in 1946, required 18,000 tubes and consumed 150 Kilowatts. ENIAC cost $2 mil-

Knowledge-based companies – objectives and requirements

ENIAC: First electronic computer (1946) • Components: 18,000 tubes • Multiplication time: 9 ms • Power consumption: 150 kw • Price: 2 Mio. $

483

Intel Pentium III (2000) • Components: 28.1 million transistors • Multiplication time: < 5ns • Chip area: 106 mm2 (0,18 µm) • Power consumption: 33 W (1 GHz) • Price: approx. $400

Microelectronics: Science with a Future

Figure 110: Microelectronics has triggered the information and communication technology revolution

lion, was as large as a living room, weighed 30 tons, and nevertheless required three milliseconds to perform a multiplication function. Just 55 years later, in 2000, a slice of silicon with a surface area of around one square centimeter has enough room for 28.1 million transistors where each transistor fulfills the same functional purpose as an ENIAC tube . This Pentium III microprocessor – the heart of today’s PC and laptops – completes one multiplication in less than five nanoseconds, requires only 33 watts for its operation, and costs around $400. If we compare the priceperformance ratio of ENIAC to that of a Pentium III and express it in terms of computer performance per Watt and Dollar, we find that these values have increased by a factor of 1013. Expressed in centimeters, this number corresponds to the distance between the earth and the sun, or about 150 million kilometers. Taking the weight of both computers into consideration, this would result in the almost unimaginable value of 1017, which corresponds approximately to the age of the universe – measured in seconds. These dimensions may truly be regarded as astronomical. In the history of applied science they represent an explosion of technical progress never seen before. The evolution of microelectronics is far from over. Today’s most complex product, the 1-Gbit memory chip with more than 2 billion components on approximately two square centimeters of silicon, can store around 64,000 pages of text, which equates to a stack of books several meters high. In ten to fifteen years 64-Gbit memory chips will be feasible. Their storage capac-

484

Claus Weyrich

ity will correspond to a tower of books several hundred meters in height. In addition to microelectronics, software has become part of the indispensable technology mix of the information age. Software brings microelectronics hardware to life and has become its symbiotic partner in the sense that each technology stimulates the other. Software technology has developed just about as rapidly as microelectronics. Furthermore, the complexity of the largest software systems, such as those used in meteorology, computer and communications technology has increased exponentially over time. Today, such systems may include up to 50 million “lines of code” and their architectures are no less impressive than those of integrated circuits.

In 2003*: • approx. 1.2 billion fixed network connections • approx. 1.3 billion mobile phone users • approx. 665 million Internet users * Source: ITU

fixed network connections

mobile phone users

Internet Hosts

Figure 111: New forms of communication: mobile radio networks and the Internet

Microelectronics and software have also enabled completely new forms of communication such as mobile radio networks and the Internet. As these technologies have evolved, the number of mobile phone users has grown to 1.3 billion subscribers worldwide, thus surpassing the number of fixed network connections, while the number of Internet users has skyrocketed to well over 500 million. It is also worth noting that the rate at which new technologies are expanding is accelerating. Take fixed network connections, for instance. It took 50 years for the number of subscribers to grow from 1 million to 100 million; but it has taken only a decade for mobile phone communication to achieve the same level of utilization. Internet use has increased even more rapidly. And without a doubt, innovation cycles will become ever shorter in the future.

Knowledge-based companies – objectives and requirements

485

Furthermore, modern information and communication technology has also structurally changed the value chain of industrial production. Increasing integration and miniaturization, together with the growing replacement of hardware functionality by software, have resulted in the value-added part of traditional production diminishing steadily in importance compared to the so-called ‘early’ and ‘late’ phases of the value chain. This is caused by the increasing variety of product versions, as well as growing performance requirements – forces that are driven by the market, which significantly increase development, marketing, product definition and services costs. In addition, all processes of the value chain have become increasingly permeated by information technology and exhibit very close interlocking with services. All in all, the volume of so-called “knowledge-based” businesses is growing by leaps and bounds.

Product-based businesses versus knowledge-based businesses The transition from a pure product-based business to a knowledge-based business represents a paradigm shift for companies. Physical products leave the manufacturer and go to the customer. Knowledge-based products, on the other hand, stay with the vendor while at the same time going to the customer. This emphasizes the fact that knowledge grows by using and sharing and is indeed raised to a higher power. This experience provides a sharp contrast to the depreciation of material products as they are used. If economies of scale, i.e. large-scale production with resulting cost-per-piece reductions, are decisive for achieving a strong market position, then even small companies have the potential to do so in specific segments of knowledge-based businesses. This is not the case where physical products are concerned. Here, inadequate capital or insufficient time often prove to be a bottleneck because companies need to allocate physical resources to production, and these resources are dependent on location. Accordingly, enterprise planning is strongly aligned to assets. In contrast to this, the bottlenecks in knowledge-based companies lie particularly in the number of knowledge carriers and, above all, in the quality of knowledge exchanges. The production resource known as “knowledge” is located predominantly in the heads of employees and, due to increasingly efficient real-time communication across time zones, is far less dependent on location. Consequently, efficient networks and relationships play a dominant role in knowledge-based companies.

486

Claus Weyrich

Nevertheless, both product-based and knowledge-based businesses have their specific strengths. However, it is also a fact that these strengths tend to lead to a symbiosis of medium-sized operations with big industries. In any case, with or without such unions, large companies are increasingly attempting to integrate the entrepreneurial culture of smaller companies into their operations through the establishment of “innovation cells” or “incubators.” Such units are designed to foster the development of new businesses in order to exploit new markets with greater flexibility and speed. On the other hand, smaller companies attempt to enter into strategic alliances with major companies in order to globalize their businesses.

Knowledge-based companies: a vision In a visionary picture, a knowledge-based company acts as a flexible and breathing network of skills whose most important raw material is knowledge. Deployed at the right time, in the right location, and always in a customized form, that knowledge can become a powerful competitive advantage. Here, the key success factor is the creation and solution-oriented application of knowledge that is available to every employee at any time by the application of a standardized and efficient information and communication infrastructure. At the same time, openness to a company-wide exchange of knowledge must be an essential component of company culture. This, however, presents a huge challenge both to company management and for employees. The basis of every organization is the structure that defines its responsibilities, underlying processes, the rolls of employees, and the values that are an essential component of the company culture. The so-called “objective pyramid” should be regarded as an element of management where the peak represents the company vision, or long-term objective that should be striven for by everyone. It goes without saying that strategy is developed in order to achieve the vision from which, in turn, operative objectives are derived. These objectives specifically describe what should be achieved when on the path to realization. Again, these objectives present a clear guideline for the development of milestones. The transition to becoming a knowledge-based company undoubtedly requires a rethink of the five basic elements that will be examined in more detail below.

Knowledge-based companies – objectives and requirements

Operative implementation

Vision

487

What do we want to achieve and what do we want to be? How do we realize our vision?

Strategy Operative goals Structures

What do we actually want to achieve (and by when) on the path to realization? Values

Organization Principles Processes

Skills Employees

Figure 112: Organization’s major elements must be aligned to its vision

Organization structures In the past, organizations were hierarchically rigid and structured in a function-oriented way with units that were optimized within themselves, such as research, development, marketing, production and sales. Today, processoriented operation dominates, using local organizational blocks, but – in the sense of a matrix –incorporating distributed and flexible responsibilities. Therefore, it becomes increasingly important to break down barriers to direct communication. This transformation toward streamlined, time-optimized processes has evolved from the realization that, as a result of ever shorter innovation cycles and global competition, time has become a crucial factor with regard to resource utilization and optimization. In the future rigid organizational structures will have to be dismantled to an even greater extent. They will be replaced with temporary, flexible and solution-oriented functional units whose characteristic features are openness, a high degree of networking, and the complete consistency of processes. Fractal structure, autonomy and self-organization, as well as networking beyond the boundaries of the company, will become imperative features. Internal and external cooperation in all phases of the value chain is increasing, as value chains develop into value networks. This does not necessarily mean that such companies will change as a whole, but only that those parts of these companies that are involved in knowledge-based businesses will be transformed by these streamlined features.

488

Claus Weyrich

As this trend evolves, it will lead to so-called virtual companies. Such companies will be defined by a a specific task or goal that will determine which sub-areas of the company or even which external partner companies are temporarily merged to produce an optimized mix of qualities and capabilities. On the one hand, the advantages of such virtual companies – often described as “dream teams” – are based on their direct orientation to solving customer problems and, on the other hand, on their flexibility with regard to changing requirements. Both of these advantages, however, require unified and efficient business processes, without which the advantages of virtual companies cannot come to fruition. Past

Present

Function-oriented

Process-oriented

Hierarchical, rigid

Distributed, flexible

• Rigid boundaries in the company • Local optimization of resource utilization

• Removal of barriers • Optimization of the timing schedule • Across the company

Future Solution-oriented

Open, networked • • • •

Autonomous Fractal Self-organizing Beyond the company

The need for internal and external cooperative agreements is increasing

Figure 113: Knowledge-based companies require further development of organizational structures

Processes Virtual companies function particularly efficiently if the entire business is understood as a value-added process independent of organizations, which is aligned to customer requirements and extends across the different valueadded steps uniformly from customer to customer. The standardization of processes and data enables virtual teams to work harmoniously with customers, suppliers and partners, both within the company as well as across company boundaries. In addition, process and data standardization are the precondition for the utilization of standardized e-

Knowledge-based companies – objectives and requirements

489

business systems that can be used to achieve added increases in efficiency, as well as cost reductions. The systematic optimization of process quality is an additional important objective in knowledge-based companies. At the same time, a differentiation must be made between constant improvement in small steps (KAIZEN) and the fundamental rethinking and radical redesign of processes (Business Process Reengineering). Standardization and optimization go hand in hand and ensure the sustainable improvement of business, management and support processes that are determined in a so-called ‘Process House’. The objective of Customer Relationship Management (CRM) is to establish lasting and profitable partnerships with the customer. Processes are defined according to the “customer life cycle” with which customers are won, developed, retained and, if necessary, won back. Supply Chain Management (SCM) refers to customer-oriented and value-increasing design, introduction and control of processes throughout value chains from the suppliers of suppliers to the customers of customers. Hence, the supply chain comprises procurement, production and delivery of products, systems, facilities, solutions and services to the customer as well as return processes and recycling. Management Processes

Business Processes Customer Relationship Management (CRM)

Supply Chain Management (SCM)

Product Lifecycle Management (PLM)

Support Processes

Figure 114: Process House

490

Claus Weyrich

Improved company profitability resulting from creation of customer benefits across the entire product life cycle is recorded in the Product Lifecycle Management (PLM) process. The term “product” stands for all types and versions of products, ranging from components and systems to customer-specific solutions and services. PLM comprises processes for product portfolio management, product definition, development, marketing, operation and servicing, to orderly product phase-out. In addition to business processes, the management and support processes essentially provide a high potential for standardization and optimization. Company objectives are defined through management processes, guidelines are drawn up for implementation, and relevant boundary conditions are created. Management processes also steer the business and support processes. Support processes are principally used for the support of the business and management processes. They indirectly contribute to the value added, but provide a high potential for synergy. The specifications to be carried out at each level described below must meet the requirements of the individual businesses. Despite a certain disillusionment after the e-business “hype” around the year 2000, the mastery of business processes and their consistent “digitalization” are and remain a fundamental necessity for companies that operate and implement knowledge-based businesses.

Expertise Collective skills such as project team capability, knowledge management, and management by target agreements develop orthogonally to the specialized competencies and are an important lever for increasing corporate effectiveness and efficiency. They represent the core corporate competences that are not visible from outside and that are difficult to imitate, yet are often decisive for the attainment of competitive advantages. The specific development of such capabilities places high requirements particularly on top management and on an organization’s personnel management. This requires, however, that participants have an increased awareness of their skills, and that all those involved see the necessity and have the confidence that they are always stronger together.

Knowledge-based companies – objectives and requirements

Internal knowledge

Productive use of existing knowledge

• Knowledge of own products and internal partners

External knowledge

• Customers • Business partners

• Core competencies

• Competition • Benchmarking

• Project experience • Technology planning

• Publicity

Knowledge must be transferable

• Startups

• R&D projects

Generation and consolidation of new knowledge

491

• Merger & Acquisitions

• Invention patents • Task forces • Benchbreaking

• Strategic alliances • R&D co-operation projects

If any, process barriers are mainly of human origin

Figure 115: Knowledge-based companies require the right mix of knowledge management elements

One of the most important core competencies of knowledge-based companies is knowledge management, i.e. the managing of knowledge carriers and the basic processes of creativity and innovation, as well as the exchange and use of knowledge. The skills resulting from internal and external knowledge can be combined with capabilities depending on requirements, the productive use of existing knowledge, as well as the generation and compilation of new knowledge, and can be represented in a matrix. The crucial factor, however, is that knowledge must be transferable. The barriers to this process are partly technical and partly sociological and psychological in nature. If, for example, a company’s information structure is perfected, but its human component is neglected, this can create cultural barriers that reduce the willingness for open communication. On the other hand, if human factors are over emphasized while the technical infrastructure is allowed to remain inadequate, the exchange of knowledge will remain far from optimal. At the same time, it is important to be aware that only human beings can convert information into knowledge, and that this empowers them in conscious and effective action. Communication that increases the amount of knowledge does not have to be restricted to telecommunication. Personal contacts are indispensable, especially since efficient media are universally available. Above all, the task of establishing contacts between people and merging knowledge carriers into teams that can solve problems is an impor-

492

Claus Weyrich

tant addition to communication networks when knowledge management is at stake. Turning ability into constructive action requires a balanced knowledge culture whose characteristic feature is that knowledge management calls for give and take, i.e. all partners must profit from one another and experience success together.

“Human being” • Information turns into knowledge • Amount of knowledge increases through personal communication • Global networking of experts in business communities • Knowledge culture: cooperation, motivation

“Technology” • Secure login (identification and authentication) • Encrypted data transmission • Assignment of roles and rights (authorization) • Personalized representation of information

Figure 116: Key factors for knowledge management

Employees Undoubtedly, among all the factors that contribute to making a company fundamentally knowledge-based, human beings are the most important. Just as important is the fact that the qualifications employees must have in order to accomplish future tasks be (of necessity) clearly defined. If the development of employee qualifications at Siemens AG over the last 30 years is considered, the percentage of college graduates has almost tripled – and not just in technical fields. The number of skilled workers has risen to a new high. In contrast, the number of so-called “blus-collar” workers has steadily declined. Due to structural changes in industry and continuing globalization, the requirements placed on college graduates have increased considerably. An intensified need to meet tough economic goals, methods and system expertise, process-oriented reasoning, and business-related social skills are demanded, together with knowledge of other cultures if possible.

Knowledge-based companies – objectives and requirements

493

Without a doubt, a solid university education is the foundation for entry into company life. But even this is no longer sufficient to meet the mounting and changing requirements of professional life. In view of rapidly growing bodies of knowledge in virtually all fields, advanced and continuous training has become a must. Only continuing education focused on new knowledge of facts and methods, as well as new technical disciplines, ensures the necessary ability to adapt to the process of dynamic change. This means that lifelong learning will gain more and more significance in the future and that employee must have the potential for further development and for the integration of new knowledge. Thus, our creed must be: Each of us is the manager of his and her own abilities. The company can support us in this, but the ultimate responsibility rests firmly on the shoulders of the individual. 100%

Semi-skilled workers Abilities that can be used over the long-term • Fundamental knowledge • Competence in methods • Technically, rather broad

Industrial

Skilled workers

Up-to-date, business-related • New facts and methods • New disciplines • Rather specialized • Implementation-oriented

50%

Technical

Univ. + Tech. College

Commercial

advanced education Education 4 ... 8 Years

Job 30 ... 40 Years

0% 1970

1980 1990 Siemens AG

1998

Time

Life-long continuous education ensures adaptation to dynamic change

Figure 117: Employee qualifications have steadily increased

Values Changes in the structure of the working environment cannot be overcome without this readiness for self-management because it is associated with the increasing acceptance of self-reliance. The hallmark of this structural change is that work processes are losing their regulated character, and it is becoming ever more necessary to quickly master new disciplines. Even everyday tasks will become more exciting, varied and enriched – but also

494

Claus Weyrich

more demanding. In this context, companies are looking not just for star employees who can shoot to the top in a narrow field, but for people with a multifaceted track record of experience and skills from activities in other areas. Individualism, the willingness to take risks, and entrepreneurial reasoning and action are the fundamental qualifications and characteristics employees need, along with understanding of the values on which a company is founded. WE STRENGTHEN OUR CUSTOMERS – TO KEEP THEM COMPETITIVE Customers

WE PUSH INNOVATION TO SHAPE THE FUTURE Innovation

WE ENHANCE THE COMPANY VALUE – TO OPEN UP NEW OPPORTUNITIES Value

WE EMPOWER OUR PEOPLE – TO ACHIEVE WORLD-CLASS PERFORMANCE People

WE EMBRACE CORPORATE RESPONSIBILITY – TO ADVANCE SOCIETY Responsibility

Figure 118: Siemens’ Corporate Principles. An organization’s values provide direction in in the operative environment

These maxims are summarized in Siemens’ Corporate Principles. The product of experience, these principles focus on the customer, innovation, company value, employee and social responsibility, and are intended to provide direction in everyday life. Such corporate principles are essential for every company, regardless of the industry in which it is active or what type of business it operates; they are a common platform that does justice to the interests of the company and the interests of the employee.

Knowledge-based companies – objectives and requirements

495

Summary Knowledge has always been a major factor in human development. As a result, it is justifiable to ask why the term ‘knowledge’ should play such a prominent role – especially with regard to the future.

Quantity Database records

Quantity

• Around 1 billion books were published up to 1995 • ... that corresponds to a volume of information of about 106 Gigabytes

Scientific journals

• The memory content of human beings is estimated to be more than 109 Gigabytes

Abstract publications

Quantity world population

• Knowledge as a capacity for effective action is unlimited

The revolutionary new factor is the rapid availability and targeted development of knowledge

Figure 119: Knowledge has been growing exponentially for centuries – many times faster that population growth

The amount of information is increasing far more rapidly than world population. But, without doubt, modern information and communication technologies have made it possible for knowledge to become available much more quickly and to be developed in a targeted way. A typical example of this is the Internet with its integrated databases, which are available in real time and that enable completely new knowledge management processes. In view of this, in future, a new paradigm known as “Economies of Knowledge” will be added to existing strategy paradigms, which include “Economies of Scale,” “Economies of Scope,” and “Economies of Speed.” This means that companies that use so-called “intangible assets” more effectively and efficiently than their competitors will operate more successfully and will achieve higher rates of returns. The goals and challenges of knowledge-based business will shape the companies of the future. These companies will be characterized by a flexible network of cells populated by competence bearers, entrepreneurs and

496

Claus Weyrich

capital managers. In form, structure and constant ability to mutate, this network will resemble an acting, breathing, living organism. As a result, organizations will become organisms consisting of structures, processes and people oriented towards the successful solution of complex tasks in knowledge-based businesses. Moreover, these organisms will need to transform their collective capabilities into core competencies, with knowledge management first and foremost. And they will need to continuously optimize themselves as learning organisms Underpinning all of this will be corporate principles that will not only define such organizations, but provide them with stability. But the analogy of organization as organism also implies that companies developing in the direction of knowledge-based enterprises require time – a fact that is sometimes difficult to accept in the fast moving present-day. As a result, in spite of the challenging pace of change, it is a justifiable consolation to ponder the fact that evolutionary growth is healthier and more resistant – indeed, an indispensable part of the present era of globalization and growing competition. In short, Darwin’s theory of the “survival of the fittest” applies today more than ever. Economies of Scope

log of costs

Return on equity

Economies of Scale Cost leadership

log of accumulated quantity

Economies Return on equity of Knowledge

Performance leadership

Use of intangible assets

Return on equity

Economies of Speed

Budget overdrawn Delayed product introduction

Costs Time

Figure 120: “Economies of Knowledge” will become a new paradigm

Hans-Jörg Bullinger, Rolf Ilg

Living and working in a networked world: ten trends

When faced with major changes in the coordinates defining the world in which they live and work, people are often wary and full of reservations – initially, at least. For this reason, it is important not simply to take people on a journey into the future, but to give them the opportunity to pick the itinerary and the speed at which they proceed from one milestone to the next. This report highlights ten key trends relating to the knowledge society, demographic developments, and the increasing degree of flexibilization in our networked world – trends with a crucial influence on how we will live and work tomorrow.

Preface The future is an issue that concerns us all. The future of work in particular should not be underestimated in its importance, given its immediate impact on everyone’s lives. How will we be working in tomorrow’s world? Is telecommuting going to be the dominant work model? Will we no longer have our own offices, and will we communicate with customers and colleagues solely through mobile information and communication media? Every era has its trends, and the management teams heading up a large number of companies today have no choice but to move with the times and the dictates of the day. The transition from an industrial society to an information and knowledge society is one such trend currently impacting on all of the world’s national economies. Here, the internationalization and globalization of business as well as working and social relationships are the primary triggers and impelling forces shaping the future and introducing changes to the way we work. A number of other trends are also discernible that are responsible for initiating and driving change in our living and working situations.

498

Hans-Jörg Bullinger, Rolf Ilg

Trend 1: Globalization is intensifying international competition It would be wrong to refer to globalization as a trend these days. It is already a fact of life. This is clearly evident every day in many different ways. For example, national sales and labor markets are being drawn deeper and deeper into a fiercely competitive global business arena. Hitherto largely protected, these markets are now shining a merciless and revealing spotlight on national shortcomings and weaknesses. At the same time, though, globalization is creating opportunities for entirely new forms of collaboration and cross-border business models. One aspect of this development is that the global availability of products is forcing businesses to fundamentally rethink how they approach their markets. Products’ quality, capabilities, and technologies have become broadly similar worldwide. In order to carve out sustainable competitive advantages in the future, businesses must now set their sights on streamlining their business processes – by merging or networking with other companies if necessary. At the same time, the privatization of large state monopolies and the ensuing deregulation processes have caused considerable turbulence in markets. The key now is to learn to ride this turbulence by trimming corporate strategy for greater flexibility, managing customer relationships tightly, and developing networks of competency.

Trend 2: Information and knowledge as productive factors For companies, knowledge is becoming a crucial growth and employment factor. This is partly evident from the fact that the jobless statistics do not just carry people with fewer qualifications but also well-trained specialists who possess skill sets other than those in demand in the labor market. So clearly, obtaining highly specialized expert qualifications is not sufficient on its own. Instead, it is important to be able to respond to the market’s needs by rapidly acquiring and offering the exact competencies it demands. Thus, the ability to develop market-focused knowledge fast and flexibly is becoming a crucial capability in the knowledge society. The term “knowledge society” is rarely defined with any degree of precision. What are the defining characteristics of a knowledge society? For simplicity’s sake, we can say that the term knowledge society or knowledge-based society pertains when value-creating processes depend to such

Living and working in a networked world: ten trends

499

a large extent on the knowledge of individual people or companies that the cultivation and management of this knowledge takes precedence over other factors. In such a situation, society’s and companies’ efforts center on developing and utilizing this knowledge and on perpetuating it by promoting lifelong learning.

Trend 3: Information and communication technology is advancing at breakneck speed Deploying information and communication technology today serves two primary goals: first, to ensure that information is available fast and in a high quality worldwide, and second, to underpin business processes. Electronic business and collaborative business today represent a range of activities that have leveraged Internet technology to introduce new ways of processing business transactions within and between companies, and will continue to do so in the future. In this context, the Internet’s potential has by no means been exploited to the full. This is due in part to the speed of Internet traffic, which is still too slow, and partly to low availability. In addition, businesses need time to adjust to working with suppliers and competing with rivals in virtual marketplaces. It is safe to assume that as the Internet becomes faster, virtual marketplaces in particular will experience high rates of growth. Here, it is important that standards are put in place that enable companies to exchange a comprehensive range of product data. Germany is playing an important role here, driving forward standardization at both the domestic and the international level. Web services, too, constitute an important trend – hardly surprising, given that web-based software components, besides opening up new and uncomplicated e-business channels, also promise to lower IT costs and simplify the problems of enterprise application integration (EAI) in the intermediate term. The possibilities afforded by integrating internal and enterprise-wide applications to optimize business processes make web services especially interesting. Although the technology is still in its infancy, it clearly has enormous potential. In many areas of the enterprise, information and communication technology has become a mainstay of efficient and effective business processes. IT, for example, has made it possible to network companies; at the same time, though, it represents a risk. Before businesses can put their faith in ITsupported business processes and communication, they need security –

500

Hans-Jörg Bullinger, Rolf Ilg

security built on reliability, authenticity, integrity, and confidentiality. If these prerequisites are not met, there is no value in rolling out specialized applications to handle e-commerce, e-banking and telecommuting, or to protect sensitive business information. IT security therefore is an issue not be taken lightly and should occupy not just software makers and network operators but also corporate management everywhere. The case of the Slammer computer virus underscores just how vulnerable IT can be: When it struck in early 2003, it infected around 75,000 systems within the first 11 minutes and then made its way onto 1.4 million hosts within just a few hours, even taking out banking terminals and forcing airlines to cancel flights. The ferocity with which the virus spread shows clearly that it is not sufficient to take a purely reactive approach to security-critical incidents and that all concerned need to be proactive and aware of their responsibilities.

Trend 4: Less manufacturing work, more knowledge-intensive services The services sector is assuming the role of an innovation catalyst to an increasing degree. The growing importance of services in a competitive context and in the labor market is arguably the most prominent feature of the economic and societal transformation currently underway. The services sector remains a crucial engine driving economic and employment growth. In Germany, the number of people in manufacturing work will drop to below 25% of the nation’s workforce, whereas the number of those employed in knowledge-intensive services will increase sharply. Numerous American corporations along with a growing number of companies in Europe are showing how systematically developing customer-focused services can be a key to success. Frequently, companies do more than just provide a particular service; instead, they offer so-called hybrid products that consist of product hardware in combination with one or more related services. Just a handful of companies have been quick to recognize that, in light of increasing globalization, merely maintaining a lead in technology, costs, or product quality will not be sufficient in the future to maintain a strong competitive advantage. By offering innovative services, they can create a unique selling proposition that sets them apart from their rivals and can pursue a strategy capable of tapping new market potential. Three particularly important areas have emerged in services sector research. Alongside benchmarking (the standard means of evaluating serv-

Living and working in a networked world: ten trends

501

ices and identifying service gaps), service engineering and standardization have now achieved prominence in Germany. Today, specialized methods, classifications, process and marketing models, and related standards for developing and rolling out new service products have become established. Businesses looking for better ways to exploit opportunities in the services market need to further improve their base of knowledge by systematically exploring, absorbing, and embracing international examples of good practice. This calls for a change of mindset in different areas: In industry, for example, companies and employees who understand the essence and the value of service and who see their future in serving customers need to commit to it. In the political arena, responsible and courageous decision makers who regard themselves as agents serving Germany as a country and as a business location need to assign high priority to the creation of a favorable environment for service providers and services. And lastly, we in society need to develop a clearer perception of ourselves as consumers and coproducers of traditional and innovative services. The same understanding of the direct association between a systematic development approach and success in the marketplace – gained in the context of classic hardware products and, in the last 20 years, with software products – will gradually take hold in the services sector, too.

Trend 5: Value-oriented company management The more intangible products become, the greater the role played by values and corporate culture in conveying these products’ qualities to society. What do we mean by “values” in this context? Values are abstract notions representing fundamental, central and general objectives and guidelines pertaining to human conduct and co-existence in social groups. Values serve as standards guiding the direction, goals, intensity and means we choose when taking any form of action. They underpin, justify and give meaning to the social norms that are an essential aspect of coordinated and predictable behavior. In a corporate context, they are company-specific, mutable, can be shaped consciously, and are elemental to corporate culture. Corporate culture is an abstract and company-specific phenomenon encompassing the values, senses of direction, and cognitive capabilities shared, upheld, and practiced by the majority of employees. In a company, these values create a foundation for people’s decisions and actions. Strategies build on these values and are thus congruent with the culture that pre-

502

Hans-Jörg Bullinger, Rolf Ilg

vails in the company. Corporate culture and values are among the foremost factors that motivate employees and encourage them to empathize with the company’s identity. If executives are seen to embrace and embody these values credibly, they can succeed in uniting the company and its employees within a nexus of values that fosters higher levels of motivation. By pursuing a common line in the values they represent and the goals they set, managers can engage employees emotionally and encourage them to feel part of the company.

Trend 6: Demographic change: people are living longer A look at demographic trends reveals that the declining birth rate and steadily increasing life spans will lead to a marked change in the age structure of Germany’s population in the intermediate and longer term. Forecasts point to a drop in the population and an attendant decrease in the number of younger people capable of gainful employment. Against this backdrop, industry can expect to see both a shortage of suitably qualified workers when recruiting new labor and a corresponding rise in the mean workforce age. Companies need to begin responding to this challenge today in order to adapt to the changes ahead. One serious issue that should be addressed in this context is the still widely held perception that older people are less capable and less innovative in their thinking. Recent research, however, offers evidence that diminished performance and innovation skills are less a function of biological aging and more the result of whether people are called upon to accomplish challenging or undemanding tasks in their working and private lives. For companies, this means investing more heavily in the immediate future in the development of human resources and in the provision of onward training for their gradually aging workforces. Simply recruiting young blood will not be enough to secure a steady influx of new knowledge. However, it has been determined that younger and older employees’ different qualifications and experience levels can complement one another well.

Living and working in a networked world: ten trends

503

Trend 7: Virtualization and decentralization: virtual company networks Many companies are currently gearing up for global competition by merging with and taking over competitors. This trend is giving rise to global companies that are leveraging their worldwide presence to develop, manufacture, and sell products and services worldwide through extensive networks of locations and business units. Researchers at the Massachusetts Institute of Technology (MIT) mapped out a scenario in which companies attained a scale and power equivalent to those of countries and had partially supplanted countries. The people heading up these companies were potentially able to wield greater power than heads of state. In the scenario, people lived, worked and consumed within the confines of “nationalistic” and seemingly rigid global corporations. As tangible as this scenario might seem, current trends show that merged companies are by no means rigid or bureaucratic and, to an increasing degree, are exhibiting decentralized, networked structures that empower employees by making them self-responsible and giving them the latitude to organize the way they work. Flattening hierarchies and introducing principles like self-organization and self-control will be defining features of organizations as they adapt to deal with markets’ and rivals’ flexibility. It is also evident that these large corporations are outsourcing many of their manufacturing or sales activities to suppliers, or are handling these tasks in collaboration with partners. In some cases, these companies are keeping just their core competencies in-house, such as research and product development, strategy development and financial control. Other value creation processes are placed in the hands of specialist external suppliers. Extremely small, independent units – so-called virtual companies of the kind now not uncommon in the media and software industries – are growing up alongside large corporate conglomerates. These virtual companies are alliances between independent, self-employed entrepreneurs often operating as one-man proprietorships. They join forces for a limited time to accomplish specific tasks or fill particular orders. Their goals are determined solely by the objectives of their current business venture. Certain technical prerequisites need to be met so as to ensure that these alliances can operate without being tied to a specific location. The immense economic advantage of this form of organization lies in its flexibility to respond rapidly and at short notice to changes in the marketplace (e.g., to demand for new fashion articles or for time-critical development projects)

504

Hans-Jörg Bullinger, Rolf Ilg

and the leeway it gives to take on board additional expertise affordably for a specific time and task.

Trend 8: Miniaturization and individualization: the smaller the better, but ultimately it is the customer’s choice The trend toward miniaturization in many products – in information and communication technology, in particular – continues unbroken. Miniaturization has become possible because research has uncovered ways to access the individual elements that make up matter, and has understood how these elements organize themselves, paving the way for industry’s conquest of the nano dimension. Current microsystems will be reduced further in size through the use of new physical instruments and procedures. Key areas of application include energy technology (fuel cells, batteries, solar cells, gas storage, etc.), environmental technology (materials cycles, disposal, cleanup, etc.) and information technology (high-density storage, high-performance processors, etc.), as well as health and aging. The pace of innovation in nanotechnology has led to a situation where products are already available on world markets even though basic research is still being conducted. Achieving further miniaturization will call for new manufacturing technologies. The smaller a device, the more important the usability issues become. In the future, for example, extremely small mobile end-user devices will only find acceptance among consumers if buyers can identify the actual benefits of miniaturized equipment and can operate it successfully. Today’s customers expect products to be tailored exactly to their needs. This applies not just to everyday necessities but also to products with a more long-term character. One prominent case in point is the automobile, which is now regarded as a showcase product made by an exceptionally customer-focused sector of industry. Also, more and more services, likewise tailored to customer needs, are being offered in connection with this product. For companies, this trend has had a marked influence on manufacturing and business processes by forcing a transition to sustained flexibility. Greater flexibility can only be achieved if accompanied by individually tailored employee training initiatives, and in the future, a commitment to lifelong learning will be increasingly important.

Living and working in a networked world: ten trends

505

Trend 9: Flexibilization: work when and where you like Recently, the notion of working within a centralized structure, at a fixed location and at fixed times is being called into question. Does the work we do have to be accomplished in this way, or will it be superseded by a style of working marked by temporal, spatial and structural mobility – a model in which we can work when and where we like? One precursor to this trend was flextime, which was introduced to give people greater flexibility over the hours they worked. In the 1990s came spatial mobility. For many, working from home, at customer locations, and on the road is now an everyday part of working life. The process has been advanced still further by the gradual breaking down of rigid corporate structures and the emergence of virtual companies. This trend toward flexibilization has already had an impact on industrial and office buildings, an impact which, given time, will eventually extend to residential buildings, too. In the 20th century, housing was primarily designed to serve the needs of home living and leisure. By contrast, 21st century housing will focus additionally on working and learning. In the area of building technology, current generations of equipment will be supplanted by new networked systems with capabilities based on information and communication technology. Advanced facility management systems will control all the technology in the home. Whereas the expectations that residential housing was required to fulfill in the past revolved primarily around basic needs such as protection from the elements and the provision of living space, a life center, and a home for at least two generations, houses today are designed to meet the specific needs of distinct target groups like single people, small families, people with disabilities, and the elderly. In the future, personal and social needs – like the conservation of resources, security, care, social interaction, communication, and integration with the world of work – will play a far greater role. Architecturally, tomorrow’s offices will no longer be centralized in classic office blocks; instead, they will be regarded as nodes in a network of work processes. Offices will be required to adapt to the changing requirements issuing from markets, organizations, and technology. This will call for spatial flexibility – to accommodate different types of offices to suit different work scenarios. Several distinct types of office will exist side by side: Market offices, for example, are a form of office in which different professional groups share the same workspace and where administrative duties

506

Hans-Jörg Bullinger, Rolf Ilg

like secretarial services are pooled. Use of flexible furnishings and fixtures creates plaza-like areas where people can swap information and share ideas, work groups can meet, and new projects can be initiated. Non-territorial offices, by contrast, provide communal workspace for a group of employees without permanently assigned desks. File cabinets, libraries, work tables, and creative space are all shared. This environment fosters teamwork and informal contacts. Employees keep their documents in special containers that are parked in a special storage area. In offices of this type, end-to-end networking with information and communication systems is absolutely essential. Festival offices are intended for particularly important work, strategy meetings and similar that need exclusive surroundings to underscore their special character. Working in a dynamic and vibrant setting the promotes creativity and at the same time provides space for relaxation helps to produce better results. These are just three examples of how tomorrow’s offices might be designed. The key to success here lies in adapting offices to suit employees’ needs so as transform offices into knowledge factories.

Trend 10: Acceleration: follow the sun One prominent example of our tenth trend is global engineering, or “follow the sun R&D.” Virtual, geographically dispersed teams work in shifts around the clock to create new products in a continuous development effort. Thanks to modern information and communication technology and the enormous bandwidths it offers, engineers can move large volumes of complex data of the kind typically produced by CAD systems between locations with ease. Corporate networking on this scale promotes an extreme acceleration in the development process. The transfer of information and knowledge will accelerate to the same degree, heating up the competition for knowledge among companies all over the world It is becoming increasingly difficult for companies to build up exclusive knowledge resources and to maintain a substantive lead in this area. As a result, even major global players can be seen to be concentrating on their core competencies. Specializing in this way facilitates the development of innovative products.

Living and working in a networked world: ten trends

507

Conclusions One question raised by the trends and developments spotlighted here concerns what capabilities society and corporations need to develop in the future. The immense speed of progress in science and industry is placing increasing demands on people. Developing tailored, customer-centric solutions calls not just for exceptional problem-solving competency but also a thorough understanding of the processes, structures, and problems involved in the work customers do. This means that the ability to explore new areas of knowledge quickly and to collaborate effectively with people from other disciplines are two crucial skills for employees. The logical conclusion is that society has to place at premium on lifelong, demand-oriented learning. Companies in the knowledge society therefore need to consider how they can make efficient use of available knowledge resources, how to develop them further, and how to make these resources available to customers and to collaboration partners on a needsdriven basis. Knowledge is becoming a strategic factor in the production process and in products, and is itself becoming a product. Successfully managing this knowledge, however, calls for a degree of technological, organizational, and cultural maturity on the part of companies; this is an absolute prerequisite when it comes to networking knowledge within the company and to turning knowledge into products. The ability to utilize and develop knowledge efficiently is also essential in order to collaborate successfully with others on knowledge-intensive projects. Those companies that succeed in aggregating and making optimum user of their knowledge, both in-house and in dealings with other companies, and in tapping that knowledge in real time will dominate the marketplace with knowledgeintensive goods and services. New forms of working are revolutionizing collaboration between people. The declining importance of concentrated, solitary work compared to communicative work in teams is introducing considerable changes to the work culture. Management by objective in combination with flexible full-time and part-time working models places entirely new demands on leadership and on collaboration between people. What is necessary is a corporate culture that is based to a considerable extent on trust and can rely on managers and employees to agree on a system of common targets that provides sufficient latitude for individuals to organize their work as they see fit. The freedoms afforded by a flexible approach to organizing work in a networked world can only create benefits for companies and their employees if built on a culture of trust.

Andreas Neef, Klaus Burmeister

Swarm organization – a new paradigm for the E-enterprise of the future

With the transition to the real time enterprise, organizational models are also undergoing a radical change. Formerly, the thought process was aligned to mechanical principles where organizations should purr along like clockwork: accurate and in step. Network aficionados thought in a more socio-technical manner. Under their leadership, hierarchies flattened out, entire levels of management evaporated and highly equipped technological teams were founded. The idea of networking mesmerized an entire generation. Everything was networked with everything else. As a result, however, the degree of complexity– both internally and externally – also increased exponentially. In contrast, in the real-time business of the future, we think socio-biologically and carry out management according to swarm logic.

Scenario 2020: vision of a normal working day This morning everything looked quite quiet. Ralph Robertson stared at the screen. The business process monitor suddenly emitted a warning signal. Apparently, simultaneously in different European major cities, there was an increased demand for these new, synthetic low-carb products. This suggested at least the current sales numbers of the ‘SuperStores’ which Robertson always monitored in real-time. Robertson recognized this effect only too well. Consumers in major cities around the world exhibit a similar purchasing pattern during the day and then suddenly, on some days, pounce on specific products almost as if by agreement. This may well have changed again by the following day. This effect is called ”Magic Synchronization”. How it exactly functions, however, remains an unsolved enigma of the developed information society. Anyway it must be dealt with immediately in order to stay on the crest of the wave and to respond to the market as quickly as possible. Robertson acquires an overview of the situation. He

510

Andreas Neef, Klaus Burmeister

displays the entire stock of low-carb products in the logistic network. In this way he knows precisely what inventory stocks are available where and which products and units are already en route on trucks or by train. At the same time, the potential production capacity for the next 24 hours is automatically requested from the manufacturer. Robertson is satisfied for the present. It doesn’t seem to be a problem to meet the projected European demand. Now it is simply a question of bringing the right number of products into the shops at the right time. Robertson is a professional. He has been doing the job of independent ”market coordinator” for an international marketing consortium for the last six years. The customers trust him and his incredible intuition for the sudden caprices of the market. ”The markets are 95% psychology”, he always says, ”because the best artificial intelligence simply can’t cope.” However, in this case, he decides to activate the “dynamic logistic optimizer” anyway, a neural network for process optimization in real-time markets. It doesn’t take long until the system has determined a target scenario. According to this, a clear intervention in the current goods flow is required. The routes of a whole series of trucks must be changed, detours made to temporary storage, inventory transferred between different stores and also the production schedule at the manufacturer must be adapted ad-hoc. Because of the vast number of actors involved – many of the logistic service providers are independent small companies – this is not so easy to enforce in the short-term. He requests an “assisted conference” with the disposition agents of the largest members of the marketing consortium. Thanks to efficient presence management, this takes place within a few minutes. Robertson briefly explains the situation while the members follow the strategy simulation on their respective terminals. After a short discussion, an electronic ballot takes place. The majority of members are in favor of Robertson’s proposal and they provide him with the software agents together with the respective security access to transfer the target scenario into the logistic network. The actual agreement and control of the modified production and logistic processes, the route control and resources disposition then proceeds autonomously between the numerous participants and the vehicles in the network. At lightning speed and supported by the agents, co-operations are agreed between the logistic service providers, cargos are accepted and truckloading reorganized. Less time-critical cargos are stored temporarily at the next available logistics center instead, in order to bring low-carb products to the point of sale as quickly as possible. Robertson tracks the processes on the business process monitor. He sees the flow of goods moving across the screen as colored lines and patterns

Swarm organization – a new paradigm for the E-enterprise of the future

511

and is – as always – amazed at how highly complex processes develop from simple objectives though the principle of self organization and that such apparent confusion still creates highly efficient results in the end. Nonetheless , he also hopes of course that he was actually right in today’s assessment of the demand for low-carb products.

Survival in real-time markets The complexity and dynamics of the market compels companies in the 21st century to act and react at breathtaking speed. The ever-increasing pace marks the reality of economic life. “Real-time markets” emerge in a multitude of industries for which alacrity is the all-important factor. In a business world where each individual product is recorded and fully tracked by means of smart tags in a ubiquitous, wireless network and where, on the other hand, information regarding customer behavior on the global markets as well as from in-house business processes is immediately made available, adapted into success strategies by the competition in the shortest period of time and in which the behavior of customers and market actors is, however, even less predictable, companies are forced to relentlessly orient their forms of organization, working processes and management methods to the laws of real-time markets. At the same time, the call for speed is in no way an expression of a culture that has forgotten how to wait and which demands that its every wish is immediately fulfilled. Speed is much more an imperative precondition to successfully mastering the synchronization requirements of a complex lifestyle. The markets of the future are fundamentally concerned with the quality of time. It is a limited commodity, an object of desire. The quality of time requires the sovereignty of time. Customers of the future will react highly indignantly to the unwanted monopoly of their lives by companies and their offers. They do not wish to be restricted by defective or complicated devices, waiting queues for service telephones, poor disposition in business or inaccurate information on the delivery times of a product within the sovereignty of their own time. On the contrary, they expect intelligent products and services which support the management of their everyday lives individually, context-sensitively and with foresight. The credo of customers regarding the real-time markets of the future reads: “Simplify my life! Here and now!” And it is precisely this demand that presents companies with enormous organizational and technical challenges.

512

Andreas Neef, Klaus Burmeister

The search for the organization of tomorrow We are used to thinking of companies and organizations in metaphors. The most successful metaphor of the past was that of the machine. It was the general principle of the industrial age and functioned incredibly well in the context of developing mass markets. At the center of the company were the machines and assembly lines that dictated the rate at which the workers had to work in unison. Detailed job descriptions, strict hierarchical management and work processes that were accurate to the second were the building blocks of the machine-based organization. The optimization objective was the continuous maximization of output. With increasing saturation of markets and new possibilities of information and communication technologies, the transition to the model of network organization began during the 80s. New technical infrastructures – especially the Internet more recently – were the inspiration. Networks consist of relationships. And, in times of diminishing rates of growth, the establishment of relationships appears to be the most important source of new value-added: the relationships to customers primarily, to suppliers, to external partners and to the public. Even within the network organization relationships play a key role. Rigid hierarchies are being replaced step-by-step by networked teams and open communication structures, machine assembly workers become knowledge workers and supervisors turn into employee coaches. The idea of the team is a prevailing ideology of the network age and relationship management is its core competency. As a consequence, the markets are no longer concerned with just the one time sale of products, but also the ongoing generation of revenue flows over enduring and extended service relationships. Although the network economy first reached maturity a few years ago, its weaknesses and the limits of its reasoning and organizational models are apparent. On the one hand, customers are increasingly irritated by the constant relationship business of the companies and are, nevertheless, becoming ever more “unfaithful” towards the companies despite intensive customer relationship management (CRM). On the other hand, the resolute implementation of network organization in the companies has led to an enormous surge in the need for coordination and hence the amount of communication. In addition, the question of distribution of competency and responsibility between centralized and decentralized organizational units of most companies is rather unclear in practice, as is the relationship between teamwork and individual performance. The potential for conflict in the network organization is correspondingly high. It is very much a question of

Swarm organization – a new paradigm for the E-enterprise of the future

513

psychology and group dynamics. Managers in such an organization must take care of the spirit of their employees just as much as the business strategies – which is, however, often too much for them. In brief: the progressive internal and external networking of the company has led to an enormous increase in social complexity. On the one hand, the increase in complexity has developed a huge potential for innovation and opens up completely new commercial business contexts. Above a certain level of networking, however, it tends to work counter productively. The network organization runs inevitably into the complexity trap. It is not without reason that complexity research today is grappling intensively with the internal structures of networks. The complexity of networks has become a problem for companies. Above all, however, the complexity caused by networking can no longer be overcome by even more networking! This fact is coming a real danger for companies, above all with regard to the requirements of real-time markets. Thus, a change of perspective, a modified model, a new more powerful metaphor on which the further evolution of companies and organizations can be oriented is required. We recommend the model of swarm organization for this purpose. Because, particularly the real-time markets of the future will involve such aspects as spontaneity, flexibility and rapid capacity to act, environmental sensitivity and adaptiveness of products and services, employees and the organization as a whole. All these aspects are related in a graphic way with the socio-biological concept of the swarm. The company of the future will no longer be a machine, nor will it correspond to the image of the network. It will appear much more organic, just like a huge swarm of independently acting individuals which, despite apparent chaotic single processes, a great deal is accomplished overall – and, in fact, in an amazingly simple way.

Understanding the swarm principle This section presents an excursion into the management techniques of nature. Generally, the swarm principle can be studied in highly organized insect populations such as ants or bees. In their alliances, they master highly complex tasks with incredible efficiency – and, indeed, without advance plans and instances of central control. They collectively achieve coordinated feats, far exceeding the ability and intelligence of the single individual. This phenomenon is also known as swarm intelligence. Examples of this

514

Andreas Neef, Klaus Burmeister

can be observed in nest building, defense, care of the brood or in the search for food. Swarm intelligence rests upon two complementary axioms: the principle of decentralization and that of self-organization, based on the strict adherence to simple rules. The swarm organizes itself through direct and indirect interactions. This interaction usually runs “peer-to-peer” in an apparently random exchange between single individuals. Information is very quickly distributed throughout the swarm by a large number of such individual contacts, but specifically in those contexts for which information is also relevant to action, i.e. ants on the other side of an obstacle which must be cleared out of the way “learn” nothing about it; however, they are informed about a huge source of food which must be exploited by the use of combined forces. The swarm has no network structure with defined “nodes” and “links”. Because there are no defined tasks for the single individual, the swarm can react very quickly and flexibly to changing environmental conditions or imminent threats. Who cooperates with whom or solves a problem is decided ‘situatively’ and autonomously according to the respective problem. Each individual is quasi a source of information as a performing unit. The swarm perceives its environment with a thousand eyes or feelers so to speak, and processes this information in a collective process of action. However, the creatures do not react to the information according to a specified operating schedule, but to a set of simple, robust and adapted rules that have proved to be successful during the course of evolution for the environmental conditions of their species. All in all, complex and often unpredictable behavior of the swarm is the result of these simple rules. The swarm evolves its problem-solving ability out of a perpetual trial and error process. In swarms of insects that are organized in a collective, as Eric Bonabeau puts it: misunderstandings and coincidences are not deficiencies. On the contrary, they contribute greatly to the swarm’s success because they enable them to discover and to explore instead of only to harvest. The swarm is always searching for new opportunities to ensure their survival. This form of autonomy feeds on errors. It keeps the colony adaptable to changes in the environment. However, self-organization also ensures the robustness of the organization. Even if a number of individuals of the species fail, the group can still fulfill its tasks. The swarm principle is a survival strategy for complex and extremely rapidly changing environmental conditions. Can this maxim, however, be usefully transferred to companies? Important to note: Humans are neither fish nor intelligent insects and do not deal blindly according to fixed rules (even if the latest brain research

Swarm organization – a new paradigm for the E-enterprise of the future

515

results show that the portion of conscious, goal-oriented action and, thus, the controllability of human behavior appears to be much less than previously thought). It is not just a question of converting companies into a type of ant colony. Yet the metaphor of the swarm is productive for the organization of the future. Its contribution is to provide new perspectives and to serve as a source of inspiration. The swarm principle actually reveals new paths of how a complex organization can be designed to remain highly flexible and therefore robust by means of self-organization, radical decentralization and on the basis of a simple and fast “peer-to-peer” infrastructure in order to successfully act as an enterprise on the real-time markets of tomorrow.

Managing the swarm organization By studying the metaphor of swarm organization, companies will no longer be seen as closed systems with distinct boundaries, defined structures and one-off defined purposes and goals. In future, companies will be thought of much more as living social organisms, which can adapt quickly and creatively to new customer requirements and markets in an ongoing collective learning process by means of networked intelligence. A swarm company is in constant touch with its environment and is therefore in a position to perceive and to react very quickly to changes. And it is able to carry out spontaneous, unforeseen actions and coordinated changes in its course at any time. Where the model of the swarm organization in the company will lead to and which concrete organizational innovations it will result in cannot be answered definitively from today’s perspective. However, the general outlines can be sketched using some basic tenets of swarm management. The fundamental idea of the swarm company is self-organization. The management concentrates on strategy. It defines strategic objectives and develops robust ground rules, value orientation or models for the organization. In contrast, the processes for the achievement of strategic objectives arise at the operational level in a self-governing optimization and adaptation process. It consists therefore of a large amount of openness for new, unconventional approaches and solutions. Errors are not regarded as failures, but are tolerated to a higher degree as an impetus for the learning process. Swarm companies reduce centralized functions to a minimum and, in contrast, strongly favor decentralized and redundant structures. Consequently, swarm logic also argues against radical downsizing or lean man-

516

Andreas Neef, Klaus Burmeister

agement because a robust, highly flexible and sustainable organization requires a certain surplus of resources, even if this leads to increased overheads and thus contradicts short-term profit objectives. Table 9: From machine-based organization to swarm organization

Infrastructure Knowledge

Yesterday Machines organization Commodity markets Order Output Control Strict hierarchy Centralistic Information flow

Co-operation

Supply chain

Organizational model Markets Basic principle Optimization Management Organisation

Today Network organization Relational economy Flexibility Relationships Coaching Flat hierarchy Client Server Knowledge management Strategic alliances

Tomorrow Swarm organization Real-time markets Spontaneity Speed Rules Decentralization/ self-organization Peer-to-Peer Distributed intelligence Microcooperations

The employees in the swarm organization have – in the framework of binding ground rules – higher commercial competencies and degrees of freedom than today. On the other hand, there are higher requirements with regard to self-motivation and flexibility. A fixed allocation to departments or teams is more the exception. Employees in the swarm company acts like independent individuals who, depending on load and competencies, are always searching for new tasks or reacting to existing bottlenecks within the organization. They are supported in all task constellations by autonomous, interacting software agents that form part of an intelligent environment as striven for by the model of ambient intelligence. In the swarm company the ability to tackle problems and act quickly is more in demand than protracted explanation processes, hierarchies and task descriptions. Relational tasks decline in importance compared to creativity, problem-solving ability and spontaneous cooperation capability. Consequently, communication structures also change. Communication in the swarm runs along the shortest paths between the people who are actually affected by a subject. The factor of company culture is crucial in such an organizational structure that more or less ensures that the company does not

Swarm organization – a new paradigm for the E-enterprise of the future

517

collapse. The model-imprinted culture is the cement that bonds the swarm organization. On the other hand, swarm companies are only enabled by an efficient technology platform. All work and communication processes are pervaded by information technology. At the same time, the differentiation between information and communication technology is omitted in an intelligent, adaptive network infrastructure. Instead of central approaches or client server architectures, the swarm organization is based on peer-to-peer technologies in particular that accommodate its logic to a high extent while they optimally support spontaneous co-operation in decentralized structures. From that point of view, today’s concepts of knowledge management are losing relevance in the swarm company. Instead of trying to organize knowledge supplies collectively at a high cost, this concerns the active design of distributed intelligence around “networked knowledge in action”. Even the production area readily places its trust in swarm logic. An open association of smaller, more flexible mini factories, which can be equipped in the shortest period of time and in which even the smallest batches and individual piece production can be economically realized, is replacing assembly lines set up for mass production. The swarm company cooperates vertically in many ways, but also horizontally in order to benefit from the advantages of real-time markets. Instead of long-term strategic alliances, this deals more with temporary micro-cooperations that are also linked ad hoc at the operational level in order to react to a current market opportunity. This can go so far that competing companies provide employees or resources in the short term in order to compensate for differences in load to their mutual advantage. The swarm organization exploits the complexity of the environment instead of cutting itself off from it. It has strengths as well as weaknesses that must be carefully balanced. It is a vision with indistinct contours that are not well defined. On the other hand, there is no way of avoiding swarm organization in the long term. Today, as the implications of swarm logic and swarm intelligence are being grappled with intensively, it is always advantageous for management. Anyhow, nature has had a couple of million years to optimize its forms of organization through evolution. In times of high-speed business, it has to go a bit quicker than that.

Juniors Group

Young professionals look to the communication of tomorrow

The communication of the future is an exciting and inspiring subject for junior scientific and technological staff. As the progeny of the technological age, today’s twenty to thirty year olds have grown up with the enormous boom in the PC market, modern communication platforms, mobile terminals and the Internet. They are developing their own visions of how future professional and private lives could appear in a virtual world. Siemens provides the so-called ‘Juniors Group’ with its own forum where young potential managers can cooperate on internationally designed projects outside of their regular working hours. Their perceptions of the communication of tomorrow includes the scenario described below since innovations nowadays consist not only of new or improved technologies, but also of far-sightedness and the ability to rethink the familiar.

Future private communication “24th May 2015, 13.00 hrs. My name is John Doe. I am sitting in my apartment and relaxing, listening to the sound of the waves – the realistic simulation of a Caribbean beach is projected in three dimensions with high resolution onto a screen. An audio system made of special material recreates the genuine sounds of the waves and other accompanying effects – like the wind in the palms and the occasional falling coconut. Combined with a special diffusor system that releases atmospheric fragrances in this living space, I really have the feeling that I am on a beach in Jamaica. Suddenly a color coordinated pop-up window appears on the screen and a pleasing female voice asks me whether she should read out a recently arrived message from a friend. “Yes”, I answer loud and clear. The computer system reads out the message in a well-modulated and undistorted voice with the proper intonation. It is really relaxing to have a completely automated, personal communi-

520

Juniors Group

cation agent in certain situations who knows exactly the friends who are welcome to contact me in my leisure time. The basis for this is my profile – a function, which like the filter in a firewall, only puts through the desired messages directly to my current whereabouts. Other incoming messages are not, of course, discarded unless I have placed the author on my “blacklist”. Messages today rarely consist of just text. Because broadband transmission is available in the majority of households, I mostly get video messages or holograms that are replayed in three dimensions if the current environment allows. Even telephone calls are mainly accompanied by video images nowadays. Now the computer system blends in my daily diary, semi-transparent but easy to read, on top of the image of the beach. Ah yes, I am invited to my father’s birthday party at my parents’ this evening. But because I still need a present, I need to go online and look on the Internet. The presentation has changed incredibly in recent years. The times of constant clicking and the never-ending search between text and confusing pop-up windows is over. Even customer service has improved: intelligent CRM agents on the basis of artificial intelligence (AI) accompany me through the virtual department store which I explore from my sofa. I announce my wishes clearly and request information. Interestingly, some online stores offer direct staff contact to “real” agents whose quality still cannot be surpassed by an AI solution. However, the virtual colleagues are ideally suited to fast searches. A look round the online shops doesn’t help because the delivery time for the product of my choice (a camera for photographing three dimensional landscape images) would be five hours. This is not acceptable. I take my “Personal Live Communication Assistant” (PLCA) from the table and go to my vehicle. I get into the car after identification by face geometry and fingerprint. I plug my PLCA in its socket. This establishes a direct, encrypted and authenticated Internet connection to my telecommunication service provider. As well as many other services, there is now a navigation assistant that will drive my car automatically to the destination if required or will change my communication profile and give me data on my surroundings. The data request about a specific vicinity is just one possibility that we could only dream of at the beginning of this century. Then, for example, setting out in the search for a specific DVD, it was possible that it would not available in shop A or shop B or even in department store C. How frustrating! Nowadays it is easy to obtain detailed information on the available services within a specific area via the data of the surroundings. These may include restaurants, hotels, garages, and also the latest in-stock article in the relevant stores. Each of the services can be requested by means of an entry

Young professionals look to the communication of tomorrow

521

in dynamic directories via the PLCA. It’s a little more complicated with consumer goods: each item is fitted with a so-called RFID chip (Radio Frequency Identification). In this way, every advertiser can provide information ‘on request’ concerning the availability of the article as well as additional information. This information base saves a lot of time and a futile search is a thing of the past. So I also inform the PLCA that I am looking for a specific 3D camera. Using specific criteria, my system works out within 10 seconds the cheapest supplier, the shortest path or ideal location to purchase the new camera, for example, and also guides me there...”

John Doe: animated vision of the future This example from the life of John Doe is a long way off, but in ten or twenty years communication will take place as described or in a similar way. However, this will require a series of changes to the technological basis. Today work is being carried out on this diligently around the world.

New interfaces between man and computer Today’s information and communication systems mostly have one thing in common: they are technically complex and, because of this, often cannot be used efficiently by users. Computers today often only react to keyboard input and mouse movements. Instead of using the natural vocal communicative ability of the user to control the equipment, special operating techniques must be learned. Indeed the further development of communication technology is increasingly dependent on the existence of efficient man-machine interfaces. As physical interfaces, however, the human voice – as well as mimicry, gesticulation and haptics – can be used for control. Even the use of human brainwaves is also conceivable. The human being has five direct input channels: sight, hearing, smell, touch and taste. Output in the opposite direction takes place over the two communication channels: voice and physical movements of the body. In contrast, machines have many possible input and output channels whose use however quickly hits the limits due to the limited ability of humans. Even indirect human communication channels such as line of sight, reasoning or emotions can be exploited in future for man-machine interaction. Different research groups are already working on an ‘affective’ computer

522

Juniors Group

with the ability to record human reactions and to detect behavior patterns. Sensors in clothing (wearable computers) register biological and physiological stimuli directly on the human body, these are then evaluated, processed in a model and converted into interaction. Thus, the system learns from the user and increasingly adjusts itself to him over time. Affective computing enables more user-friendliness and more efficient work through less errors and faster information processing. Biometry also uses new interfaces and thus enables security applications based on unique human characteristics. In addition to a person’s signature and voice, for example, these can include the fingerprint, geometry of the hand, shape of the face or structure of the retina. With the brain-computer interface already under trial, the computer can even be controlled by the power of thought. To do this, the user wears a helmet with 128 electrodes, which are placed on the scalp, and his brainwaves are measured. The signals of this electroencephalogram (ECG) are amplified and, after filtering out glitches, are transferred to a computer and then converted into control commands using specially developed learning and signal processing algorithms. Intelligent software agents, already under development in many research centers around the world, represent an additional approach in the field of human-machine interaction. In future, they will become capable of carrying out negotiations, of fulfilling orders and also of communicating autonomously with other agents. The crucial difference to conventional computer programs is that the virtual problem solver is equipped with a profile of the user wishes, can become independently active and can make independent decisions. Agent technologies support the user by assuming routine tasks as his representative and providing him with recommendations for decisions. At the same time, they record changes and react to external events. Different digital assistants are active in personalized agent systems. The application agent is like a middleman. He must analyze and interpret the tasks of the user, then decide which function agents are suitable for working out a solution and in which sequence they should be activated because these have the special knowledge for their respective function, such as an Internet search function or navigation. Therefore, the agents must be in a position to cooperate and to further delegate partial tasks. An important precondition for the communication of such a software robot is a “smart” Internet. This semantic network not only sends characters but also understands their significance. This makes it the ideal biotope for agents who require a reservoir of knowledge sources.

Young professionals look to the communication of tomorrow

523

A particularly intensive experience of the man-machine interaction is provided by virtual reality (VR), i.e. immersion in a computer-generated three-dimensional environment. Not only can this be carried out using a special data helmet or through projections, but also through holographic objects “floating in the air”. VR applications make it possible to work with holographic projections in the same way as real objects. Their potential uses are manifold – from virtual product development and building simulation to support of operations in the clinic. VR technology enables completely new services, contributing to increases in productivity by significantly shortening the long path from concept to finished product. Even the draft version is available in virtual space in three dimensions. The subsequent product can be examined from any perspective, modified by disparate teams if required, and tested as often an necessary until it meets all requirements. Tele-immersion is a mix of real and virtual environments for presentation as well as for interaction. The application fields of these mixed realities range from Telemeetings and Collaborative Engineering & Design, and Teletraining and medical applications to entertainment programs. Communication channels between man and machine Five direct input channels: sight, hearing, smell, touch taste. Two direct output channels: speech movements (of the body)

In addition to direct communication channels, the man-machine interface of the future will also address indirect channels (e.g. line of sight, brain: reasoning, emotions, ...)

In contrast, machines have any number of possible input and output channels. They are limited in man-machine communication only by the number of addressable human channels.

Figure 121: Man-machine communication: How humans and technology can interact

Even in the case of ‘Augmented Reality’, real images have computer data superimposed on them. This enables hidden information to be made visible on objects of all types. Whether with data glasses or a holographic projec-

524

Juniors Group

tion – virtual data appear superimposed on the actual surroundings. Thus, for example, context-related virtual building plans can be projected directly before the eyes of the fitter. He no longer needs to pore over a thick pile of paper instructions.

The ubiquitous computer Communication via the Internet usually takes place today using special devices such as the PC, organizer, telephone or mobile phone. They represent the eye of the needle that severely restricts the exchange of information and do not suit human habits. Why do we not obtain digital information directly from our surroundings? From things that we are familiar with – such as furnishings, clothes, household devices, light switches or book covers? Then, for example, the washing machine itself would know when it should stop spinning for particular textiles. Or the shelf in the supermarket that lets the sales personnel know when it must be refilled. Researchers around the world are busy today with the vision of ‘ubiquitous computing’. So-called ‘Embedded Devices’ are embedded in larger systems and take over control functions, fulfill specific tasks and communicate wirelessly with one another over the Internet: The miniature computer has sensors and actuators to monitor and steer specific processes on its own. A textbook example for the increasing spread of these invisible aids is the modern automobile: Here, embedded devices already take over engine management and control safety systems such as airbags and ABS or can be found in the GPS navigation system. As a result of integration in communication networks, they will soon also be able to exchange information and to access resources on the Internet at any time. Even in the near future such communication-capable microprocessors will be manufactured very cheaply and introduced into the environment in their millions or invisibly integrated in objects. Embedded devices are advancing the concept of ubiquitous computing primarily in three subject fields: SmartTags are small radio-based objects based on RFID transponder technology (Radio Frequency Identification) that support identification and logistic processes in industrial applications. SmartDevices will soon be found everywhere – in clothing and in furnishings, in industrial and in private applications. In the SmartOffice, workplaces fitted with embedded devices will lead to more efficiency and improved working conditions.

Young professionals look to the communication of tomorrow

525

Smart Devices will give everyday objects that extra something in future – because they cannot just communicate with humans and other “smart” objects in a suitable way; they know where they are at that moment, what other objects are in their vicinity and what is currently happening in their surroundings. There are many possibilities for use: from the support of inspections for service personnel in industrial plants, and contact-free reading of bus tickets and IDs, to intelligent household equipment and means of transport. For example, the use of specialized sensors – perhaps for noise, infrared radiation, vibration or smell – is conceivable for the support of inspection, or even the three-dimensional representation of objects under inspection. And the forwarding of inspection plans to service personnel as well as guided tracking with automatic logging. Clothes made from material containing conductive fibers have been experimented with for a long time in the field of wearable computing. These fibers, which change their electrical resistance when stretched, enable new man-machine interfaces. They record movements of the body and cause functions to be triggered, for example, by slight tugging of a piece of clothing. What is also exciting are the developments in the area of ‘Body Area Networks’ – here the human body itself is used as the medium for the transmission of very low currents. Just touching a device or object alone can transmit a unique identification – that is fed from a wristwatch to the body for example. Access authorization for example can be issued in this way to enable individual configuration of devices to be carried out or services to be billed.

One terminal for various tasks While it will certainly take a few decades for the complete realization of ubiquitous computing visions, other developments are gaining a faster entry into our everyday lives. Perhaps even the unification of today’s disparate communication tasks into a central terminal. In the life of John Doe, this is the Personal Live Communication Assistant (PLCA): • PLCAs replace today’s PDA (Personal Digital Assistant) and the mobile phone. These devices exchange data via a variety of media (e.g. WLAN, UMTS enhancements, etc.). A radio-controlled mini microphone (e.g. in the shape of a tasteful tiepin) combined with an earpiece enables high quality communication.

526

Juniors Group

• Moreover, the PLCA provides a virtual keyboard where light pens can be used for data entry. • An integrated mini beamer projects 2D images in the quality we know today from high performance systems for giant screens. • A virtual reality interface makes it possible to connect a helmet for experiencing virtual reality and a data glove directly to the PLCA. • Even an electronic heads-up display on which information is projected and that could also be used for a video conference are part of the product range. • Of course, important office applications (organizers, etc.), MP3 players and digital cameras (2D) are also integrated in the device. • Despite their small size, these components have the same possibilities as today’s PC.

Figure 122: This is how the working environment appears with the PLCA when no other visualization devices are available

Working across borders Looking at the working environment, each company uses internal communication mechanisms, processes and technologies in order to guarantee effective information exchange internally as well as with partners and customers. The employees of large international companies nowadays often work in so-called virtual teams that are either engaged in joint projects within a country or across national borders. In 2004, many technologies are already being used to enable and facilitate virtual cooperation: messaging,

Young professionals look to the communication of tomorrow

527

document sharing as well as telephony (IP or PSTN-based) are normal means for exchanging information. However, their possibilities are still greatly restricted: • Usually only voice is transmitted in a virtual meeting, where the quality is very often only average and perhaps occasionally even poor through delays, particularly in the international arena. • Documents can normally only be changed by one person and are solely provided for reading (“read-only”). • There is no common area where they can be creatively processed – the possibilities of a writing desk are limited in this respect. • However, in certain situations, a real meeting is essential such as in the case of a workshop where high travel costs may be incurred in the worst case. • Many of the technical approaches today are limited in use due to the low bandwidths available. • The user-friendliness of conferencing systems and document sharing is not always optimum and potential man-machine interactions are very restricted. On the basis that information technology is developing along the lines of Moore’s Law, there are new technologies that clearly alleviate or even eliminate these disadvantages. One of the most interesting is VR – Virtual Reality. The following situation is a possible example: A virtual team develops a concept for a customer project. All team members come from farflung corners of the country. The workshop manager creates a virtual room and then defines a provisional “environment”. This can be a virtual office with an appropriate set up or John Doe's Jamaican beach – which the team members would certainly prefer. Then a virtual table and a presentation screen are added to the beach. Now there is a virtual location – a place for creativity and motivated work. A further advantage is that no office space needs to be rented when looking for this locality. Equipped with a VR helmet and a data glove, any employee can participate in the meeting at any location with an appropriate network connection. All presentation tools are available virtually, e.g. a pointer pen, a laser pointer or a virtual flipchart. The documents relevant to the project are simply loaded in the conference. Such a scenario can be expanded: new products can be developed in the virtual room, models can be made to come alive, ideas and products can be

528

Juniors Group

presented. Not all members in such a virtual reality need to be human – even artificial intelligence in the shape of software assistants are helpful in such a project. Team building could now take place in the VR. Even knowledge management and training can be realized more graphically, more realistically and more intuitively. In addition, many barriers to virtual cooperation can be Figure 123: Example of an advanced video reduced by VR. Such traits as conference – even outside a virtual room mood and mimicry are recorded by VR sensors and reproduced in the virtual mirror image of the user. And, if the common office can be replaced by a comfortable environment, this creates a high degree of motivation for participating employees – a factor that is often underestimated nowadays. So that projects in the virtual room are not too strenuous for the eyes of the members, the presentation in the helmet system provides high quality with the finest resolution. Work in such virtual rooms can only be made comfortable in the long run using appropriate high quality and high-resolution displays.

Key technical factors for company communication of the future Today, the first approaches are being made to deal with communication traffic and its related challenges. Visions such as LifeWorks from Siemens and solutions such as the HiPath OpenScape real-time software suite are important cornerstones in supporting focused, effective communication. The HiPath OpenScape user has, for example, a program available on his desktop where he can individually manage his communication and availability. His personal profile can specify to the entire network via which media he will be available and by whom. In this way, he prevents futile contact attempts and, consequently, saves valuable time for customers, business

Young professionals look to the communication of tomorrow

529

partners and colleagues. Conversely, he can always be kept up to date on the availability of his colleagues and can contact them directly or to call up an ad-hoc audio or video conference. It doesn’t matter whether the employee participates in the team conference via his mobile phone or via an extension in the office. With HiPath OpenScape for example, it is also possible to allow authorized callers access to his own appointment book. Therefore, for example, the system allows an important customer to enter the next free appointment without needing to reach the employee or his secretary personally. The software informs the addressee of the new appointment and sends an automatic confirmation to the customer. The LifeWorks concept enables the user to work in a considerably more efficient way. He retains full control of his availability at all times so that he can combine his professional and private life better. However, visions alone do not solve today’s problems. The OpenScape software, for instance, must be available across all networks and all media, including Wireless LAN (WLAN), UMTS, etc. Ideally, the support by carriers goes beyond the company boundaries as well adding the accessibility of all terminals and PCs with different operating systems. Open standards and interfaces to other applications, scalability and high availability are vitally important for the guaranteed future of a communication solution. One of the future’s most important factors in information technology will be security. Currently we are undergoing the strongest independently acting attack codes in the history of information technology. IT security is a continuous cycle made up of the combination of technology, processes and humans. This chain is only as strong as it’s weakest link. Different factors decide the issue in protecting computer systems against potential attacks in the future: • Future products, primarily software applications, will include more integrated security features than ever before. • Scan & Block technology will become an essential component of IT infrastructure components. • There will be enormous growth in the area of security services. This includes both the professional and managed services segments. In the long term, a change from transport protocol Ipv4 to Ipv6 (also IpnG) is to be expected. This change will probably not be implemented in a Big Bang Rollout. In fact, this development will be implemented in mixed

530

Juniors Group

forms and step-by-step transformations. It will be interesting to see who is the first to have the courage to take this step: company customers, carrier customers or product manufacturers. In any case considerable improvements will be achieved by this change: • The protocol security will be reinforced by corresponding IPSec integration. This is already implemented in the standard. • The protocol also supports QoS (Quality of Service) mechanisms. • IP address space shortage will no longer exist. As a result, existing problems related to diverse communication protocols (such as H.323 and SIP) with NAT (Network Address Translation) will be solved. It will also be interesting to see which product provider will appear in future with the solutions demanded by the market. Only one thing is certain: information technology will continue to change. The success of the technologies used will be influenced by additional added value for the company and acceptance by their employees. This is also true of the end user market. Customer orientation, time to market, quality in the face of rapid prototyping and faster innovation cycles, cost savings and, most of all, the technological innovations themselves will be the determining forces.

Georg Berner

Information and communication in 20XX

The following article is taken from the book “Management in 20XX” (ISBN: 3-89578-230-0) by the author Georg Berner, published by Publicis Corporate Publishing. On the basis of the recently commenced conversion from an industrial to an information society, Berner’s initial theory is that nothing but systematic and holistic involvement with the future can provide the necessary guiding principles for commercial trading today. Only in this way can the business world react lastingly and quickly to the change – in other words: industry helps shape the change successfully. In order to provide a fund of new ideas, strategies and perceptions for the decision maker, Berner links the technical and social perspectives already foreseeable today with the reality of present day living spaces. In the area of discord between vision and reality, fields of action are opening up which today’s companies must occupy in order to survive in and through the change. In the abridged version of the chapter “Information and Communication” reproduced here with the kind permission of the publisher, Georg Berner deals with the development of communication technology and data transmission. A crucial benchmark for the development of existing network technologies is the data volume that the human being himself can process in his neural network. As long as the available solutions cannot match the 250 Mbit/s transmission capacity of the human being, unwanted delays or losses in quality will occur. According to the author, this provides development potential for network and access technologies.

Information and communication As we make our way into the 21st century, we can no longer shut ourselves off from the innovations of the information age. Its effects are evident in all aspects of our lives, and many people are asking what it has in store for us next. It is useful in this context to begin by considering the demands placed

532

Georg Berner

on networks, because the evolution of networks is driven by the information and communication requirements of people and machines.

Human, man-machine and machine-machine communication Most consumers are not terribly interested in network technology; they take as little interest in the digital nervous system as they do in our own. When they use the Internet, people primarily just want to communicate reliably with others, obtain information or engage in secure transactions. Basically, they just want everything to work without them having to concern themselves with the underlying technology (see figure 124).

Connecting people and machines Connecting people

Connecting machines

Communication

Network Information

Services

Applications

Figure 124: Demands on the network

Corporate processes

(Mobile) bandwidth, quality of service, security

Electronic/mobile business

Information and communication in 20XX

533

But it is not just people who need to obtain information and communicate. In our everyday life there are more and more situations in which there is an information chain between machines. Take a drinks machine, for example. Wouldn't it be nice if the machine notified the supplier before its supplies ran out? The supplier's computer would then automatically work out a route for the delivery vehicle, taking in all the empty drinks machines. Banks are already beginning to move in this direction with their automatic teller machines. This example provides a good illustration of how communication between machines can make our lives easier by relieving us of routine work. Regardless of whether we are talking about communication between people, between machines or between people and machines, the essential prerequisites are security, reliability and broadband networks. The latter are required in order to handle the high levels of data traffic involved. Unfortunately, the Internet of today is still unstable, unreliable and insecure. The table on the opposite page shows the requirements that the various interest groups will have of networks up to the year 2020 or so. However, further advances in communication networks depend on more than just our requirements. In addition to technological progress, as outlined in the previous sections, the development of the kind of applications that will demand such advances is also essential.

534

Georg Berner

Table 10: What we can expect from communication networks in the next 20 years Bandwidth

Access

Quality of service Immediate, uninterrupted connections, bandwidth as required

Payment

Growing demand for more complex accounting and bookkeeping, payment per bit, basic monthly charges, payments through third parties possible Little difference to today

Services always ready for use, cheap access for large sensor and control networks, new, sophisticated applications (providing medical assistance, for example) Support of current market developments with new devices

New services

Consumers

Realistic virtual Access from worlds with inter- anywhere, the activity last few meters often wireless, multifunctional applications

Corporate customers

Extremely high demand for broadband capacity, simulation of alternative solutions

Wireless access in the building, integration of wireless and wired intelligent devices and users

Providers of telecom applications

Rapid increase in high bandwidths, bidirectional interactive entertainment and interaction, more communication between machines

Networked portable devices commonplace, high bandwidths with high mobility through virtual reality

High quality of services for all applications, high service availability

Vendors of telecom accessories

Typically follow the demands of the market and applications

Systems that demand highquality services

Strategies of telecom service providers

More bandwidth probably needed, but will attempt to exercise sophisticated control over capacity utilization

Timely testing of new equipment and technologies soon after development, financial support for applications with multiple access through providers Will try to support a diverse range of multifunctional devices

Will deliver perfect quality to enable bundled and higher-value services, will attach greater importance to the security of services

Will offer invoicing for complex services, possibility of free services, basic charge and hidden costs will increase this requirement

Will offer services that bring higher income, will create greater network intelligence to control service offerings

Content providers

Will force demand for bandwidth up, initially little need for interaction

Will penetrate the market for computers and devices with their content, content will extend beyond the entertainment sphere

Normal to high demands in terms of service quality

Will strive for accurate, automatic tracking of usage (necessary because of the increasingly complex situation as regards rights, fees, charging and subscriptions)

Interested in services where utilization is high and regular

A simple, easily understood invoice, some free services

Multimedia services, communication from anywhere, unlimited mobility, multiple broadband connections per household High availability Simple, compre- New electronic on high-speed hens-ible display and mobile busiconnections, use of outgoings, ness models, of technologies linked to the virtual presthat are always bookkeeping ence/meetings ready for opera- system tion

Information and communication in 20XX

535

Human senses Human perception has an important role to play in discussions about the requirements placed on networks because it is a critical factor in the shaping of future access networks. What kind of bandwidths are required for human communication? Let's take as our starting point the five senses: vision, hearing, smell, taste and touch. The capacity of our senses is in no way overtaxed by new technologies, as shown in figure 125. If you add up the maximum possible bandwidths that can be picked up by each of our senses, you get a total bandwidth of around 200 Gbps, but when it comes to transmission from the senses to the brain, we can "only" manage a bandwidth of around 250 Mbps. Signals that go beyond the capacity of our senses and cannot be passed to the brain are simply omitted. Technologies such as MP3 for music and MPEG2 and MPEG4 for video permit similarly high rates of data compression for transmission. Senses

Bandwidth of receptors

Neural transmission

Eyes

200 Gbps

200 Mbps

Ears

4 Gbps

2 Mbps

Skin

1,5 Gbps

10 Mbps

Tongue

150 Gbps

11 Mbps

Nose

20 Gbps

30 Mbps

∑ =

Approx. 200 Gbps

Approx. 250 Mbps

Figure 125: The capacity of human perception

Only at speeds lower than 240 Mbps does our brain notice an information deficit. Music sounds a little flat, for example, colors are no longer brilliant, and virtual reality is no longer perceived as such. Our ears can already be supplied with a sufficient flow of information electronically, but it is our eyes that are particularly demanding: even high-quality 3D cinema films leave them relatively unimpressed. This is why work on developing improved forms of projection is being carried out around the world, involv-

536

Georg Berner

ing large, high-resolution screens, projectors or displays built into spectacles. Theme parks and high-quality PC games are moving in the direction of appealing to as many senses as possible. In addition to pictures and sound, PC control devices are now available that provide feedback, thus involving the sense of touch, while many theme parks now offer simulations of motorbikes, surf boards and so on that actually move, thus intensifying the experience. The following conclusions can be drawn from the figures presented above: • If you were simply interested in the best possible sound quality, there would be little point in going to a concert. Unless you are sitting in the middle of the front row, the audio signals you receive do not match the quality of a CD. And the other senses are hardly involved. As far as your senses are concerned, it is more worthwhile to go to the opera or the ballet because then your eyes also take in information, resulting in higher utilization of the bandwidth available to your senses. And if you round off the evening with a meal, your senses of taste and smell also get something out of it. • New computers are not being bought just for rational reasons. A processor with a clock speed of no more than 1 GHz is sufficient for most office applications; higher performance is only required for some of the more sophisticated games and videos. Consequently, it is our senses or those of our children that determine what we buy.

Network technologies So now we have some idea of the bandwidth of human senses. If we want to transmit virtual reality between people, we can assume a maximum bandwidth of 250 Mbps per person, which is very high compared to the bandwidth of current transmission systems. Broadband information and communication technologies increase productivity and improve the quality of life – just as electricity did when it was first introduced. By 2015, experts expect broadband connections to have reached 70% to 80% of households in the world's more advanced countries. Broadband infrastructures, wherever they are in place on our planet, are a significant competitive advantage, driving forward the knowledge socie-

Information and communication in 20XX

537

ty and the development of digital markets, transactions and services.

Access networks The required bandwidth for access networks can be provided in a number of different ways: via ISDN (Integrated Services Digital Network), analog modems or cable modems for cable television networks, for example. Some companies also use satellite connections. The more recently introduced transmission technologies include GPRS (General Packet Radio Services), UMTS (Universal Mobile Telecommunications System) and wireless LAN (local area network) for mobile applications, as well as various DSL (Digital Subscriber Line) technologies. BWA, which stands for broadband wireless access, and MWAN (wireless metropolitan area network) can provide data transmission speeds in the two- to three-digit Mbps range. UMTS and wireless LAN provide high-speed wireless access using the 5 GHz band for domestic connections, thus permitting both outdoor and indoor applications. Some access technologies have the characteristics of one-way streets, and these are adequate for radio and television, whereas communication and the Internet always require back channels. From where we stand today it is not possible to see clearly which of the above technologies will gain widespread acceptance. Wireless LAN has had initial success but has no chance of being used by people on the move, regardless of whether they are on foot or driving a car. Blanket coverage is also a problem, as is transmission reliability. And on a different note, it is worth questioning the extent to which we are prepared to use electronic devices in our everyday lives just because it is possible to develop them. As we have seen, the Internet is crucial to the information and communication industry. There are basically two ways of accessing it: by using a landline or by using wireless technology. What direction will the development of access networks take? Economic factors will play an important role here. For operators the access networks are the most expensive part of the overall network, which is why they use existing networks for as long as they possibly can. A widespread business model involves leasing the lines of other operators. The copper networks of telephone operating companies can transmit data at speeds of up 50 Mbps over short distances, and the average life expectancy of the copper cables used is 50 years. In view of the age of these cable systems, which were laid in the sixties and seventies, huge investments can be expected in access networks in ten to twenty years in most Western countries. By that

538

Georg Berner

time, investment in replacement cables will be necessary anyway, and the new networks will be primarily optical. Until that happens, however, the existing copper cable networks will continue to be used for data transmission, and only their capacity utilization will be improved. A number of technologies have been developed for this purpose, providing access at different speeds. Figure 126 shows the different times required to download an e-mail, a song or a video lasting an hour using these different technologies: • For a simple e-mail all the existing technologies are adequate. • To download a song or a high-resolution image, you need to use at least ISDN on a landline connection or GPRS for a wireless connection. • On the other hand, TV-quality videos can only be transmitted in sufficient quality using ADSL (Asymmetric Digital Subscriber Line) on a landline or UMTS without one. If you transmit moving images at lower transmission speeds, you have to accept a considerable loss of quality. The number of networked machines is rising much more quickly than the number of networked people, and the associated additional communication on the various networks is boosting demand for bandwidth.

Transmission networks When it comes to data transmission, the issue is how much bandwidth can be transmitted across what distance. Those who work in the field will have heard of the Shannon theorem, which showed that physical limits restrict data transmission on a conventional telephone line to only a few Mbps, whereas the coaxial cable used for television aerials can transmit several hundred Mbps. The optical fiber, on the other hand, opens up the prospect of much greater bandwidths. In the field of optical networks (photonics), the performance of systems has improved more quickly than in microelectronics. The transmission capacity of optical systems has, in fact, doubled every 10 or 12 months. There are now optical-fiber products available that permit data transmission speeds of over 7 terabits per second along a single optical fiber. Yet optical fiber is still a long way from reaching its limits. Two- to three-digit volumes of terabits per second can be transmitted over longer distances – which means the entire capacity of a human brain

Information and communication in 20XX

3k E-mail

3M MP3 or photograph

300 M 1 h video (in TV-quality)

GSM

9,6 k

2,5 s

42 min

3 days

PSTN

56 k

0,4 s

7 min

12 hours

GPRS ISDN

115 k 128 k

0,2 s

3,5 min

6 hours

UMTS ADSL Cable WLAN

2M 8M 30 M 80 M

0,01 s

12 s

20 min

1 ms

1s

30 s

Fiber

800 G

30 ns

30 µs

3 ms

Wireless Wired

Live Video Codecs begin at 32 kbps

Byte bps

539

Time for downloading

Figure 126: Download times using different technologies

could be transmitted in a second. On the whole, optical transmission systems have so far been used only in long-distance networks to connect large centers of population or countries. However, they are now beginning to penetrate urban areas. In industrial areas more and more optical fiber rings or metropolitan fiber-optic networks are being installed for business-to-business use. In the next 10 years, optical systems will slowly begin to conquer access networks. But why not any quicker? To begin with, the existing copper lines have to be replaced with optical cables, which involves huge investments. After all, 70% of the investments made in telecommunication networks are accounted for by access networks, the main cost being for earthworks. Furthermore, optical switching technologies have not yet reached maturity. A closer examination of optical transmission systems reveals that signal processing is still entirely electronic. In today's networks the only active optical elements apart from laser transmitter diodes and optical receiver diodes are optical repeaters. For the future there are visions of end-to-end optical networks in which the signals will only be processed optically and there will no longer be any place for electronics. However, it is still unclear when this vision will turn

540

Georg Berner

into reality. Today's optical components are large and still somewhat temperamental and the first purely optical transistor was only developed a few years ago. Electronics will therefore still be in use for many years to come for switching and signal processing, and it will be some time before the first fully optical long-distance network is developed and put into operation. The transmission capacities of wireless systems are also really quite impressive. Given a cell size with a diameter of 200 meters, for example, 7.5 Gbps can be transmitted per square kilometer.

Corporate networks Corporate LANs (local area networks) are having to meet growing requirements. The network hardware is often designed to meet maximum bandwidth requirements, which make it expensive to purchase and operate. By making use of "computing on demand", a company can operate more efficiently and cost-effectively. Sometimes referred to as grid computing, this involves the use of high-speed networks inside and outside the company. The computers and servers involved cooperate to distribute the load intelligently and thus improve the efficiency of computer deployment. In much the same way as utility companies distribute power, this involves collaboration between geographically distributed resources. As far as software is concerned, we are moving in the direction of Web services – universal software programs that can be used directly and made available to anyone either inside or outside the company. Large software companies have agreed on standardized protocols and interfaces for these. Consequently, in the future it will be possible to call programs as Web services over the Internet rather than have them stored locally on our PCs.

Mobile networks The latest generation of mobile networks is characterized by broadband access, efficient utilization of the entire frequency and global multi-standards. In the field of mobile access networks, in particular, there are a number of technologies to consider, such as GSM, GPRS and UMTS, as well as their subsequent generations. If you are moving slowly and within a limited area – in a hotel or an airport, for example – other technologies such as WLAN (Wireless Local Area Network), BWA (Broadband Wireless

Information and communication in 20XX

541

Access), BRAN (Broadband Radio Access Network) or MMAC (Multimedia Mobile Access Communication Systems) are also an option. Technologies for short distances measured in meters, such as radio tags, Bluetooth, DECT and WLAN, enable all intelligent devices to be networked. The intelligent homes of the future will be based on these standards. Terminal devices will also be able to set up an ad hoc network in order to work together. So far, Europe has been the driving force behind these technologies, with the USA still lagging behind a little. There are already more mobile phone users than landline users, and surveys of trends indicate that in future we will be surfing the Internet more from our mobile phones than from our computers. The first applications for UMTS networks are already upon us: locationbased services, video telephony, improved WAP (Wireless Application Protocol) and video streaming. Mobile office services, online games and video surveillance are also available. Other trend-setting applications are local television programs, video conferences, chat and music streaming. Although there are still some doubts about the business plans for UMTS, work is already underway on the next mobile network generation. The 4th generation mobile network is expected after 2012, promising speeds of 10 to 100 Mbps. This only needs to be increased by a factor of two to reach "hi-fi" quality for all our senses. The ad hoc networks mentioned above are a really clever idea, involving mobile devices that network with each other autonomously and exchange information. So don't be surprised in the future if your increasingly intelligent equipment starts talking about you!

Internet The Internet is the most important platform in the information and communication industry, and, as such, it is accelerating the process of fusion taking place between the previously separate worlds of information technology, telecommunications, media and entertainment. In the next ten years it is likely that the world's entire stock of public knowledge, including all its historical documents, will have gone digital. All the world's books, pieces of music, pictures, publications and so on will be available on the Web. And in order to cope with the immense flow of information this will entail, we will need powerful, reliable networks. If we compare the development of landline and mobile applications since the nineties, we can see that landline applications have always appeared

542

Georg Berner

120

100

80

60

40

Millions of customers in the US

before the equivalent mobile applications. The reason for this is that landline networks have always had a lead over mobile networks in terms of functionality, speed and costs. Thus, in each case, the transmission of text, pictures and audio and video files became possible over mobile networks 3 to 5 years after landline networks. But mobile networks catch up. We live in a fast-moving age. Daily we learn about new technologies, products, company startups, mergers, joint ventures and alliances. The rate of change is increasing day by day. It may be difficult for us to grasp the consequences of this increasing rate of change, but it is illustrated clearly in figure 127. New media catch on much more quickly now than they used to. It took 90 years before the telephone had 50 million users. The mobile phone, on the other hand, reached the same total in eleven years, while the Internet took only five. In brackets: Time to reach 50m customers

Internet (5 years)

TV set (15 years)

Radio (40 years)

Cable TV (10 years)

Computer (13 years)

20 Telephone (90 years) 0 1922

1950

Mobile phone (11 years) 1980

Wireless Internet (2 years)

1995

Figure 127: Penetration of the US market by information and communication technology

Some suspect that the Internet age is already beyond its peak following the industry's slump in 2001 and 2002 and the slower growth in Internet hosts. The number of Web pages was obtained on the basis of a comprehensive analysis of domain names and includes the pages of organizations and companies as well as private individuals. The statistics are published by the Internet Software Consortium, a non-profit-making organization. Even when all public domain knowledge is digitized, the Internet will still be far from complete. It will still be a number of years after that before broadband access is available everywhere.

Information and communication in 20XX

543

Next-generation networks Building the Internet of the next generation will require expertise in voice transmission (i.e. telecommunications) in addition to the existing expertise in data transmission that we already have from the Internet. There is no doubt that voice transmission will set the standards for the convergence of data and voice networks. This is because it is much more difficult for today's data specialists to transmit voice traffic using the Internet Protocol (IP) than it is for voice specialists to transmit data. The reasons for this are historical. Switching technology in telecommunications has developed at a terrific rate since the invention of the telegraph in 1840. In 1884 telephone calls were still switched by hand, but the first automatic switching systems came onto the market as of 1935, which marked the advent of electromechanical switching systems. Analog switching technologies dominated the market for a short time in the mid-seventies, but once digital switching became available at the end of the seventies, the revolution of modern, packet-oriented router systems kicked in. Data packet-oriented networks such as Ethernet emerged in the field of computer networking. So as far as stability, reliability and security are concerned, the Internet does not yet come up to the standards of the telephone. We need a better Internet – the Internet of the next generation – that is as reliable as the telephone, as powerful as the computer and, perhaps, as mobile as the mobile phone. Expertise in the field of voice transmission is of crucial importance because it is voice transmission rather than data transmission that will set the standards – both technical and commercial – for the convergence of voice and data transmission. Standards – in terms of reliability, service quality and real-time transmission – are significantly higher for voice transmission. What are the forces shaping the networks of the next generation? • To start with, there are the technological advances in multiplexing, computer performance and intelligent systems, for example. • In addition, optical networks and the growing scope of wireless communication are raising network capacities. • However, with their increasing demands for mobility and greater bandwidths, consumers are also playing their part. • Governments and regulatory authorities are demanding general access to

544

Georg Berner

networks, which they want to achieve through deregulation and privatization. • Further driving forces are applications, services and interactive multimedia services, since new applications will have to be developed for communication between machines in the course of the conversion of all data and voice traffic to IP.

Convergence In spite of all the efforts being made to achieve convergence between data and voice networks, a supernetwork to meet all needs cannot be expected to emerge in the near future. If experience is anything to go by, there is no reason to expect that one network technology will replace all others. As history has shown, electronic media have not replaced print media, for example. Similarly, televisions and video recorders have not replaced the cinema, television has not replaced radio and so on. But one thing is certain: all media will be digital. Instead, different technologies exist side by side and may even complement each other and lead to new cases of media convergence: cameras that can work as telephones, for example, televisions that can surf the Internet or telephones that can also shoot films. This principle of coexistence also applies to the field of telecommunications. In the final analysis, only what meets the needs of consumers and businesses will really catch on. And generally that is something simple that works reliably. Today's Internet is a cheap version of what we can expect in the future. The Internet of the future will bring entirely new qualities to the table. Ultimately, it will have to be characterized by the following attributes: always-on access, ubiquity, speed, multi-media, multi-sensor, reliability, simplicity, intelligence and, of course, intuitive handling.

List of authors

Ferri Abolhassan

Dr. Ferri Abolhassan studied computer sciences at the University of the Saarland before moving to the USA to join IBM. After receiving his doctorate in 1992, he spent several years in managerial positions at SAP Retail Solutions, ending this stint as Managing Director. From 2001 until 31st of December 2004 Dr. Abolhassan has been the Co-Chairman and CEO of IDS Scheer AG. In this position, he was responsible for International Business and Marketing. Shai Agassi

Shai Agassi, a software entrepreneur, joined SAP in 2001 and was appointed to the SAP Executive Board in April 2002. As head of the Technology Platform organization, he oversees development of the integration and application platform SAP NetWeaver and SAP xApps. Agassi co-leads the Suite Architecture Team with Peter Zencke, the head of the Application Platform & Architecture (AP&A) group at SAP. Georg Berner

Georg Berner is in charge of strategic marketing at Siemens' Information and Communications Executive Office. Prior to this post, he had held a variety of leadership positions in research and development, marketing and sales within the Siemens Group. At various times, Berner has served as Vice President Innovation Field, Corporate Research & Development, Vice President Worldwide Marketing at Siemens Nixdorf, and Vice President Strategic Marketing at the Information and Communication Networks Group. Before joining Siemens, Berner was employed as Worldwide Account & Sales Manager at Texas Instruments. Steve Blood

Steve Blood is a Vice President and Research Area Leader at Gartner Research. He is accountable for leading the enterprise voice agenda, which includes tracking market migration to IP telephony, messaging and unified

546

List of authors

communications. His specific research coverage focuses on multi-channel contact center applications, the evolution of telephony in converged enterprise networks and the development of communications-enabled business processes. Hans-Jörg Bullinger

Professor Hans-Jörg Bullinger is President of the Munich-based Fraunhofer Society. He graduated and earned a doctorate in mechanical engineering and later qualified to lecture in the same subject. Professor Bullinger held the Chair of Industrial Engineering and Human Factors at the distancelearning University of Hagen from 1980 through 1982. The University of Stuttgart then offered him its Chair of Industrial Engineering and Human Factors in 1982. At the same time, he took over as head of the local Fraunhofer Institute for Industrial Engineering (IAO). In addition, from 1991 through 2002, Professor Bullinger led the Institute for Industrial Engineering and Technology Management (IAT) at the University of Stuttgart. Klaus Burmeister

Futurologist Klaus Burmeister apprenticed as an electrician before studying political science in Hamburg and Berlin. Since 1986, he has held senior positions at various institutions focused on scientific futurology. Burmeister founded Z_punkt GmbH The Foresight Company (Essen, Karlsruhe and Berlin), of which he is now an Executive Manager. He advises corporate clients on social change issues and long-term corporate development. Johann Csecsinovits

Having trained as a mechanical engineer, Johann Csecsinovits joined what was then Steyr Daimler Puch Fahrzeugtechnik (now Magna Steyr Fahrzeugtechnik AG & Co KG) in 1990. 1992 brought a move to the Information Systems & Organization line. Csecsinovits has been in charge of User Service & Systems, Infrastructure and Telecommunications since 1994 and has for years been involved in outsourcing projects.

List of authors

547

Tom Davies

Tom Davies is Senior Vice President at Current Analysis, Inc. Davies' primary focus is advising technology companies on their growth strategies and competitiveness. Prior to joining Current Analysis, he held senior positions with several leading software, services, and system integration companies. Davies writes and speaks regularly on the topic of competitive strategy and marketing. Jörg Eberspächer

Professor Jörg Eberspächer is Dean of the Faculty of Electrical Engineering and Information Systems at the Technical University of Munich, where he was given the Chair of Communication Networks in 1990. Having graduated and earned a doctorate in electrical engineering, he worked in research and development in the field of communication networks as of 1976. Professor Eberspächer chairs or is a member of various research and scientific committees and holds a guest professorship at Tongji University in Shanghai. Claudia Eckert

Claudia Eckert is Professor of Computer Science at the Technical University of Darmstadt, where she has held the Chair of Security in Information Technology since 2001. She is Director of the Fraunhofer Institute for Secure Information Technology since 2002. Professor Eckert concentrates her research on IT security topics, current and future network issues, and operating software for IT systems. Jürgen Frischmuth

Jürgen Frischmuth has been a member of the Group Executive Management Board at Siemens Business Services since October 2002. In this capacity, he is responsible for sales strategy and marketing as well as for the relevant corporate functions for global sales. Frischmuth also spearheads the company's international solutions business and oversees the regions Germany, Austria, Belgium and Italy. Earlier, he had held an array of senior international posts in marketing, sales and business administration. In 1999, he took the senior management job at Siemens Business Services

548

List of authors

Deutschland. In this capacity, Frischmuth systematically transformed the company from a solutions business to a full-service IT provider that covers every link in the IT service chain. Walter Fumy

Dr. Walter Fumy studied mathematics and computer sciences. He joined Siemens in 1986 and became Vice President Security Technology at the Enterprise Systems division of Siemens Communications in 1999. He is responsible for a wide range of areas, from cryptographic matters through the evaluation of innovation content to the analysis of market opportunities for new security technologies and business ideas. Dr. Fumy is the author of numerous technical publications and has for many years held senior positions in various international committees and forums. Thomas Ganswindt

Thomas Ganswindt is a member of Siemens' Corporate Executive Committee, in which he oversees the Group's IC divisions. His graduate degree in mechanical engineering was followed by two years at the Fraunhofer Institute for Production Systems and Design Technology in Berlin. His career with Siemens began at the Automation Group in 1989, since when he has held a number of posts throughout the company. He was appointed to lead Siemens Information and Communication Networks (ICN) in 2001 and, a year later, became a member of the Managing Board of Siemens. Thomas Ganswindt is a member of the UN's Information and Communications Task Force and published his first book – “Innovationen – Versprechen an die Zukunft” – in 2004. Barry Gilbert

Barry Gilbert is the Chief Operating Officer of Current Analysis, Inc. Prior to joining the company in 2003, Gilbert spent 25 years in the high technology space focused primarily in the market research and consulting industry, where he has been an analyst, service director, consultant, and entrepreneur. He has held executive level positions with Giga Information Group, Strategy Analytics and Bowne Global Solutions.

List of authors

549

Frank E. Gillett

Frank E. Gillett holds a master’s degree in management from the MIT Sloan School of Management and a master’s degree from the MIT Technology and Policy Program. He also earned a bachelor of science degree in mechanical engineering (minoring in political science) from Duke University. As part of the Computing Systems Team at Forrester, Gillett has been analyzing strategic problems in corporate IT infrastructures since 1998. His current focus is on next-generation data center architectures and related strategies. Josef Glasmann

Dr. Josef Glasmann studied electrical engineering at the Technical University of Munich, where he also earned his doctorate in 2003 on the subject of "Resource Management for Real Time Traffic in IP Networks". Dr. Glasmann has been employed at a medium-sized company since 2004, where he is in charge of consulting, planning and development in relation to IT and telecommunications networks for corporate customers. Martina Göhring

Dr. Martina Göhring studied business at the University of Stuttgart, where she went on to work as a scientific officer at the Institute for Industrial Engineering and Technology Management (IAT). This was followed by a stint at the Fraunhofer Institute for Industrial Engineering (IAO) in Stuttgart and, more recently, by membership of the institute's management team. In May 2004, Dr. Göhring became Managing Director of CentreStage GmbH, a company spun off from the institute. Her new company focuses on e-learning, total customer relationship management, e-marketing and performance management. Manuel Görtz

Manuel Görtz studied electrical engineering and information technology at the Technical University of Darmstadt. He combines his research activities at the university with his experiences as senior development engineer at a SME (KIMK GmbH) in the area of Voice over IP and innovative communication media. In both environment he initiated novel activities in context aware computing. Manuel Görtz currently finalizes his doctorate on con-

550

List of authors

text-sensitive seamless communication services. Michael-W. Hartmann

Michael-W. Hartmann is a seasoned expert with over 20 years' experience in the hotel industry. He studied at the hotel vocational schools in Heidelberg, Germany, and Cornell, USA, before completing his MBA in London and Henley. After periods of service at a range of hotels, Hartmann switched to hotel-related consulting and software firms such as Fidelio, Hogatex and Burns & Gimble. He came to Siemens in 1997 and, in May 2003, took charge of the vertical Business Hospitality line at Siemens Communications. Claus E. Heinrich

Professor Claus E. Heinrich joined SAP in 1987 and was appointed to its Executive Board in 1996. He heads up the Manufacturing Industries Business Solution Group. Since becoming the company's personnel chief in 1998, he has also been responsible for human resources and labor relations. Thomas Hess

Professor Thomas Hess has been Director of the Institute for Business Information Systems and New Media at the Ludwig Maximilians University in Munich since 2001. Earlier, he had headed the New Media and Corporate Networks workgroups at the Institute of Business Computing at the University of Göttingen. From 1996 through 1997, Professor Hess served as assistant to the Executive Board of Bertelsmann, focusing his activities on online business. Today, he coordinates Munich-based research into the Internet economy and co-publishes trade journals. Rob House

Rob House is Head of Collaboration and Integrity Solutions at Siemens Communications in the United Kingdom. Rob is responsible for the OpenScape, Security and Data Solutions product areas. Product and Project Managers in these areas report to him. Rob also defines and implements Siemens Communications' data and security solutions strategy. Previously, Rob spent five years with Energis as Senior Hosting Operations Manager. Rob has a BSc (Hons.) in

List of authors

551

computing and management from Loughborough University and an ITIL Manager's Certificate. He is also a fully qualified Unix (Solaris) engineer. Rolf Ilg

Since 1991, Dr. Rolf Ilg has led the Knowledge Transfer Competence Center at the Fraunhofer Institute for Industrial Engineering (IAO) and at the University of Stuttgart's Institute for Industrial Engineering and Technology Management (IAT), with which the former collaborates. Dr. Ilg belongs to the management team of either institute. He studied mechanical engineering at the Technical University of Darmstadt and earned his doctorate at the University of Stuttgart. Clemens Jochum

Professor Clemens Jochum has been Executive Director at Deutsche Bank since 2001 and Group Chief Technology Officer since May 2003. He also lectures in business computing at the University of Frankfurt/Main, where he is Vice Chairman of the E-Finance Lab. After earning his doctorate and conducting four years’ research in the USA, Jochum was awarded an honorary professorship in Darmstadt in 1993. His non-academic career began when he set up his own software company in 1982. Two years later, he moved to the Beilstein Institute and became its President in 1985. From 1994 onward, he served as Managing Director of a global joint venture company and, in 1998, became Managing Director of Deutsche Bank’s financial marketing company. Juniors Group

The Siemens Communications Juniors Group is an international network of young professionals, that contributes to the business of Siemens Communications on a project basis while developing its members' skills, knowledge and experience through training and project work. The following members of the Siemens Communications Juniors Group co-authored the article " Young professionals look to the communication of tomorrow": Koen Delvaux, Heidi Ganser, Matt Thomas and Holger Viehöfer.

552

List of authors

Wolfgang König

Prof. Dr. Wolfgang Koenig is dean of the Faculty of Economics and Business Administration at Goethe University in Frankfurt where he also heads the Institute of Information Systems. Since 2003 he chairs the E-Finance Lab Frankfurt am Main. Prof. König serves as Editor-in-Chief of the leading German IS journal “Wirtschaftsinformatik”. His research interest include standardization, networks, and e-finance. Friedrich Kopitsch

Dr. Friedrich Kopitsch has been Vice President and Manager of System Integration at Siemens Business Services since 1999. Prior to this appointment, he held numerous management positions in software and IT at a series of companies (including Siemens Nixdorf and BHF-Bank) over many years. Dr. Kopitsch's professional career began at AEG Telefunken in 1966, after he had completed his studies of physics and mathematics. Bernd Kuhlin

Bernd Kuhlin heads the Enterprise Systems division of Siemens Communications and was one of the founding fathers of the HiPath architecture. After joining Siemens in 1984, he held a series of positions, some of them in the USA. He served as Vice President Home Automation at Siemens Bosch Hausgeräte from 1995 through 1998, followed as of 1999 by a spell as Vice President of the LAN Network division. Kuhlin took charge of the Enterprise Networks division of Siemens Information and Communication Networks in 2001. Jörg Luther

Jörg Luther, a system and applications programmer by trade, has worked as a journalist for network and Internet publications for over a decade. Since July 2000, he has covered themes such as Linux/Unix, LAN, Internet and security in his capacity as editor of the IT magazine “tecCHANNEL”.

List of authors

553

Andy W. Mattes

Andy W. Mattes is a member of the Group Executive Management Board at Siemens Communications and President and CEO of Siemens Communications, Inc., USA. Mattes studied business administration in Munich before joining Siemens, where he initially held various sales and management positions. He led Enterprise Networks from 1999, until he was appointed to the Group Executive Management Board in 2001. He is now responsible for global Enterprise business. Mattes also lectures at the Technical University of Munich and, in 2003, published the book "21st Century Communications". Thomas Mendel

Dr. Thomas Mendel is a graduate of business studies at Mannheim University and holds a Ph.D. in computer networks from the University of Wales in Swansea. Mendel's research focuses strongly on infrastructure management. He came to Forrester through its acquisition of the Giga Information Group, which he had joined in 2001. Michael Meyer

Dr. Michael Meyer is head of Strategic Alliances Management and Vertical Solutions Practices at the Enterprise division of Siemens Communications. After completing his studies in Hamburg, he worked for an international management consultant before moving to Siemens in 1989. In a variety of publications, Dr. Meyer repeatedly showed how modern corporate processes can be optimized by the close integration of real time communication and business applications. Harald Müller

Dr. Harald Müller studied electrical engineering. He came to Siemens in 1996 and served in a variety of technical system and product planning and development positions in the field of information and communication systems. Müller today works in Corporate Development.

554

List of authors

Andreas Neef

Futurologist Andreas Neef studied computer sciences, philosophy and business administration. Since 1992, he has been advising companies on social change issues and long-term innovation strategies. At the turn of the millennium, he also became Executive Manager of Z_punkt GmbH The Foresight Company (Essen, Karlsruhe and Berlin). Rita Nøstdal

Rita Nøstdal has been a scientific officer at the Fraunhofer Institute for Industrial Engineering (IAO) since 1993. In this capacity, she heads up consulting and research projects in the fields of corporate management, performance management, controlling and the management of collaborative ventures. Nøstdal also lectures in management information systems at the University of Stuttgart. Christian Oecking

Christian Oecking, a member of the Executive Board of Siemens Business Services GmbH & Co. OHG and the global head of the company's IT outsourcing and operator business, speaks and writes frequently on the subject of strategic outsourcing. The strategic aspects of outsourcing projects and how these projects impact the value of the partner companies involved form the focus of his work. Oecking earned a graduate degree in mechanical engineering from the University of Dortmund. Andreas Pfeifer

Dr. Andreas Pfeifer earned degrees in material sciences and business studies before his career in management consulting began in 1988. Today, he is a Managing Director and Partner at Accenture, a global management consultancy. Dr. Pfeifer spearheads the activities of Accenture's Electronics & High Technology practice in German-speaking Europe, where he is responsible for senior assignments in the area of strategy and business architecture consulting.

List of authors

555

Arnold Picot

Professor Arnold Picot teaches business administration and heads the Institute for Information, Organization and Management at Ludwig Maximilians University's Munich School of Business. He also chairs the "Münchner Kreis", an international communication research body, and is a member of numerous scientific advisory councils. Professor Picot is the (co)-editor of several journals, series and anthologies. Frank T. Piller

Dr. Frank Piller is member of the faculty of the TUM Business School at Technische Universität München (TUM). His research areas are technology & innovation management and marketing of innovation with particular interests in customer integration, mass customization, personalization, relationship marketing, and open innovation. He is the author of numerous other papers and has written / edited six books. He graduated summa cum laude with a Ph.D. from the School of Business Administration at the University of Wuerzburg, Germany in 1999, where he received his master's degree in 1994, too. Ralf Reichwald

Professor Ralf Reichwald holds the Chair of Information, Organization and Management at the TUM Business School, Technische Universität München. His research focuses on organizational behavior, human resources development, leadership systems, the application of new information and communication technologies in the corporate sector, and innovation in the service sector. Professor Reichwald serves as an expert adviser on several academic committees and chairs a number of technical committees at federal and regional levels, as well as on behalf of industry associations. Thomas Renner

Thomas Renner has been with the Fraunhofer Institute for Industrial Engineering (IAO) since 1990, where he heads the Electronic Business Competence Center. This center develops innovative solutions for intercompany collaboration and networking and for electronic business transactions. Ren-

556

List of authors

ner studied electrical engineering at the University of Stuttgart and at Oregon State University, USA. Jörg Sauerbrey

Professor Jörg Sauerbrey studied electrical engineering and computer sciences. He joined Siemens in 1993, holding a series of positions in the field of information and communication systems. He is currently in charge of product management for IT security solutions at Siemens Communications. Martin Schallbruch

Martin Schallbruch studied computer sciences, law and sociology. He managed an IT service center at the Humboldt University in Berlin from 1992 through 1998, before taking a three-year post as office manager on behalf of then undersecretary of state Brigitte Zypries at the Federal Ministry of the Interior. Appointed as IT Director for this ministry in 2002, he is now responsible for IT security, IT strategy, e-government, passports, ID cards and resident registration. Dieter Spath

In 2002, Professor Dieter Spath took over as head of both the Fraunhofer Institute for Industrial Engineering (IAO) and the Institute for Industrial Engineering and Technology Management (IAT) at the University of Stuttgart. He had earlier held a professorship at Karlsruhe University's Institute of Machine Tools and Production Science and served as Managing Director of the KASTO Group. Ralf Steinmetz

Professor Ralf Steinmetz worked for over a decade in industrial research and development in the area of networked multimedia systems and applications. He has, since 1996, been head of the Multimedia Communications Lab at Technical University of Darmstadt. His thematic focus is on such topics in multimedia communications which contribute to his vision of real “seamless multimedia communications”. With over 200 refereed publications to his name, he was awarded the rank of Fellow of both the IEEE in 1999 and the ACM in 2002.

List of authors

557

Christof M. Stotko

Dr. Christof M. Stotko is the global marketing manager of EOS, a worldwide leading company producing Laser-Sintering systems for e-Manufacturing. Prior to this he was a research fellow at the Chair of Information, Organization and Management at the Technische Universität München (TUM). In this function mass customization was the key theme of his research work. Before moving to TUM, Stotko worked as a consultant to the automotive industry at Roland Berger Strategy Consultants and at IBM Deutschland GmbH. Jeff Swartz

Jeff Swartz is the President and Chief Executive Officer of Current Analysis, Inc. and a member of its Board of Directors. He joined Current Analysis as President and Chief Operating Officer shortly after its incorporation in January 1997 and was named CEO and President in January 2000. He brings with him more than 25 years of entrepreneurial and management experience in the market research, competitive intelligence and consulting fields, including time spent in senior positions at Arthur D. Little Decision Resources, BIS Strategic Decisions, and Giga Information Group. Heinz Thielmann

Professor Heinz Thielmann studied communication, data and control systems engineering and earned his doctorate at the Technical University of Darmstadt. At Philips, he led communication systems R&D from 1974 through 1984, before spending the next ten years as General Manager of the company's global Network Systems Business Unit. Since 1994, Professor Thielmann has headed up the Fraunhofer Institute for Secure Telecooperation with international partners in research and industry. IT security and real time secure enterprises are the focus of his work. Susan J. Unger

Susan J. Unger is Senior Vice President and Chief Information Officer at DaimlerChysler. She bears global responsibility for directing systems and computer hardware strategy and planning, systems applications development, data center operations, and telecommunications network operations.

558

List of authors

Unger has served in executive management since joining Chrysler Corporation in 1972. She is active in an advisory capacity to academic, government and industry organizations and was named “CIO of the Year” by Salomon Smith Barney in 2002. Bernd Voigt

Professor Bernd Voigt studied mathematics and computer sciences. He has held various positions at the Lufthansa Group, which he joined in 1992. He was appointed Senior Vice President and CIO in 1999, assuming responsibility for the Group's Information Management and New Media activities. Effective October 1, 2002, Professor Voigt took charge of the Infrastructure Services business segment of the Lufthansa Systems Group. Anette Weisbecker

Dr. Anette Weisbecker, a private lecturer, studied computer sciences at the Technical University of Darmstadt. After gaining several years' experience in industry, she joined the Fraunhofer Institute for Industrial Engineering (IAO) in 1988. She is now an assistant director and also heads its Software Management Competence Center. Tim Weitzel

Dr. Tim Weitzel is Assistant Professor at the Institute of Information Systems at Frankfurt University where he is in charge of the sourcing projects of the E-Finance Lab. Previously, he headed a variety of research and consulting projects on E-Business and B2B integration, outsourcing, EFinance, standardization, XML/EDI and E-HR. Claus Weyrich

Professor Claus Weyrich studied physics and was employed by Siemens' research laboratories in 1969, where he held senior positions in semiconductor development and material sciences. He took charge of Technology Research and Development in October 1994 and has, since 1996, headed Siemens' Corporate Technology department. Professor Weyrich was appointed to the Managing Board of Siemens on October 1, 1996.

List of authors

559

Alexander Zeier

Dr. Alexander Zeier studied business administration at the University of Würzburg and earned an engineering degree from Chemnitz Technical University. He later earned his doctorate in business computing at the University of Erlangen-Nuremberg. Dr. Zeier currently heads up strategic projects at SAP and lectures at a Bavarian institute of higher education.

Printing: Strauss GmbH, Mörlenbach Binding: Schäffer, Grünstadt