Agent-based Supply Network Event Management (eBook)

Roland Zimmermann (Autor)

eBook Download: PDF

2006 | 2006
X, 329 Seiten
Springer Basel (Verlag)
978-3-7643-7487-7 (ISBN)

Infos

'Supply Chain Event Management (SCEM)' is one of the major topics in application-oriented Supply Chain Management. However, many solutions lack conceptual precision and currently available client-server SCEM-systems are ill-suited for complex supply networks in today's business environment. Agent-based proactive information logistics promises to overcome existing deficits by providing event-related information to all participants in the distributed environment. Hence, follow-up costs of disruptive events are significantly reduced for all network participants and performance of a supply network is increased.

In this book a thorough analysis of the event management problem domain is the starting point to develop a generic agent-based approach to Supply Network Event Management. The main focus lies on practical issues of event management (e.g., semantic interoperability) and economic benefits to be achieved with agent technology in this state-of-the-art problem domain.

Inhalt

Contents 6
Preface 10
Chapter 1 Introduction 12
Chapter 2 Event Management in Supply Networks 16
2.1 Problem 16
2.1.1 Event-related Information Logistics 16
2.1.2 Supply Networks 20
2.1.3 Formal Specification of the Problem 25
2.2 Requirements of an Event Management Solution 28
2.2.1 General Requirements 30
2.2.2 Functional Requirements 31
2.2.3 Data Requirements 33
2.2.4 Implications 35
2.3 Potential Benefits 35
2.3.1 Benefits for Single Enterprises 35
2.3.2 Analysis of Supply Network Effects 38
2.3.3 Benefits for Supply Networks 41
2.3.4 Summary on Potential Benefits 43
2.4 Existing Approaches 45
2.4.1 Tracking Systems 45
2.4.2 SCEM Software 50
2.4.3 Conclusion on Existing Approaches 58
Chapter 3 Information Base for Event Management 60
3.1 Data Model 60
3.1.1 Representation of the Supply Network Domain 60
3.1.2 Aggregation and Refinement of Status Data 68
3.1.3 Disruptive Event Data for Decision Support 72
3.1.4 Extendable Data Structures 75
3.2 Semantic Interoperability 76
3.2.1 Requirements for Semantic Interoperability 76
3.2.2 Existing Approaches 79
3.2.3 Ontology for Supply Network Event Management 81
3.3 Data Sources 85
3.3.1 Data Bases 86
3.3.2 Internet Sources and Web Services 89
3.3.3 Radio Frequency Identification Technologies 93
Chapter 4 Event Management Functions 98
4.1 Information Gathering in Supply Networks 98
4.1.1 Trigger Events 99
4.1.2 Inter-organizational Information Gathering 100
4.2 Proactive and Flexible Monitoring 107
4.2.1 Critical Profiles 108
4.2.2 Discovery of Critical Profiles 111
4.2.3 Continuous Assessment of Critical Profiles 116
4.3 Analysis and Interpretation of Event Data 124
4.3.1 Basic Approach 124
4.3.2 Data Interpretation with Fuzzy Logic 126
4.3.3 Aggregated Order Status 127
4.3.4 Assessment of Disruptive Events 131
4.3.5 Adjustment of Milestone Plans 133
4.4 Distribution of Event Data 138
4.4.1 Alert Management Process 139
4.4.2 Alert Decision Management 140
4.4.3 Escalation Management 144
4.4.4 Selection of Recipient and Media Type 147
4.4.5 Selection of Content 150
4.5 Event Management Process 152
4.5.1 Event Management Functions 152
4.5.2 Distributed Event Management in Supply Networks 154
Chapter 5 Agent-based Concept 156
5.1 Software Agents and Supply Network Event Management 156
5.1.1 Introduction to Software Agents 156
5.1.2 Benefits of Agent Technology for Event Management 160
5.1.3 Related Work in Agent Technologies 162
5.2 Agent Oriented Software Engineering 165
5.2.1 Approaches 165
5.2.2 AUML for Supply Network Event Management 168
5.3 Agent Society for Supply Network Event Management 172
5.3.1 Roles and Agent Types 172
5.3.2 Agent Interactions 177
5.3.3 Institutional Agreements 184
5.4 Coordination Agent 186
5.4.1 Structure 186
5.4.2 Behaviors 187
5.4.3 Interactions 192
5.5 Surveillance Agent 193
5.5.1 Structure 193
5.5.2 Behaviors 195
5.5.3 Interactions 199
5.6 Discourse Agent 200
5.6.1 Structure 200
5.6.2 Behaviors 201
5.6.3 Interactions 204
5.7 Wrapper Agent 206
5.7.1 Structure 206
5.7.2 Behaviors 207
5.7.3 Interactions 209
Chapter 6 Prototype Implementations 212
6.1 Generic Prototype 212
6.1.1 Overview 213
6.1.2 Ontology Integration 217
6.1.3 Coordination Agent 222
6.1.4 Surveillance Agent 228
6.1.5 Discourse Agen 233
6.1.6 Wrapper Agent 235
6.2 Supply Network Testbed 237
6.2.1 Simulated Enterprise Data Base 237
6.2.2 Simulator 238
6.3 Industry Showcase 240
6.3.1 Overview 240
6.3.2 Coordination Agent 242
6.3.3 Surveillance Agent 247
6.3.4 Wrapper Agent 251
Chapter 7 Evaluation 254
7.1 Concept 254
7.1.1 Constraints to an Evaluation 254
7.1.2 Multi-dimensional Evaluation 255
7.2 Analytical Evaluation 259
7.2.1 Effects of SNEM Cycles 259
7.2.2 Costs of Event Management 261
7.2.3 Cost-Benefit-Model and Benchmarks 264
7.2.4 Supply Network Effects 269
7.2.5 Event Management with Profiles 270
7.2.6 Conclusions 277
7.3 Experimental Evaluation 278
7.3.1 Reaction Function 278
7.3.2 Experimental Results 281
7.3.3 Cost-Benefit Analysis 284
7.3.4 Conclusions 286
7.4 Showcase Evaluation 287
7.4.1 Prototype Assessment 287
7.4.2 Analysis of Follow-up Costs 289
7.4.3 Conclusions 293
7.5 Summary - Benefits and Constraints 294
Chapter 8 Conclusions and Outlook 298
8.1 Supply Network Event Management 298
8.2 Further Research Opportunities 300
8.2.1 Object Chips for Supply Network Event Management 301
8.2.2 Event Management in other Domains 303
8.2.3 Integration and Acceptance Issues 303
Appendices 306
References 320

Textauszug

Chapter 7 Evaluation (p. 243.244)

An evaluation of the SNEM concept and its related prototype implementations has to consider specifics of the supply network domain, especially its structural complexity. In the following, an approach to evaluation is selected which permits to consider benefits and constraints of agent-based SNEM from different analytical perspectives. The primary focus of all evaluation activities is on economic benefits to be achieved by an agent-based solution to the SNEM problem.

7.1 Concept

Several constraints exist which prevent realization of a large-scale field trial for agentbased SNEM and thus to provide a single evaluation activity. Hence, three different perspectives ranging from a theoretical model to laboratory experiments and an assessment of an industry showcase are used to provide an overall assessment of the economic potential for agent-based event management in complex supply networks.

7.1.1 Constraints to an Evaluation

7.1.1.1 Complexity of the Domain

Evaluation of the SNEM concept is restricted by the inherent complexity of the supply network domain. The multitude of partners who cooperate in a realistic supply network and the autonomy of each of these partners (see section 2.1.2.4) prohibit testing a prototypical system in a realistic setting. A typical situation in a network consists of at least several suppliers for each single manufacturer. Each manufacturer himself is again part of an even larger supply network not to speak of logistics service providers. This results in a multitude of relationships between supply network partners. Aside from prohibitive costs of implementing prototypes in such an environment for a large number of enterprises it is likely that most partners will not agree beforehand to participate in a large-scale trial with- out any initial indication which realistic benefits a SNEM solution can provide. This situation precludes field trials that would cover multiple levels of a supply network, because empirical data on the fulfillment processes of all levels is hardly available. Only data on a limited scope (e.g. the business case depicted in section 2.4.1.3) is available for an evaluation.

7.1.1.2 Planning of Reactions

A second constraint that adds additional complexity to the evaluation of the SNEM concept is the fact that economic benefits measured in monetary units are achieved only indirectly. The following argument uses the formal specification (see section 2.1.3) of the problem: The following argument uses the formal specification (see section 2.1.3) of the problem: A SNEM system satisfies the implicit demand Dq for information on disruptive events DE by a message , Ms. is the input for any reaction that is supposed to minimize the consequences CSQ of the disruptive events DE. These negative consequences and their reduction through event management can be measured in monetary terms, as illustrated by the calculation of potential benefits in section 2.3.

Produkt-Details

• Erscheint lt. Verlag: 16.3.2006
• Reihe/Serie: Whitestein Series in Software Agent Technologies and Autonomic Computing
• Reihe/Serie: Whitestein Series in Software Agent Technologies and Autonomic Computing
• Zusatzinfo: X, 329 p.
• Verlagsort: Basel
• Sprache: englisch
• Themenwelt: Mathematik / Informatik ► Informatik
• Themenwelt: Wirtschaft ► Betriebswirtschaft / Management ► Logistik / Produktion
• Schlagworte: Agent Technology • logistics • Network event management • Networks • Supply Chain • Supply Chain Management
• ISBN-10: 3-7643-7487-X / 376437487X
• ISBN-13: 978-3-7643-7487-7 / 9783764374877