Supply Web Co-Ordination By An Agent-Based Trading Network With Integrated Logistics Services

Andreas Gerber*, Christian Russ and Matthias Klusch

German Research Center for Artificial Intelligence, Stuhlsatzenhausweg 3, 66123, Saarbrücken, Germany

Received 30 October 2002; revised 20 February 2003; accepted 28 February 2003. ; Available online 3 May 2003.

Abstract

B2B interactions, like electronic negotiations and auctions between suppliers and customers, could be significantly improved by enabling the participants to adapt their bidding strategies to current logistics information (e.g., about transportation condition, cost or dates) while the negotiation goes on. We present an approach of an agent-based information and trading network (ITN) called CASA for dynamic production and sales of timber; the integrated services for logistics and e-commerce are efficiently coordinated by appropriate types of holonic structured intelligent agents of the network. We introduce the agent-based architecture and describe how the agents build their plans and optimize them afterwards. For optimizing their plans, the agents use various market-based negotiation mechanisms, i.e., several auction mechanisms and the simulated trading mechanism described in detail in this article. The effects of the different mechanisms on resulting cost and surplus have been evaluated by various simulation runs in different competitive and co-operative settings. It turned out that the simulated trading mechanism as well as matrix auction mechanisms for two or three items are especially suitable for supply web co-ordination and optimization tasks.

Author Keywords: Supply web co-ordination; Agent-based trading network; Integrated logistics services

*Corresponding author.

www.ABA-RD.org
"Agents for Business Automation" Ó Research & Development