DESIGN AND ANALYSIS OF AGENTS FOR SUPPLY CHAIN MANAGEMENT: EXPERIENCES FROM THE TRADING AGENT COMPETITION

I. Dogan,∗ J.-W. Yoo,∗∗ R.B. Chinnam,∗ Y. Jia,∗ and G. Vanteddu∗

Keywords

Supply chain management, multi-agent modelling, Trading Agent Competition, dynamic supply chains, supply chain planning

Abstract

Rapid globalization in recent decades and the consequent demands make it essential for supply chains (SCs) and their managers to account for real-world dynamics in managing businesses to successfully compete and emerge as market leaders. The vast majority of SC models discussed in the literature are “static : greatly stylized, generally local and decision specific. While they present a number of important insights into the working of SCs, they are constricted by the model-related assumptions and replication- related problems. The dynamic nature of the business environment is subjecting SC players to a much more volatile market environment than ever before. Agent-based solutions are particularly relevant in this changing market context. Agent-based approach for supply chain management (SCM) was given a boost with the unveiling of the “TAC-SCM competition in 2003 by Carnegie Mellon University and the Swedish Institute of Computer Science through the Trading Agent Competition (TAC). Based on our competition experience for 2 years, we advocate in this paper the importance of incorporating real-world marketplace dynamics into SC modelling through agent- based solutions. We present innovative strategies adopted by some of the successful teams in the competition and detail important aspects of our own agent.

Important Links:

Go Back