An agent-based dynamic information network for supply chain management

Abstract

One of the main research issues in supply chain management is to improve the global efficiency of supply chains. However, the improvement efforts often fail because supply chains are complex, are subject to frequent changes, and collaboration and information sharing in the supply chains are often infeasible. This paper presents a practical collaboration framework for supply chain management wherein multi-agent systems form dynamic information networks and coordinate their production and order planning according to synchronized estimation of market demands. In the framework, agents employ an iterative relaxation contract net protocol to find the most desirable suppliers by using data envelopment analysis. Furthermore, the chain of buyers and suppliers, from the end markets to raw material suppliers, form dynamic information networks for synchronized planning. This paper presents an agent-based dynamic information network for supply chain management and discusses the associated pros and cons.