Improving supply-chain collaboration by linking intelligent agents to CPFR

The standpoint of this research lies in the study of the Collaborative Planning Forecasting and Replenishment (CPFR) process for trading partners (belonging to the same supply chain) who are willing to collaborate in exchanging sales and order forecasts. The hurdles that arose in implementing CPFR in field applications indicate the need for providing collaboration process with an intelligent tool to optimize negotiation. For this purpose, a literature review and classification have been carried out for autonomous agents used to managing supply-chain processes. On the basis of the classification, two multi-agent models are proposed, according to different degrees of agents' capabilities. To evaluate the strengths coming from an intelligent system embedded within the CPFR process, several experiments in different conditions have been conducted using a simulation tool. The analysis of results indicates that the agents-driven negotiation process (by comparison with CPFR without intelligent agents) benefits in terms of costs, inventory level, stock-out level and sales.