Multi-Agent Process Coordination for Supply Chain Management

One of the problem faced by each actor of a supply chain is determining the right production and/or stock levels in order to minimize costs while assuring a given customer satisfaction level. In this setting single actors of the supply chain pursue their own different goal. Indeed, each actor takes its own decisions in autonomous way but it is influenced by the decisions made by the others; for example one of main aspect influencing the decision process is demand which is not only originated at the last downstream layer of the supply chain but also at the intermediate ones. This originates the need of a coordination between actors to achieve high efficiency in the whole system. If there is no coordination entity, the coordination action may be obtained through opportune interaction protocols. In this paper, we analyze the interactions between entities and how the communication engraves on the efficiency of the whole supply chain. We propose a multi-agent model to analyze the possibility to realize cooperation between the production entities of the system based on coordination; in this model every agent represents an autonomous entity of the supply chain. The proposed multi-agent model is tested on an italian extended factory in which the actors cooperate to produce compressors for refrigerators.