A community of agents as a tool to optimize industrial districts logistics

The aim of this paper is to find an optimal solution to operational planning of freight transportation in an industrial district. We propose a system architecture that drives agents - the industrial district firms - to cooperate in logistic field, to minimize transport and environmental costs. The idea is to achieve logistics optimization setting up a community made of district enterprises, preserving a satisfactory level of system efficiency and fairness. We address the situation in which a virtual coordinator helps the agents to reach an agreement. The objectives are: maximizing customers satisfaction, and minimizing the number of trucks needed. A fuzzy clustering (FCM), two Fuzzy Inference System (FIS) combined with a Genetic Algorithm (GA), and a greedy algorithm are thus proposed to achieve these objectives, and eventually analgorithm to solve the Travelling Salesman Problem is also used. The proposed framework can be used to provide real time solutions to logistics management problems, and negative environmental impacts.