he aim of this paper is to find an optimal solution for freight transportation in an industrial district taking into account the point of view both of agents and customers, and trying to reduce the environmental pollution. We propose a system architecture that drives agents to cooperate in minimizing costs, and environmental negative impacts. The idea is to achieve an optimization of logistics by setting up a community made of district enterprises. We address the situation in which a virtual coordinator helps the agents to reach an agreement among each other. The objectives are: maximizing customers satisfaction, and minimizing the number of trucks needed. A fuzzy clustering, a fuzzy/genetic system and a greedy algorithm are thus proposed to achieve these objectives, and eventually an algorithm to solve the Travelling Salesman Problem.