Multi-agent infrastructure for distributed planning of demand-responsive passenger transportation service

Demand-responsive transport services are systems that assign users' specific transport requests to different vehicles enabled to fulfill the required service. In order to contain costs, maximize profits or get a higher service level, different planning algorithms have been described in literature. In our work we propose a multi-agent architecture, which adopts a mixed planning model. This model is an integration of centralized and negotiation-based decision-making approaches. The advantage of using this model respect to a centralized approach is that the estimation of the utility function is avoided (the client is involved in the final decision by means of a negotiation process). The model shows the advantages of a complete decentralized approach in giving results that are not so far from the optimum for the whole system.