On demand transport system's approach as a multi-agent planning problem

In recent years, the “on demand transport-ODT” systems are increasingly popular in many cities around the world. An ODT system is defined by three components: the infrastructure of the city, the vehicles and the clients requests. Clients formulate requests for transportation between a pickup and a drop off places. These requests must be served in time by satisfying a set of constraints. We propose a new approach to resolve ODT problem as a multi-agent distributed planning problem. Our model reduces the exponential complexity of the problem, when solved by a centralized approach, to be polynomial. Our mutli-agent model is a strong first step to have a realistic ODT system taking into account environment and transport infrastructure conditions.