Intelligent fleet management of autonomous vehicles for city logistics

Consideration is given to the management of a fleet of autonomous vehicles (AVs) designed to support freight delivery in urban areas. The fleet of AVs is assumed operating within the general framework of an Intelligent Transportation System (ITS). Creating the technologies necessary for such ITS raises many important and complex problems related to research areas such as automated reasoning, perception, and control among others. Two of them, namely, AVs’ routing and AVs’ motion planning , are the main themes of this work. In this paper, we propose a novel approach for their formation and composition in the context of an ITS. Routing and motion planning decisions for AVs are assessed considering the latest traffic conditions existing on the road network. Since these conditions are usually vague and therefore difficult to be measured accurately, fuzzy logic concepts are incorporated into ITS to simulate the uncertainty in traffic networks. Solutions to the fuzzy routing and fuzzy motion planning for AVs are finally derived from a hybrid approach that combines the A-star algorithm with two modified genetic algorithms. Simulated experiments over the city of Patras (Greece) show the efficiency of the developed approach.