IEEE Smart Cities Webinar - A Distributed Traffic Management System for End-to-End Dynamic Routing of Intelligent Vehicles

The proposed distributed control system generates and frequently updates a single integrated and coherent view of the network state, independent of market penetration rate and vehicle communication characteristics. The performance of the routing system is compared to a dynamic pre-trip shortest path routing in the downtown Toronto under recurrent and non-recurrent congestion conditions. Results show that the proposed routing system can reach near social optimal state, by having 56% higher throughput rate while reducing the average travel time by 40%. The proposed system manages the incidents better (15% lower travel time) and at the same time minimizes its effects on the network.