Toward a V2I-based Solution for Traffic Lights Optimization

The demographic trends and the consequent increase of urbanization lead us to answer to the growing need for mobility, by ensuring high levels of safety and travel comfort, as well as consumption reduction. In this regard, the optimal design of traffic light settings plays a fundamental role in the definition of solutions to achieve a safe and sustainable mobility in urban environments. The efficient operation of the traffic light network involves different approaches that can be classified in terms of the considered network configuration, modelling framework and strategy. In this paper, a vehicle-to-infrastructure (V2I) cooperation for traffic management is considered, in order to maximize the intersection capacity. Thus, the flow variability detected through V2I communications is taken into account by the responsive traffic light plans. The proposed strategy is compared with a video analytics-based solution performed by cameras monitoring the intersection, with the aim to evaluate the performance robustness of both camera and V2I-based approaches. Our numerical results reveal that outdoor conditions severely impact the quality of camera-information inferred from the captured images, while V2I-based solution is able to achieve better performances in terms of intersection capacity. A Channel occupancy evaluation is also conducted in order to further validate the proposed solution.