Improving fluidity in cities with intelligent traffic lights
Author: Matthewson Quentin
Connected Autonomous Vehicles (CAVs) will get more and more present on the roads in the future. They are a great way to improve traffic fluidity and thus reduce pollution. As long as other non-autonomous vehicles are present, traffic lights will be mandatory to coordinate the flow of cars. Nowadays, in most countries, traffic lights are unable to adapt dynamically to traffic-related needs like events, traffic jams, road work, and more.
Communicating traffic lights could be the solution to improve fluidity and promote fast and dynamic adaptations. Companies are starting to fund research to prove the utility of such an infrastructure and aiming to improve the technologies involved.
An ongoing study that started last year with the High School of Landscape, Engineering, and Architecture (HEPIA) in collaboration with Geneva Public Transport (TPG) has determined that the overall benefits of such technologies would reduce the waiting time by 13% in cities using dynamic traffic lights. The simulated environment used to make this study is SUMO, alongside TraCI (Traffic Control Interface) and Veins. Since Geneva is home to pilots involving CAVs, it seemed interesting to see what type of smart city improvements could take place. To do so, problematic intersections have been identified in the city's heart, and a mapping of the area was created to match today's reality. Geneva does not yet have Vehicle-to-Everything (V2X) communications for the traffic lights implemented in its infrastructure. Still, these studies aim at promoting the benefits that could be gained for every road user.
Figure 1: Simulated road network, northern part of Geneva
Future works will involve Vehicle-to-Infrastructure (V2I) communications improving city flux with a supervision organ. All intersections will be analyzed using traffic density that can be provided by connected vehicles. A parallel study shall look into V2I improvements in a node network not relying on centralized supervision, with each node being an intersection. nIoVe will help provide cyber-security advice and proofs-of-concept in the real-world pilot environment that will be applied to these simulations.
V2X : https://arxiv.org/pdf/2003.07191.pdf