AI-Based Traffic Management Systems for Autonomous Vehicles

Abstract

Traffic management is one of the most significant aspects of modern urban planning and smart cities. In big cities, there are a large number of vehicles that cause numerous issues such as a high risk of accidents, traffic jams, air pollution, and so forth. Recently, traffic management systems have drawn much attention from both academic scholars and practitioners who are working in the domains of computer science and transportation. The problem of traffic management requires a real-time adaptive and efficient decision-making system. Most of the traffic management systems work focuses on recommending traffic light control methods with the intention of reducing congestion and waiting times. Given the availability of a large number of vehicles and some historical and current states of traffic, a traffic management system can suggest the best phases at every intersection and update the recommendations as needed. Therefore, it is important to develop novel AI-based traffic light control approaches to increase computational efficiency and robustness as well.