AIoT Based Green Road System for Smart Cities

Abstract

In the last few years, Algerian cities have been facing increasing traffic congestion due to rapid urbanization and population growth. This congestion translates into severe delays, accidents, and environmental pollution, as present traffic management systems are not capable of coping with the increase in the number of vehicles. In our work, we proposed a smart traffic control system with the help of AI and IoT technology to regulate and dynamically optimize the flow. In this study, with the cooperation of the municipalities, smart cameras were placed at important intersections to take a real-time traffic flow image. We, therefore, employ the YOLOv8 object detection model for vehicle detection and classification. Bytetracker tracking helps to count and trace the vehicles, and that way, the movements can be calculated accurately. The processing is done on Nvidia Jetson Nano devices, which allows real-time optimization of the traffic light and gives priority to emergency vehicles. A solution to achieve reduced congestion, improved road safety, and a lower carbon footprint is. Upon preliminary deployments, promising results have shown that this can be effectively scaled and adapted for broader traffic management improvements across Algerian cities.