Traffic control system modelling to control traffic lights using smart camera

An intelligent system enhances the current traffic light control system by detecting the number of vehicles at intersections, whether congested or not. Presently, the system identifies the number of vehicles and adjusts the green light duration to prioritize congested sides of an intersection. However, it is limited to capturing the number of vehicles stopping before the red light, unable to account for vehicles beyond the red light or measure vehicle speeds. Consequently, fast-moving vehicles may benefit from excess green light time, potentially increasing the risk of accidents. To address this issue, the DKI Jakarta Traffic Control System Management Unit implements a green light time limit to ensure equitable road usage at intersections. This approach, though effective, still falls short in managing congestion at specific critical points requiring advanced solutions. Artificial intelligence (AI) offers a promising remedy, enabling the system to monitor vehicle speeds and predict traffic density before and after red lights. This study adopts a prototype model to simulate real-world conditions using toy cars and scaled road scenarios. The prototype integrates available tools and materials to create a traffic control system equipped with intelligen t traffic light management. Using the YOLOv8 algorithm, the prototype successfully detects vehicles waiting at red lights and measures the speed of approaching cars, paving the way for improved traffic regulation in real-world applications.