Real-Time Dynamic Traffic Optimization Network Using IoT and Machine Learning Technologies

The increasing vehicular traffic in urban areas challenges traditional traffic management systems. This paper presents an advanced system utilizing Internet of Things (IoT), Reinforcement Learning (RL), computer vision, and blockchain technologies. It dynamically optimizes traffic lights using real-time data from ultrasonic and sound sensors for efficient vehicle detection and emergency vehicle prioritization. A YOLO-based computer vision model accurately counts vehicles at intersections, while RL adapts light cycles to reduce congestion and improve traffic flow efficiency. Blockchain ensures data integrity and tamper-proof records of traffic operations. Simulation results show significant traffic delay reductions, with future plans for real-world deployment in smart city environments. The system effectively addresses issues related to traffic congestion, emergency vehicle delays and inefficient traffic light coordination, leading to smoother traffic flow and reduced response time for emergency services and the detection accuracy has improved.