AI-Driven Real-Time Traffic Management Using YOLOv12 for Dynamic Signal Control and Intelligent Surveillance

Urban traffic congestion continues to pose serious challenges for cities worldwide, contributing to longer travel times, increased fuel consumption, higher greenhouse gas emissions, and greater risks of accidents. One major limitation of conventional traffic light systems is their reliance on fixed schedules, which do not adapt to realtime traffic conditions and often worsen congestion during peak hours. This study presents a realtime, AI-driven traffic light control system designed to respond dynamically to changing traffic flows. Using data collected from a network of cameras–capturing vehicle density and intersection load–the system applies an adaptive algorithm to adjust signal timings based on current conditions. Simulation experiments conducted across a range of urban traffic scenarios show promising results, including a 51.48% reduction in average vehicle waiting time and a 25.96% improvement in intersection throughput. These outcomes demonstrate the potential of intelligent traffic control systems to improve road efficiency and support more sustainable urban transportation.