Enhanced Traffic Management System Using AI-Based Ambulance Detection

— Abstract—As cities grow, traffic congestion has become a major issue, especially when it comes to getting emergency vehicles like ambulances through busy streets. In this paper, we propose a new system that uses AI to help solve this problem. The system detects ambulances in real time through video feeds, using the YOLOv8 model, and then changes the traffic light to green in the lane the ambulance is traveling in. Instead of using OpenCV to just count the number of cars, the system takes a look at how many vehicles are in each lane and changes the light timings to keep traffic moving smoothly. When we tested it, the system performed well, with an impressive recall of 0.96 and an F1-score of 0.84, which shows it’s both accurate and quick to react. By looking at things like confusion matrices and precision-recall graphs, we confirmed that the system works well. In the end, this approach can help ambulances get through traffic faster and also make the overall road system run smoother.