DEEP LEARNING-BASED ADAPTIVE TRAFFIC CONTROL FOR AMBULANCES

Abstract

This paper presents an AI-based intelligent traffic management system designed to prioritize emergency vehicles, particularly ambulances, in congested urban environments. The motivation behind this work stems from increasing traffic congestion and delayed emergency response times, which significantly impact survival rates in critical situations. The proposed system utilizes YOLOv8 for real-time ambulance detection from video streams and integrates it with an adaptive traffic signal control mechanism. Upon detection, the system dynamically switches traffic lights to create a clear path for the emergency vehicle. The model is implemented using Python and OpenCV and tested on real-time video datasets. Experimental results demonstrate high detection accuracy (92.4% precision) and rapid response time (0.7 seconds), proving the effectiveness of the system in reducing delays. The solution is scalable and can be extended with IoT and GPS-based integration for smart city deployment.