Optimized Adaptive AI-Driven Traffic Management with Emergency Vehicle Prioritization using YOLObased Detection and Fuzzy Logic Control

The increasing pace of urban development, coupled with exponential growth in vehicle ownership has created unpredictable challenges in managing traffic flow at urban intersections particularly when emergency vehicles requires a rapid passage. Traditional traffic management systems which depend on a predetermined timing cycles that demonstrates the fundamental limitations in responding to dynamic traffic patterns that leads to a prolonged queue length, inefficient use of resources and increased pollution caused by vehicles waiting at signals with their engines running. This research introduces a smart and well-designed traffic management solution for four-way intersections that combines advanced computer vision capabilities through YOLOv8 which makes a human like decision reasoning implemented through fuzzy logic. This system processes a live camera feed to simultaneously monitor the vehicle density across the lanes, identifies the presence of emergency response vehicles and tracks the total waiting durations of each lane. These real time metrices are fed into a fuzzy inference engine that gives human judgment to dynamically determine signal priority, a guaranteed passage for ambulances and fire trucks while balancing congestion mitigation and preventing any lane from experiencing indefinite delays. This system has been trained with using a comprehensive dataset of 3,000 images of real time traffic scenarios and also by using the augmented variation images. Comparative analysis shows that our developed system over the fixed-time method systems in signals is of a measurable reduction in mean queue wait duration, enhanced intersection throughput and allowing faster passage for emergency vehicles. The system’s inherent adaptability, low computational requirements and scalability makes it highly suitable for integration into modern smart city traffic systems.