Intelligent Traffic Signal Optimization using Sound-Based Ambulance Detection and Green-Light Extension

Abstract

Urban traffic congestion significantly hinders ambulance mobility, particularly at intersections where static signals restrict timely passage. This study presents a sound-based intelligent traffic signal optimization framework that integrates siren detection with adaptive green-light extension. The system utilizes IoT audio sensors with Mel-Frequency Cepstral Coefficient (MFCC) feature extraction and a Support Vector Machine (SVM) classifier, achieving 98% detection accuracy in distinguishing ambulance sirens from urban noise. Direction identification reached 96.3% accuracy, while distance estimation maintained an average error of 5.8 m. A novel dynamic green-light timer, incorporating speed, distance, and error-correction metrics, reduced unnecessary signal holding by over 50%, lowering cross-traffic disruption from 5% to 2%. Field tests demonstrated sub-second latency and effective convoy handling for multiple ambulances. These results establish the framework as a lightweight, scalable, and cost-effective solution for real-time ambulance prioritization in resource-constrained urban environments.