AI-Based Adaptive Traffic Signal Control System for Mixed Traffic Conditions in Colombo

This research proposes an AI-based adaptive traffic signal control system designed for mixed traffic conditions in Colombo, Sri Lanka. The proposed system uses YOLOv8 for vehicle detection and Passenger Car Equivalent (PCE)-weighted traffic-demand estimation to support dynamic traffic signal timing decisions. The approach aims to reduce vehicle waiting time, queue length, and traffic congestion compared with conventional fixed-time traffic signal systems. The proposed framework considers the diverse vehicle types and traffic patterns commonly observed on Sri Lankan roads. PTV VISSIM is identified as the simulation environment for evaluating the proposed approach against fixed-time signal control. This work provides a foundation for developing intelligent, data-driven traffic management solutions suitable for urban transportation in Sri Lanka.