This project addresses the critical issue of optimizing traffic flow in urban areas, a challenge that has significant implications for congestion, environmental impact, and urban livability. At the core of our approach is the deployment of a deep reinforcement learning agent tasked with controlling the flow rate of vehicles into a specified urban area. Utilizing the Simulation of Urban Mobility (SUMO) platform as a dynamic environment, we implemented the Proximal Policy Optimization (PPO) algorithm to train our agent. The agent’s objective was to minimize average travel time by dynamically adjusting traffic flow rates, with a reward system designed around the number of vehicles present in the area at any given time. The traffic in the simulated area is nonhomogeneous, reflecting the complex and varied nature of real-world urban traffic patterns. Our results show the use of RL perimeter control manages to minimize travel time and reduce traffic in a ‘rush hour’ like simulated scenario, demonstrating the potential for real-world applications to enhance urban mobility and reduce congestion effectively .