An Integrated AI-Driven Framework for Smart Urban Traffic Management: Towards Sustainable, Efficient, and Safe Cities

Rapid urbanization has intensified traffic congestion, emissions, and safety concerns, necessitating intelligent solutions for sustainable urban mobility. This paper proposes an integrated AI-driven framework for smart urban traffic management that combines deep learning, reinforcement learning, and graph-based optimization into a unified architecture. The system leverages real-time data from multiple sources including cameras, GPS devices, and IoT sensors—to enable predictive traffic forecasting, adaptive signal control, and network-wide coordination. Evaluated using real-world datasets from Tehran, Barcelona, and a synthetic city, the framework demonstrates significant improvements over conventional methods: average travel time reduced by 34%, fuel consumption and CO₂ emissions decreased by 24%, and over 15 incidents prevented daily. These results highlight the framework’s effectiveness in enhancing efficiency, sustainability, and safety in modern cities. The modular design supports scalability and extensibility, offering a practical pathway toward smarter, greener, and safer urban environments.