Dynamic Signal Timing Optimization for Left-Turn Maneuvers via Vehicle-Infrastructure Communication

In densely populated urban areas, left-turn maneuvers at signalized intersections are a critical bottleneck, affecting both vehicular and pedestrian flows and contributing to unsafe conditions. The integration of Full Self-Driving (FSD) vehicles adds another layer of complexity, requiring enhanced communication and coordination strategies. This study proposes an innovative approach that employs Vehicle-to-Infrastructure (V2I) communication to dynamically optimize signal timings for intersections featuring both protected and unprotected left turns. The effectiveness of this approach was validated using a benchmark traffic network and real-world traffic data. By optimizing signal phases through a tested reinforcement learning framework, an average of 15% improvement was observed in terms of the average speed. This research not only offers a robust solution for the seamless incorporation of FSD vehicles into existing urban traffic systems but also addresses the longstanding challenges associated with left-turn maneuvers. The study's findings have profound implications for the future of smart, efficient, and safe urban mobility.