Dr. Madhur Behl is an Associate Professor in the Departments of Computer Science and Systems and Information Engineering at the University of Virginia, specializing in real-time machine learning and control for cyber-physical systems. His research focuses on the design and implementation of intelligent systems, with applications in autonomous vehicles and racing. He leads UVA’s Cavalier Autonomous Racing team and co-founded the F1/10 International Autonomous Racing Competitions to advance research and education in autonomy. Dr. Behl earned his Ph.D. in Electrical and Systems Engineering from the University of Pennsylvania. On last Thursday, he shared insights on “Automated Extraction of Temporal Logic Based Trajectory Embeddings for Enhanced Autonomous Vehicle Safety.”
Dr. Behl’s presentation focused on the Trajectory-to-Action Pipeline (TAP), a scalable method for deriving Scenario Description Language (SDL) labels from large trajectory datasets. Leveraging machine learning and natural language processing, TAP automates the extraction of interpretable scenario descriptions from raw sensor and video data. This approach enhances the benchmarking and safety analysis of autonomous vehicle (AV) systems by enabling the classification and comparison of AV behavior across similar real-world situations, such as sudden lane changes, pedestrian crossings, or unexpected vehicle stops. By improving the transparency of AV decision-making, TAP contributes to a standardized approach for evaluating AV safety, aligning with the mission of the Commonwealth Cyber Initiative Central Virginia Node to support cutting-edge, impactful research in AI and cybersecurity.
