Intelligent Transportation & Mobility Systems

Our research in intelligent transportation and mobility systems focuses on developing data-driven methods for safer, more efficient, adaptive, and resilient transportation. We combine transportation data, sensing, machine learning, simulation, and decision-support methods to study public transit, multimodal mobility, transportation operations, and system performance under both routine and disruptive conditions.

Public Transit Sensing & Passenger Analytics

Multi-Source Passenger Load Estimation for Public Transit

We develop data-driven methods for reliable passenger load estimation in public transit by integrating automatic passenger counting (APC), wireless sensing, and contextual information. Our work combines machine learning, physical constraints, and adaptive multi-source data fusion to maintain reliable passenger-load trajectories under changing and imperfect sensing conditions.

This research aims to reduce error accumulation and improve robustness when sensing streams are sparse, biased, or inconsistent, supporting more reliable crowding information, transit operations, and data-driven planning.

APC · Multi-Source Sensing · Data Fusion · Machine Learning · Public Transit