Urban Space Through Public Transportation

How people actually use public transportation in urban space reveals far more than simple ridership counts. Every boarding, alighting, and transfer encodes a story about where residents live and work, which neighborhoods are well-connected and which are isolated, and how transit demand fluctuates across the hours and seasons of urban life. By mining large-scale smartcard records, GPS traces, and sensor data from buses, trains, and trams, our research reconstructs the spatial and temporal structure of transit usage at the city scale.

Our work develops methods to identify dominant travel corridors, characterize the diversity of trip purposes embedded in transit flows, and detect how external events — from extreme weather to major city developments — reshape usage patterns across the network. These insights allow us to move beyond static origin–destination matrices toward dynamic, context-aware models of urban transit demand that capture how the city breathes and changes over time.

Understanding usage patterns is ultimately a foundation for better decision-making. We translate spatial usage analytics into actionable guidance for transit agencies: optimizing route design and service frequency in response to real demand, revealing underserved communities whose travel needs are not met by existing networks, and anticipating how urban growth and land-use changes will shift future ridership. The goal is a public transportation system that is deeply legible to the city it serves — responsive, equitable, and built around how people truly move.