Public data, private insights: field-validated approaches to estimating e-scooter travel patterns

We propose a methodology for estimating shared e-scooter trip origins and destinations using publicly available General Bikeshare Feed Specification (GBFS) data. This approach tackles challenges including GPS update lags, randomized vehicle IDs, and jurisdictional boundary constraints. By leveraging GBFS attributes, we developed a scalable framework to analyze mobility patterns. Field experiments in Fairfax, VA, validated the algorithm’s accuracy, with cross-verification using provider data confirming reliable trip activity capture. Scalability was demonstrated by applying the methodology to GBFS data in Washington, D.C., showing adaptability to larger, denser urban environments. The results reveal dynamic spatiotemporal patterns, such as concentrated activity in downtown cores during peak hours and heightened recreational demand at locations like the National Mall during non-peak periods. This methodology provides urban planners and activity center managers with actionable insights for optimizing fleet deployment, enhancing first- and last-mile connectivity, and informing infrastructure planning. Requiring no proprietary datasets, it relies solely on public GBFS data, ensuring broad applicability across cities with compatible feeds. The findings advance shared micromobility research by offering a practical, replicable tool for analyzing usage patterns, empowering stakeholders to optimize micromobility systems and integrate them into sustainable, multimodal urban transportation networks.