An Algorithm for Estimating Origins and Destinations of Shared E-Scooter Trips from Public Data-Feeds

Abstract

We present an algorithm that estimates shared e-scooter trip origins and destinations using public General Bikeshare Feed Specification (GBFS) data. The model addresses challenges like GPS lag and randomized vehicle IDs. The algorithm is field-tested and validated using data for the city of Fairfax, VA. When applied to Washington, D.C., a larger city, the algorithm demonstrates scalability for larger, complex urban settings. Our methodology equips urban planners and activity center managers with a robust tool for analyzing micromobility patterns, enhancing first- and last-mile connectivity, and optimizing transportation systems---without relying on proprietary data from operators.