Estimating an e-scooter origin-destination model leveraging Yelp POI data

Dr. Abolfazl Karimpour, Assistant Professor of Transportation Engineering at SUNY Polytechnic Institute and Director of the Transportation AI Research Lab (TRAIL), is lead author of a new study that advances understanding of e-scooter mobility in urban environments.

Published in Environment and Planning B: Urban Analytics and City Science, the article, “Estimating an e-scooter origin-destination model leveraging Yelp POI data for enhanced urban mobility insights,” presents the first calibrated gravity-inspired machine learning model for e-scooter trips. The work was completed in collaboration with researchers from the University of Louisville.

Using Louisville, Kentucky, as a case study, the research integrated e-scooter trip data, traffic analysis zone (TAZ) boundaries, and crowdsourced Points of Interest (POI) data from Yelp. The model revealed that e-scooter flows are strongly influenced by the presence of restaurants, bars, coffee shops, and shopping centers, while parks and museums have less effect. Distance between zones was shown to reduce trip volumes, reflecting the short-range nature of micromobility travel.

The study demonstrates how POI data can be combined with trip records to forecast demand, assess infrastructure needs, and optimize fleet rebalancing. The approach provides a framework for cities to evaluate how urban changes, such as the addition of new commercial areas, may affect e-scooter usage.

This research contributes new tools for urban planners and transportation agencies seeking to integrate micromobility into broader transportation systems.