ISTL researchers attend Vulnerable Road User conference to discuss mode choice analysis using e-bikes
By CSEE staff
Published June 15, 2026
Faculty and students from the Stephen Still Institute for Sustainable Transportation and Logistics (SS ISTL) traveled downstate to Tarrytown, New York, June 2 and 3, 2026, to attend the biannual Walk, Bike and Roll New York State (NYS) Symposium and present recent research findings.
The theme of this year’s event was “Protecting Our Most Vulnerable Road Users.” The symposium attracts more than 150 local, state, federal and private agency professionals who are devoted to building strong communities where people of all ages and abilities can safely and comfortably walk, bike, and roll.
The goal of the symposium is to empower NYS communities with the knowledge, tools and partnerships needed to design safe, accessible, and active transportation options for everyone. Experts from academia, industry and government led sessions that promoted discussion and showcased innovative solutions. Symposium outcomes include actionable ideas intended to expand walkable and bikeable communities and promote universal accessibility.
Austin Angulo, assistant professor, and Ye Wang, PhD candidate, both from the Department of Civil, Structural and Environmental Engineering, attended the 2026 symposium. Ther abstract was titled “A Temporal Graph Neural Network for Dynamic Demand Forecasting and Mode Choice Analysis: A Case Study of Citi Bike in NYC.”
In the research, Angulo and Wang explored the rapid expansion of shared micromobility systems, particularly the integration of electric bicycles, or e-bikes, alongside traditional fleets. Demand imbalance often forces users to abandon trips or travel to unsafe, distant alternatives, directly affecting the safety and sustainability of the transportation network.
The study addresses this challenge by applying advanced artificial intelligence (AI) techniques to predict aggregate station-level demand at the macro level and model individual user mode choice between e-bikes and classic bikes at the micro level.
Findings indicate that individual choice is driven primarily by immediate physical context, including trip distance and membership status rather than network dynamics and station history. The findings also show that continuous-time modeling is essential for capturing the irregular and often spatial context-dependent nature of e-bike demand at the station level.
The Walk, Bike and Roll symposium is organized by the Institute for Traffic Safety Management and Research, through the partnership of the following organizations: Federal Highway Administration, National Highway Traffic Safety Administration, NYS Governor’s Traffic Safety Committee, NYS Department of Transportation, NYS Department of Health, NYS Department of State, New York Bicycling Coalition, SADD Inc., NYS Association of Metropolitan Planning Organizations and Cornell Local Roads Program.
