Author ORCID Identifier

0000-0001-6842-6890

Document Type

Dissertation

Date of Award

5-31-2026

Degree Name

Doctor of Philosophy in Transportation - (Ph.D.)

Department

Civil and Environmental Engineering

First Advisor

Janice Rhoda Daniel

Second Advisor

Athanassios K. Bladikas

Third Advisor

Branislav Dimitrijevic

Fourth Advisor

Joyoung Lee

Fifth Advisor

Lazar Spasovic

Abstract

Micromobility devices—namely e-scooters and e-bikes—have rapidly gained popularity in the United Sates, rising from 35 million annual shared rides in 2017 to over 133 million in 2023 (NACTO 2024). But also increasing is the number of injuries associated with these devices; however, most research emphasizes injury and demographic patterns rather than crash characteristics that would inform prevention strategies. Most existing crash research also relies on small sample sizes, lacks nuance distinguishing between involved parties (motorists, pedestrians, single device), and fails to distinguish between bicycle and micromobility crash patterns despite micromobility devices often being instructed to use conventional bicycle facilities. These constraints make it difficult to determine meaningful crash predictors and countermeasures tailored to meet e-scooter and e-bike safety needs.

This dissertation aims to: 1) identify safety-related crash characteristics for e-scooters and e-bikes; 2) identify key differences between bicycle and e-bike or e-scooter crashes; and 3) explore countermeasures to address the identified safety concerns. The objectives are met through a unique "big data" analysis of New York City crashes occurring from 2020 to 2024 (n=34,952). Six binary logit regression models determine the likelihood of crash severity and crash vehicle. The four "crash severity" models identify significant variables that determine the likelihood of a fatal or severe injury (FSI) crash (rather than non-FSI) when the collision involves either: 1) an e-bike and motor vehicle, 2) an e-scooter and motor vehicle, 3) a single e-scooter or e-bike or 4) an e-bike or e-scooter and a pedestrian. The two "crash vehicle" models determine the likelihood that a crash will involve either 1) an e-scooter or 2) e-bike rather than a bicycle. Odds ratios (OR) describe the likelihood of an explanatory variable to predict the crash outcome.

Crash severity models indicate that significant factors vary by crash scenario. Single-device crashes had the highest concentration of FSI crashes (23% FSI) and were significantly more likely to result in FSI in the afternoon (OR=2.0) and along truck routes (OR 2.0). Pedestrian-involved FSI crashes were also relatively more common (12% FSI) and were more likely when the e-bike or e-scooter operator disregarded a traffic control device (OR 3.0) or was speeding (OR 2.6) and when the pedestrian crossed against the signal (OR 2.2). E-scooter versus motor vehicle crashes (6% FSI) had significantly higher odds of FSI if the e-scooter operator disregarded the traffic control device (OR 2.1) and when the crash occurred on a truck route (OR 1.7). E-bike versus motor vehicle crashes (6% FSI) were more likely to result in FSI overnight (OR 1.9), when the e-bike operator disregarded a traffic control device (OR 1.9) and when the motor vehicle driver improperly passed the e-bike (OR 1.3), which contradicts a common perception thought that only e-bikes are at fault and motorists are not. When predicting crash vehicle likelihood, e-scooter and e-bike crashes were more likely than bicycle crashes to involve unsafe speeds, improper motor vehicle turns, e-bike or e-scooter operator distraction and occur overnight.

As states and cities grapple with policies that address micromobility safety, this research provides a starting point for how to improve design and policy decisions pertaining to public rights of way and facilities that e-bikes and e-scooters share with pedestrians and motorists. Furthermore, it provides a unique paradigm for understanding safety factors that vary by user and vehicles involved.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.