Skip to main content
UMD College of Behavorial & Social Sciences UMD College of Behavorial & Social Sciences
MENU
  • Home
  • Research
  • People
  • Grant Support
  • Resources
  • About
Search

Main navigation

  • Home
  • Research
  • People
  • Grant Support
  • Resources
  • About

Search our site:

The built environment and traffic collisions in the United States

Breadcrumb

  • Home
  • Featured Content
  • The Built Environment and Traffic Collisions In The United States
Quynh Nguyen

In a research article published in the British Medical Journal of Injury Prevention, Faculty Associates Quynh Nguyen, Thu Nguyen, and colleagues, investigated the relationship between built environment characteristics and traffic collisions across the United States using Google Street View data. Their research explores how characteristics like sidewalks, streetlights, street greenness, and road types influence collision rates. Using Google Street View images, data from the 2019-2021 Fatality Analysis Reporting System, and data from Washington D.C.'s District Department of Transportation, the study employed adjusted Poisson regression models to assess associations between built environment characteristics and traffic collisions while controlling for sociodemographic factors at the census tract level. 

The team found that "the presence of sidewalks, streetlights, street greenness and single-lane roads was associated with marked reductions in collisions-related outcomes." For instance, compared to the lowest tertile, areas with the highest tertile of sidewalks had 70% fewer fatal collisions. They also found that similar reductions were observed for street greenness, streetlights, and single-lane roads, at 26%, 30%, and 50%, respectively. 

The findings from this study suggest that sidewalks, streetlights, street greenness, and single-lane roads could mitigate motor vehicle collisions, particularly those involving cyclists and pedestrians. This study demonstrates how analyzing street features with data algorithms can identify factors that reduce road collisions. As stated by the authors, the results from this study can guide interventions that aim to improve the safety of the roads, thereby "[creating] environments that foster health and improve the safety of roadways."


 
Nguyen QC, Alirezaei M, Yue X, et al. (2024). Leveraging computer vision for predicting collision risks: a cross-sectional analysis of 2019–2021 fatal collisions in the USA. Injury Prevention. Published Online First: 06 June 2024. doi: 10.1136/ip-2023-045153.


 

Published on Wed, 04/23/2025 - 12:41

College of Behavioral & Social Sciences

Maryland Population Research Center

2105 Morrill Hall • College Park, MD 20742
Phone: 301-405-6403

For site issues, contact the webmaster.

Contact MPRC

Links
  • UMD Land Acknowledgement
  • UMD Staff Directory
  • UMD Web Accessibility
  • Alumni
© 2026 College of Behavioral & Social Sciences. All Rights Reserved.
Login