Modeling and Predicting Geospatial Teen Crash Frequency

Road traffic crashes have emerged as a significant public health concern in this century. Nearly 3,700 people die in road crashes each day (WHO, 2016). As per the National Highway Traffic Safety Administration (NHTSA), in 2018, 2,476 teens aged 13-19 were killed in road crashes. As per North Carolina Department of Transportation (NCDOT)’s “2018 Crash Facts”, teen drivers were involved in 51,061 road crashes during the year, resulting in 80 teen deaths and 11,776 teen injuries in North Carolina. Road crashes are the leading cause of teen deaths in the United States (Centers for Disease Control and Prevention, 2019). Behavioral factors, lack of adequate driving experience as well as the driving environment contribute to teen crash frequency. For example, the driving challenges in an urban area and a rural area are entirely different for a new teen driver. Various types of land uses have different teen trip generation and attraction potential. Teen travel activity is higher at schools, commercial areas, recreational centers, etc. While such location-based features play a key role and can explain the spatial heterogeneity in teen crash frequency, studies exploring the relationship between teen crash frequency and driving environment are found to be very limited. The purpose of this research is to investigate the relationship between road network, demographic, and land use characteristics on teen crash frequency. The influence of location-specific indicators on teen crash frequency is difficult to capture from currently available safety performance functions (SPFs). Also, the global regression estimates (all data are used to develop a single model) may not result in accurate estimates at certain locations. Geospatial methods like geographically weighted regression (GWR) can help generate localized SPFs (develop a model specific to the location) by capturing the spatial variations in the explanatory variables and accurately estimate teen crash frequency. The research outcomes can be used to not only estimate teen crash frequency by accounting for spatial variations in the explanatory variables but also assist with planning engineering, enforcement, and education activities. The research outcomes may also lead to a higher emphasis on region-specific or localized crash prediction models.

    Language

    • English

    Project

    • Status: Active
    • Funding: $37500
    • Contract Numbers:

      69A3551747127

    • Sponsor Organizations:

      Office of the Assistant Secretary for Research and Technology

      University Transportation Centers Program
      Department of Transportation
      Washington, DC  United States  20590
    • Managing Organizations:

      Office of the Assistant Secretary for Research and Technology

      University Transportation Centers Program
      Department of Transportation
      Washington, DC  United States  20590
    • Performing Organizations:

      Mineta Consortium for Transportation Mobility

      San Jose State University
      San Jose, CA  United States  95112
    • Principal Investigators:

      Pulugurtha, Srinivas

    • Start Date: 20210216
    • Expected Completion Date: 20211231
    • Actual Completion Date: 0
    • USDOT Program: University Transportation Centers

    Subject/Index Terms

    Filing Info

    • Accession Number: 01765060
    • Record Type: Research project
    • Source Agency: Mineta Consortium for Transportation Mobility
    • Contract Numbers: 69A3551747127
    • Files: UTC, RiP
    • Created Date: Feb 18 2021 6:19PM