Non-Motorist Safety at Highway-Rail Grade Crossings: Developing a Crash Prediction Model with Integrated Non-Motorist Exposure
Non-motorist users at highway-rail grade crossings (HRGCs) include pedestrians, bicyclists, wheelchair users, skateboarders, and push scooter users, among others. Incidents involving non-motorist users at HRGCs are often underreported or overlooked, yet statistics reveal that they significantly contribute to overall fatalities and injuries at these locations. Pedestrians and bicyclists are particularly vulnerable at HRGCs due to the lack of adequate protective barriers or warning devices. In 2022, the Federal Railroad Administration (FRA) recorded 2,202 crashes at HRGCs, leading to 269 fatalities and 827 injuries nationwide. Furthermore, during the same year, there were 1,157 reported incidents of pedestrian rail trespassing, resulting in 606 fatalities and 551 injuries. These numbers emphasize the need for a comprehensive understanding of the risks associated with non-motorized users at HRGCs and identification of crossings where non-motorized users may be susceptible to crashes. In 2020, the Federal Railroad Administration (FRA) developed a crash frequency model for HRGCs; however, it does not consider non-motorist characteristics in its crash prediction. It is anticipated that adding components pertaining to pedestrians, bicyclists and other non-motorist users will improve the model's overall crash prediction. This improvement has the potential to guide more efficient resource allocation and HRGC’s safety decision-making. Furthermore, the existing FRA model relies on conventional statistical analyses. In contrast, the proposed research aims to explore the efficacy of robust artificial intelligence (AI) based models for crash prediction at HRGCs. The research team seeks to compare traditional statistical prediction models with AI-based models for crash predictions at HRGCs; the investigation aims to discern the superior performance of these models by evaluating their precision and fitness according to established criteria. These models will not only encompass physical characteristics of HRGCs grade crossings but would also integrate dynamic elements such as train and vehicular traffic, along with non-motorist exposure and its associated factors. The non-motorist exposure data will be acquired upon the successful completion of Phase I within the broader research framework. Anticipated outcomes of this study include improved HRGCs’ crash frequency prediction model. This research will contribute to a deeper understanding of safety hazards associated with HRGCs, considering various dynamic attributes related to traffic and trains, acknowledging the vulnerability of non-motorists at HRGCs. The proposed model would be developed through a comprehensive approach, incorporating policy perspectives on grade crossing safety, thorough data reviews, statistical analyses, AI-based techniques, and rigorous validation processes. Ultimately, the research findings are expected to empower transportation agencies to implement proactive safety measures, assisting in reducing the frequency of crashes and promoting the overall well-being of both motorized and non-motorized HRGCs users.
Language
- English
Project
- Status: Active
- Funding: $219336
-
Contract Numbers:
69A3552348340
-
Sponsor Organizations:
Office of the Assistant Secretary for Research and Technology
University Transportation Centers Program
Department of Transportation
Washington, DC United States 20590University of Nebraska, Lincoln
1400 R Street
Lincoln, NE United States 68588 -
Managing Organizations:
University of Nebraska, Lincoln
1400 R Street
Lincoln, NE United States 68588 -
Project Managers:
Stearns, Amy
-
Performing Organizations:
University of Nebraska, Lincoln
1400 R Street
Lincoln, NE United States 68588 -
Principal Investigators:
Khattak, Aemal
Aman, M
Farooq, M
- Start Date: 20240601
- Expected Completion Date: 20250831
- Actual Completion Date: 0
- USDOT Program: University Transportation Centers Program
Subject/Index Terms
- TRT Terms: Artificial intelligence; Crash data; Crash risk forecasting; Cyclists; Pedestrian safety; Predictive models; Railroad grade crossings; Railroad safety
- Subject Areas: Data and Information Technology; Pedestrians and Bicyclists; Planning and Forecasting; Railroads; Safety and Human Factors;
Filing Info
- Accession Number: 01924845
- Record Type: Research project
- Source Agency: University Transportation Center for Railway Safety
- Contract Numbers: 69A3552348340
- Files: UTC, RIP
- Created Date: Jul 22 2024 7:49AM