Related Experiment Video
Updated: Jan 18, 2026
Trajectory Data Analyses for Pedestrian Space-time Activity Study
Published on: February 25, 2013
Nighttime pedestrian safety in different communities: Application of artificial intelligence techniques
Sheikh M Usman1, Asad J Khattak1, A Latif Patwary1
1Department of Civil & Environmental Engineering, The University of Tennessee, Knoxville, United States.
Introduction:
Pedestrian safety is a growing concern in the United States transportation sector, with around 7,500 pedestrian crash fatalities reported in the United States in recent years. Pedestrians are at an even higher risk of crashes at night due to limited visibility and alcohol impairment of the drivers or pedestrians. The U.S. Department of Transportation (USDOT) has integrated six unique indicators (Economy, Health, Language Proficiency, Resilience, Environmental, and Transportation Access) at the census tract level to promote safety and achieve its Vision Zero goal. This study integrates these six indicators with pedestrian crash data from 2016 to 2019 in North Carolina. The pedestrian crash data are extracted from police reports using the Pedestrian and Bicyclist Crash Analysis Tool, which provides high-quality, detailed crash-type descriptors, resulting in a unique and comprehensive pedestrian crash database.
Methods:
The study applies rigorous methods for analysis, including the inference-based ordered logit model, to quantify key correlates of nighttime pedestrian crash injury severity in communities. To improve forecasting of pedestrian crashes and the resulting injury severity for planning purposes, an Artificial Intelligence (AI) based heterogeneous ensemble method, "Stacking," is applied with an Ordered Logit model and machine-learning techniques, Gradient Boosting, Decision Tree, and Random Forest as the base learners.
Results:
The stacked model integrates the predictive advantages of the base learner models and yields better predictive accuracy (78.85%) than the best-performing base learner (73.56%). The model results reveal associations between the Economy and Transportation Access indicators, roads without lights, pedestrian crossing violations, alcohol impairment, and nighttime pedestrian crash injury severity.
Conclusions And Practical Applications:
The study findings and the application of AI methods can assist safety practitioners in implementing targeted interventions to improve roadway infrastructure and overall road user safety.
