Identifying built environment factors influencing driver yielding behavior at unsignalized intersections: A naturalistic open-source dataset collected in Minnesota

TL;DR

This paper addresses pedestrian safety at unsignalized intersections by building a large naturalistic dataset from 18 Minnesota sites (over 3,000 driver–pedestrian interactions) and examining how built-environment factors influence driver yielding. Using a logistic regression framework with stepwise feature selection, the study identifies significant predictors of yielding, including vehicle speed, crossing width, and nearby land uses such as restaurants/bars, parks, and schools, as well as parking presence and opposite-direction yielding. The authors demonstrate that the model achieves high predictive performance ($P(Y=1)$ via logistic regression) with an ROC AUC around 0.88 and provide actionable insights for road-safety planning and automated-vehicle design. By publicly sharing the open-source TIM-based dataset and the analysis, the work supports policymakers, planners, and researchers in Minnesota and beyond to design pedestrian-friendly environments and to improve vehicle responses at unsignalized crossings.

Abstract

Many factors influence the yielding result of a driver-pedestrian interaction, including traffic volume, vehicle speed, roadway characteristics, etc. While individual aspects of these interactions have been explored, comprehensive, naturalistic studies, particularly those considering the built environment's influence on driver-yielding behavior, are lacking. To address this gap, our study introduces an extensive open-source dataset, compiled from video data at 18 unsignalized intersections across Minnesota. Documenting more than 3000 interactions, this dataset provides a detailed view of driver-pedestrian interactions and over 50 distinct contextual variables. The data, which covers individual driver-pedestrian interactions and contextual factors, is made publicly available at https://github.com/tianyi17/pedestrian_yielding_data_MN. Using logistic regression, we developed a classification model that predicts driver yielding based on the identified variables. Our analysis indicates that vehicle speed, the presence of parking lots, proximity to parks or schools, and the width of major road crossings significantly influence driver yielding at unsignalized intersections. Through our findings and by publishing one of the most comprehensive driver-pedestrian datasets in the United States, our study will support communities across Minnesota and the United States in their ongoing efforts to improve road safety for pedestrians and be helpful for automated vehicle design.

Figures (10)

• Figure 1: TIM sensor attached to a light pole during data collection at site 17.
• Figure 2: Location ID vs. yielding outcome showing different yielding rates among sites.
• Figure 3: Visualization of predictor variables: Event Features
• Figure 4: Visualization of predictor variables: Pedestrian and vehicle characteristics
• Figure 5: Visualization of predictor variables: Signs and markings