A Rural Road Accident Probability Model Based on Single-Vehicle Hazard Properties including Hazard Color and Mobility: A Driving Simulator Study

Sheikholeslami, Sina; Boroujerdian, Amin M.; Asadamraji, Morteza · 2020 · Crossref

DOI: 10.1155/2020/8826374

archive: archived pipeline: cataloged

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the lack of precise understanding regarding how specific properties of non-vehicular hazards—such as color, size, and mobility—influence the probability of single-vehicle accidents on rural roads. While road design standards account for perception-reaction time (PRT), they often overlook how hazard characteristics interact with ambient light and driving speed to affect driver visibility and reaction. The research aims to quantify these effects to improve road safety models and design guidelines. The researchers employed a fixed-based driving simulator to conduct a controlled experiment with 90 licensed drivers. Participants navigated four distinct scenarios (two daytime, two nighttime) on a simulated two-way, three-lane rural road, encountering an average of 14 hazards per scenario. Hazards included pedestrians, animals (e.g., camels, cows), and stationary objects (e.g., rocks), varying in size, color (green, yellow, or other), and mobility. The study utilized a Linear Probability Model (LPM) to analyze the binary outcome of accident occurrence, incorporating interaction terms to capture the combined effects of hazard properties and environmental conditions. The model demonstrated a high goodness of fit, with an R-squared value of 0.583 and a correct prediction ratio of 70.41%. Key findings indicate that hazard properties significantly modulate accident likelihood. Specifically, encountering a dark, small, stationary hazard at night (e.g., a 0.5 m × 0.5 m stone) is approximately 23% more likely to result in an accident compared to a large, moving, light-colored hazard during the day (e.g., a 1.5 m × 2 m camel). Color plays a critical role, with green-colored hazards being 27% less likely to cause accidents at night compared to other colors. Mobility also reduces risk; mobile hazards are associated with a lower accident probability than fixed ones, particularly in daytime conditions. Additionally, the study found that every 10 km/h increase in speed raises the accident likelihood by 1.9%, while each subsequent hazard encounter reduces the crash probability by 0.84%, suggesting a learning effect. The significance of this research lies in its demonstration that hazard-specific attributes, rather than just road geometry, are critical determinants of single-vehicle accident risk. The findings support the use of LPM over traditional count models for analyzing binary accident outcomes in controlled environments. By identifying that green hazards and mobile objects reduce risk, the study offers actionable insights for road design, such as optimizing roadside vegetation colors and managing animal crossings. The results highlight the need to integrate hazard property data into safety assessments, particularly for rural roads where non-vehicular hazards are prevalent.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.8-27b-gittensor on 2026-08-23; verification: pending re-verification.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• roadway lighting effects
• hazard perception
• vehicle conspicuity
• motorcycle conspicuity

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: crash risk outcomes