A comprehensive safety analysis for gaze fixation of drivers to outside scene

Sarkar, Abhijit · 2022 · Crossref

DOI: 10.54941/ahfe1002481

archive: archived pipeline: cataloged verified

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

Summary

This study investigates the impact of visual distraction caused by outside objects on driver safety, addressing a gap in research regarding how external stimuli contribute to crashes and near-crashes. While internal distractions like cell phone use are well-documented, external distractions have historically accounted for a significant portion of crashes. The research aims to quantify gaze fixation patterns, durations, and object types during safety-critical events compared to baseline driving, specifically focusing on visual attention directed through the right windshield. The analysis utilized data from the Strategic Highway Research Program 2 (SHRP2) Naturalistic Driving Study (NDS), the largest publicly available naturalistic driving dataset. The researchers selected 666 crash/near-crash (CNC) events and 446 baseline (BL) events where driver gaze was fixated through the right windshield for durations between 2 and 5 seconds. This range was chosen because distractions exceeding 2 seconds are considered safety-critical, while fixations longer than 5 seconds often indicate stationary vehicles. After manual annotation of object type, ego vehicle motion, and driver head movement, the final dataset comprised 617 valid CNC events and 410 valid BL events. Statistical analysis compared gaze fixation characteristics across these two groups. The results indicate distinct differences in gaze behavior between safety-critical and baseline driving. In CNC events, drivers primarily fixated on dynamic roadway objects, with light vehicles accounting for 39% of fixations, followed by trees/vegetation (16%) and traffic signs (15%). In contrast, BL events showed higher fixation on stationary objects, such as buildings (17%) and traffic signs (18%), with light vehicles comprising only 30%. Drivers in CNC events exhibited longer gaze fixation durations, with a median of 2.60 seconds compared to 2.40 seconds in BL events. Specifically, fixations on billboards and intersections were significantly longer during CNC events. Furthermore, 73% of CNC events involved visible head movement, primarily directed at other vehicles during maneuvers like lane changes or turns, whereas BL head movements were more often directed at buildings and intersections. Notably, drivers in BL events looked at stationary objects 89% of the time, compared to only 44% in CNC events. The study concludes that fixation on dynamic roadway objects contributes more significantly to safety-critical events than fixation on roadside stationary objects. This finding aligns with driver modeling theories suggesting that processing dynamic objects requires continuous cognitive updates and evidence accumulation, increasing cognitive load and distraction risk. The results imply that practitioners and safety researchers should prioritize understanding how dynamic external stimuli, such as moving vehicles and complex intersections, disrupt driver attention and increase crash risk.

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 qwen3.6-27b-prismaquant on 2026-06-09; verification: verified.

Topics

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

• visual
• external distraction
• eye movements scanning
• gaze based attention detection
• looked but failed to see
• attention allocation

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: behavioral performance data, observational prevalence
• Methodological Resource: tool software