Assessment of Secondary Tasks Based on Drivers’ Eye-Movement Features

Yao, Ying; Zhao, Xiaohua; Feng, Xiaofan; Rong, Jian · 2020 · OpenAlex-citations

DOI: 10.1109/access.2020.3010797

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 addresses the critical issue of distracted driving by developing a visual distraction assessment method for in-vehicle secondary tasks. Motivated by the rising prevalence of mobile phone usage and the significant proportion of traffic accidents attributed to distraction, the research aims to evaluate the visual load imposed by various tasks. The authors focus on visual distraction, identified as having a greater impact on driving behavior than manual or cognitive distractions, and seek to refine existing evaluation criteria that rely on limited general visual characteristics. The study specifically investigates how driver age and experience interact with different secondary tasks to influence visual distraction levels. The researchers conducted driving simulator experiments with 34 participants, divided into two groups: 15 experienced older drivers (average age 47.6 years) and 19 less experienced younger drivers (average age 23.05 years). Participants performed five secondary tasks: navigation, tuning the radio, replying to a text message, replying to a voice message, and making a phone call. Visual data were collected using an eye tracker at 60 Hz. The study employed the AttenD algorithm, which models an "attention buffer" that depletes when drivers look away from the field relevant for driving and refills upon returning gaze. An influence degree index ($INC_{dis}$) was calculated based on the depletion of this buffer to quantify the visual distraction caused by each task. Statistical analyses, including ANOVA, were used to compare visual features across tasks and driver groups. The results demonstrated significant differences in visual features among the secondary tasks. Navigation imposed the lowest visual load, while replying to text messages and making phone calls resulted in the longest total eyes-off-road time and highest proportion of long glances. Tuning the radio and replying to text messages also exhibited the longest mean glance durations. Most tasks, except navigation, exceeded the National Highway Traffic Safety Administration’s guidelines for safe visual distraction, particularly regarding the proportion of glances longer than 2.0 seconds. Furthermore, experienced older drivers showed longer eyes-off-road times and mean glance durations compared to younger drivers, indicating a more pronounced impact of visual distraction on their driving behavior. However, older drivers also exhibited a higher frequency of glances, suggesting they monitored traffic conditions more frequently despite the distraction. The study concludes that the proposed AttenD-based assessment method effectively differentiates the visual distraction levels of various secondary tasks and accounts for driver demographics. The findings highlight that complex tasks like texting and calling pose significant safety risks by exceeding established visual distraction thresholds. The interaction between driver experience, age, and task type underscores the need for tailored safety standards. This assessment framework provides a foundation for developing visual-manual standards for in-vehicle information systems, aiding in the design of interfaces that minimize distraction and enhance driving safety.

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 2026-06-25; verification: verified.

Topics

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

• visual manual
• distraction detection algorithms
• visual
• temporal
• auditory
• manual

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: measurement protocol