Improving Methods to Measure Attentiveness through Driver Monitoring

NHTSA · 2022 · ROSA P / United States. Department of Transportation. Office of the Assistant Secretary for Research and Technology

archive: archived pipeline: cataloged verified

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

Summary

This research addresses the critical safety challenge of driver inattention, which contributed to an estimated 3,142 fatalities in 2020. The study aims to improve the accuracy of Driver Monitoring Systems (DMS) by refining algorithms that measure a driver’s attention level in real time. Accurate differentiation between distracted and attentive driving is essential for DMS to effectively identify distraction and prompt drivers to refocus on the roadway. The project was conducted by researchers at the Virginia Tech Transportation Institute (VTTI) under the Safety Through Disruption (Safe-D) National University Transportation Center. The methodology leveraged a naturalistic dataset from a proprietary study originally conducted for General Motors. This dataset included DMS data, full-time video, and vehicle parameters such as brake, throttle, and steering wheel position. Researchers developed a ground-truth dataset by manually reviewing 1,367 ten-second driving events to determine the driver’s distraction level at the end of each clip. These levels were categorized, and glance locations were mapped to identify visual attention patterns associated with each distraction state. The core algorithm development focused on calculating attention levels by increasing or decreasing buffer values based on the driver’s gaze location. Once buffers reached specific thresholds, the system classified the driver as either “attentive” or “inattentive.” The study evaluated the algorithm’s performance by comparing its output against the manually determined ground truth, measuring accuracy via mean-squared error (MSE). Results indicated that the highest MSE occurred at low speeds for both “not distracted” and “very distracted” events. This finding suggests that glance patterns at low speeds differ significantly from those at high speeds, likely due to complex maneuvers such as turning, navigating intersections, or driving through pedestrian-heavy areas. To address these discrepancies, researchers developed algorithm variations that incorporated additional vehicle parameters, including speed, steering wheel input, brake pressure, and throttle pressure, to provide necessary environmental context. The findings conclude that effective assessment of driver attention requires, at a minimum, the integration of both glance location and vehicle speed. While current tools can determine inattention, the research highlights that no single tool can immediately and correctly identify every instance of distraction. Consequently, algorithms must be designed with a clear understanding of their limitations and the constraints of their data sources. This work underscores the importance of contextualizing gaze data with vehicle dynamics to improve the reliability of DMS in real-world driving scenarios.

Key finding

Algorithm error in classifying driver attention was highest at low speeds, where glance patterns during turns and intersections differed sharply from high-speed driving.

Methodology

naturalistic

Sample size: 1367

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. Discovered via bulk_ingest_rosap on 2026-05-23 (7 acquisition events logged).

StageOutcomeToolModelPromptAttemptsCompleted
discover success rosap 2 2026-05-23
archive success 1 2026-05-23
extract success cached 2 2026-06-10
clean success 1 2026-06-01
chunk success 1 2026-06-01
embed success 1 2026-06-02
enrich success 1 2026-05-23
promote success 1 2026-05-23
summarize success llm qwen3.6-27b-prismaquant summ-v5 3 2026-06-10
tag success vector_similarity 19 2026-06-11
verify success 3 2026-06-10

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

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

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).