Using Driver State Detection in Automated Vehicles

Gaspar, John; Schwarz, Chris; Kashef, Omeed; Schmitt, Rose; Shull, Emily · 2018 · ROSA P / Safety Research Using Simulation (SAFER-SIM) University Transportation Center

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 study investigates the efficacy of driver state monitoring systems (DMS) in mitigating safety risks associated with conditional automation (SAE Levels 2–3), where drivers must remain ready to retake control. The research addresses two primary questions: whether continuous attentional alerts can maintain driver situation awareness (SA) and improve performance during automation failures, and whether state-contingent takeover requests—providing earlier warnings to distracted drivers—enhance takeover success. The experiment utilized the National Advanced Driving Simulator (NADS-1) at the University of Iowa, featuring a production camera-based DMS integrated into a 2014 Toyota Camry cab. Twenty-four participants engaged in a 40-minute automated drive while performing a secondary trivia task to induce disengagement. The study employed a between-subjects design with three groups: a baseline group with no driver state feedback; an Attentional Maintenance (AM) group that received alerts when looking away from the road for 30 seconds; and a State-Contingent Takeover (SCT) group that received earlier warnings during takeover events if classified as distracted. Participants encountered sudden automation dropouts and planned takeovers, with performance measured via freeze probes for SA, gaze metrics, and takeover response times. Results indicated that the AM system significantly improved driver engagement and safety outcomes. Drivers in the AM group demonstrated higher freeze probe accuracy (80% vs. 73%) and greater percent road center gaze (35% vs. 15%) compared to the baseline group. Crucially, AM drivers responded faster to sudden automation dropouts, exhibiting quicker hands-on-wheel and steering response times, which resulted in significantly smaller maximum lateral lane deviations. Conversely, the SCT approach yielded limited benefits. While SCT drivers tended to place their hands on the wheel earlier during dropouts, this did not translate to faster steering responses or significant improvements in planned takeover events compared to the baseline. The authors suggest a potential ceiling effect, where earlier warnings did not further accelerate steering actions once drivers were already re-engaging. The study concludes that continuous attentional monitoring is an effective strategy for maintaining driver situation awareness and improving reaction speeds during unexpected automation failures. However, modifying takeover requests based on driver state offers marginal utility, as earlier warnings do not necessarily accelerate the physical execution of control. These findings highlight the potential of DMS data to mitigate out-of-the-loop performance problems in partially automated vehicles, though further research is needed to optimize alert modalities and account for individual differences in technology experience.

Key finding

Drivers receiving continuous attentional maintenance alerts demonstrated higher situation awareness and faster, more stable responses during sudden automation failures compared to baseline drivers, whereas state-contingent takeover requests provided no significant performance benefit.

Methodology

simulator

Sample size: 24

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 (6 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 2 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).