DrivingResearch.

Explicit behaviors affected by driver's trust in a driving automation system

Liu, Hailong; Hiraoka, Toshihiro; Tanaka, Seiya · 2019 · arXiv (Cornell University)

DOI: 10.48550/arxiv.1906.03831

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 explicit behavioral manifestations of driver trust in Driving Automation Systems (DAS), specifically Adaptive Cruise Control (ACC), to address the safety risks associated with "over-trust." Over-trust occurs when drivers rely on automation beyond its functional capabilities, potentially leading to accidents. The authors posit that outward physical behaviors, particularly foot placement and intervention timing, serve as observable indicators of a driver’s internal trust state. The research aims to identify these behaviors to eventually enable real-time trust monitoring and over-trust prevention. The experimental design utilized a driving simulator equipped with a driver monitoring system using OpenPose to track the motion of participants' right feet and legs. Thirteen licensed drivers participated, though data from three were excluded due to equipment issues. Participants evaluated their trust in real-time using steering wheel paddles, categorizing their state as completely trusting, moderately trusting, or completely distrusting. The experiment consisted of three scenarios: the first two induced trust through smooth ACC performance, while the third introduced a dangerous event where the ACC failed to respond to a stationary obstacle, requiring immediate driver intervention. Participants were divided into two groups based on their trust patterns: Group 1 (high trust) and Group 2 (cautious/low trust). The results confirmed two primary hypotheses. First, regarding foot placement, Group 1 participants were significantly more likely to place their feet away from the pedals when completely trusting the ACC, whereas they kept their feet on the pedals when distrusting. Group 2 exhibited the opposite behavior, maintaining physical readiness regardless of mental trust levels. Second, regarding intervention timing, higher trust levels correlated with delayed brake reaction times during the dangerous event. Statistical analysis revealed a significant difference in reaction times between collision and non-collision groups ($p=0.01$). Furthermore, a strong positive correlation ($r=0.75$) was found between the integral trust score and brake reaction time, indicating that higher trust directly delayed operational intervention. Consequently, participants with higher trust were more likely to collide because they intervened too late for the ACC to handle the hazard. The study concludes that explicit behaviors, specifically foot position and intervention delay, are reliable indicators of driver trust in DAS. The findings demonstrate that over-trust leads to dangerous delays in taking control during system failures. These results provide a foundation for developing machine learning models that predict driver trust states from motion data, which could be used to design systems that alert drivers or adjust automation levels to prevent over-trust-related accidents.

Key finding

Higher trust in adaptive cruise control leads drivers to move their feet away from pedals and significantly delays their operational intervention during dangerous situations.

Methodology

simulator

Sample size: 13

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 author_sweep_intake on 2026-05-28.

StageOutcomeToolModelPromptAttemptsCompleted
discover success author_sweep — — 2 2026-05-28
archive success canonical_url — — 1 2026-06-04
extract success cached — — 3 2026-06-10
clean success clean — — 1 2026-06-04
chunk success chunk — — 1 2026-06-04
embed success embed Qwen/Qwen3-Embedding-8B — 1 2026-06-04
enrich success — — — 1 2026-05-28
promote success — — — 1 2026-06-04
summarize success llm qwen3.6-27b-prismaquant summ-v5 2 2026-06-10
tag success vector_similarity — — 15 2026-06-11
verify partial — — — 2 2026-06-10

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

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).