Real-World Evaluation of the Impact of Automated Driving System Technology on Driver Gaze Behavior, Reaction Time and Trust
DOI: 10.1109/ivworkshops54471.2021.9669230
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 gap in research regarding driver behavior in real-world scenarios involving Automated Driving Systems (ADS), specifically focusing on the transition from automated to manual control. While previous studies relied heavily on simulated environments, this work investigates how non-driving-related tasks (NDRT) and vehicle speed impact driver gaze behavior, reaction time to take-over requests (TOR), and self-reported trust in automation. The motivation stems from the introduction of SAE Level 3 conditional automation, which allows drivers to engage in secondary tasks but requires them to safely regain control when the system exceeds its operational design domain. The researchers conducted field tests with 14 participants using a Toyota RAV4 equipped with an Openpilot semi-ADS system. Experiments took place on controlled tracks in Austria at two speeds: 30 km/h and 50 km/h. Participants performed four conditions: a baseline (no secondary task), a visual task (Stroop Color and Word Test), a manual task (retrieving screws from a bag), and a combined visual-manual task (writing text backwards on a smartphone). Data were collected using eye-tracking glasses, vehicle sensors, and post-task questionnaires. Key metrics included reaction time to TOR, frequency of gaze switches between the task and the road, average time eyes were on the road or task, and self-reported trust levels. Statistical analyses included One-Way ANOVA, Bonferroni post-hoc comparisons, Wilcoxon signed-rank tests, and Student T-tests. The results demonstrated that both the type of NDRT and vehicle speed significantly influenced driver behavior. Reaction times were fastest in the baseline condition (approx. 0.8 seconds) and increased significantly during secondary tasks, with the visual-manual task yielding the longest reaction times (approx. 1.6 seconds at 30 km/h). Gaze behavior varied markedly; drivers in the baseline condition looked at the road for an average of 15.6 seconds, whereas those in visual tasks looked at the road for less than 1.2 seconds. Higher speeds (50 km/h) led to more frequent gaze switches from tasks to the road in visual and visual-manual conditions compared to 30 km/h. Additionally, reaction times decreased slightly as speed increased for baseline, visual, and visual-manual tasks. Self-reported trust in the automation was significantly lower when performing secondary tasks compared to the baseline, particularly at higher speeds, though speed alone did not significantly alter trust levels in the baseline condition. The study concludes that real-world driver responses to TOR are heavily dependent on cognitive and manual load, with combined visual-manual tasks posing the greatest challenge for timely control takeover. The findings highlight that while drivers may glance at the road more frequently at higher speeds, their ability to react quickly is compromised by secondary tasks. This underscores the necessity for robust driver monitoring systems and careful design of TOR cues to ensure safety in conditional automation scenarios, as trust and situational awareness degrade significantly when drivers are distracted.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-08-09 |
| archive | success | unpaywall | — | — | 2 | 2026-08-09 |
| extract | success | cached | — | — | 3 | 2026-08-10 |
| clean | success | clean | — | — | 1 | 2026-08-09 |
| chunk | success | chunk | — | — | 1 | 2026-08-09 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-08-09 |
| enrich | success | semantic_scholar | — | — | 1 | 2026-08-09 |
| promote | success | — | — | — | 1 | 2026-08-09 |
| summarize | success | llm | qwen3.6-27b-nvidia | summ-v5 | 2 | 2026-08-10 |
| tag | success | vector_similarity | — | — | 10 | 2026-08-11 |
| verify | success | — | — | — | 2 | 2026-08-10 |
Summary generated by qwen3.6-27b-nvidia on 2026-08-10; 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).
- Empirical Findings: behavioral performance data
- Methodological Resource: measurement protocol, tool software