Training to Improve Drivers’ Behavior When Partial Driving Automation Fails
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 addresses the safety challenges associated with Level 2 (L2) partial driving automation, specifically focusing on drivers’ ability to recognize system limitations and execute timely control transfers. As L2 systems operate within specific Operational Design Domains (ODD), drivers must transition from passive supervisors to active operators when automation fails. The research was motivated by evidence that standard user manuals are insufficient for conveying these critical limitations, leading to poor situational awareness and delayed takeover responses. The primary objective was to design and evaluate an interactive training program to improve drivers’ situational awareness and takeover performance when L2 systems reach their ODD limits. The researchers employed a between-subjects experimental design involving 36 participants randomly assigned to three training conditions: PC-based interactive training, user manual reading, or placebo training. The PC-based program utilized the "3M" method (Mistake, Mentoring, Mastery), an interactive approach where participants practiced identifying takeover scenarios, received real-time feedback on errors, and repeated tasks until mastery was achieved. This training covered eight scenario types, including sharp curves, intersections, and stationary objects. Following training, all participants drove through 10 post-test scenarios in a high-fidelity fixed-based driving simulator. Performance was measured using binary takeover success rates and Situational Awareness Rating Technique (SART) scores. Results demonstrated that the PC-based training significantly outperformed both the user manual and placebo conditions. Participants in the PC-based group successfully took back control in 91.71% of required scenarios, compared to only 27% for the user manual group and 23.71% for the placebo group. Statistical analysis confirmed a significant main effect of training type on takeover success. Furthermore, the PC-based group exhibited significantly higher mean overall SART scores (22.03) than the user manual group (15.20) and placebo group (10.84), indicating superior situational awareness. Notably, the interactive training did not induce false alarms; no participants in the PC-based group unnecessarily took control during non-hazardous scenarios. The findings imply that interactive, experiential training is far more effective than passive information delivery for preparing drivers for partial automation. The study highlights the inadequacy of owner’s manuals in conveying safety-critical system limitations and suggests that manufacturers should adopt interactive training methods, such as the 3M approach, to enhance driver education. These results support the development of comprehensive training programs for dealerships and driving schools to mitigate risks associated with automation failures. Future work should aim to validate these findings through on-road studies and larger sample sizes to improve generalizability.
Key finding
Drivers in the PC-based training group successfully took back control in 91.71 percent of takeover scenarios, significantly outperforming the user manual group (27 percent) and placebo group (23.71 percent).
Methodology
simulator
Sample size: 36
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).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| 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.
- automation surprise
- automation
- situational awareness
- takeover transitions
- automation complacency bias
- mode awareness
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).
- Applied Guidance: countermeasure evaluation
- Theoretical Contribution: conceptual framework, computational model