Predicting Driver Takeover Performance in Conditional Automation (Level 3) through Physiological Sensing
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 critical safety challenge of driver takeover performance (TOP) in Level 3 conditional automation, where vehicles manage driving tasks but require human intervention upon request. The research aims to determine if multimodal physiological sensing can predict a driver’s readiness and performance during takeover events. The motivation stems from the need to ensure safe transitions between automated and manual control, particularly when drivers are engaged in secondary tasks that may impair their ability to react promptly. To investigate this, the authors conducted a driving simulator experiment involving 20 participants. The simulation, developed using the Unity engine and a ProSimu T5 Pro simulator, incorporated three types of secondary tasks (observing, 1-back, and 2-back cognitive tasks), three takeover events (obstacle, police car, and fake alert), and two traffic densities (40 and 80 vehicles per mile). Physiological data, including Skin Conductance Level (SCL), Heart Rate (HR), Frontal Asymmetry Index (FAI), and Mental Workload (MWL), were collected alongside vehicle performance metrics such as reaction time, maximum acceleration, and time-to-collision (TTC). The findings reveal distinct physiological patterns during takeover periods. SCL, MWL, and HR generally increased following takeover alerts, indicating heightened sympathetic arousal and mental cost. In contrast, FAI, an indicator of engagement, decreased, suggesting drivers became distracted from driving by secondary tasks. Harder secondary tasks (2-back) resulted in higher average MWL and HR but lower engagement compared to easier tasks. Takeover events also influenced physiological responses; for instance, the "fake alert" scenario elicited a slower SCL response, and the obstacle event triggered higher HR than other scenarios. High traffic density significantly increased average MWL. Correlation analysis showed that while SCL and HR correlated with vehicle performance metrics like reaction time and acceleration, no single physiological feature dominated the prediction of TOP. Furthermore, significant individual differences in physiological baselines were observed, highlighting the necessity of data normalization. The study concludes that physiological sensing provides valuable insights into driver state during conditional automation, supporting the development of personalized, real-time prediction models for takeover readiness. The results emphasize that secondary task difficulty and traffic density significantly impact driver mental workload and engagement. The authors recommend incorporating all measured physiological features for prediction models and stress the importance of accounting for individual differences through standardization. This work lays the foundation for adaptive alert systems that can enhance safety and trust in Level 3 autonomous vehicles by accurately assessing driver readiness before critical takeover requests.
Key finding
Physiological responses such as heart rate, skin conductance, and mental workload exhibit distinct patterns during takeover periods that correlate with vehicle performance metrics, though significant individual differences necessitate personalized modeling approaches.
Methodology
simulator
Sample size: 20
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.
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: physiological data, behavioral performance data
- Theoretical Contribution: conceptual framework