Psychophysiological responses to takeover requests in conditionally automated driving

Du, Na; Yang, X. Jessie; Zhou, Feng · 2020 · Crossref

DOI: 10.1016/j.aap.2020.105804

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 drivers’ psychophysiological responses to takeover requests (TORs) in SAE Level 3 conditionally automated driving, addressing a gap in existing literature that primarily focuses on behavioral metrics. The research aims to determine how non-driving-related tasks (NDRTs), traffic density, and TOR lead time influence internal driver states such as cognitive workload, attention, and emotion. By utilizing continuous, non-invasive physiological measures, the study seeks to provide a more comprehensive understanding of the takeover transition process than behavioral data alone can offer. The experiment employed a within-subjects design with 102 university students participating in a high-fidelity fixed-base driving simulator. Each participant experienced eight takeover scenarios involving varying conditions: cognitive load (low vs. high, manipulated via 1-back and 2-back memory tasks), traffic density (light vs. heavy), and TOR lead time (4 seconds vs. 7 seconds). Psychophysiological data, including gaze behaviors, heart rate (HR) and heart rate variability (HRV), galvanic skin responses (GSRs), and facial expressions, were recorded during both the automated driving stage and the takeover transition stage. Data analysis utilized linear mixed models and chi-squared tests to examine the effects of independent variables on physiological markers. Results indicated distinct physiological patterns associated with specific conditions. During the automated driving stage, high cognitive load resulted in lower HRV, narrower horizontal gaze dispersion, and shorter eyes-on-road time compared to low cognitive load. During the takeover transition, a 4-second lead time led to significantly inhibited blink numbers and higher maximum and mean GSR phasic activation compared to a 7-second lead time, suggesting increased stress and arousal. Additionally, heavy traffic density induced more frequent heart rate acceleration patterns than light traffic, indicating a defensive or rejection response to the driving environment. Correlation analysis revealed that higher GSR activation and fewer blinks were associated with more negative emotional valence, while greater emotional engagement correlated with larger heart rate differences between the NDRT and takeover stages. The findings demonstrate that psychophysiological measures can effectively and continuously indicate drivers’ internal states, including workload, attention, and emotional responses, during automated driving transitions. These results support the integration of such measures into future driver monitoring systems and adaptive alert systems, offering a real-time, non-intrusive method to assess driver readiness and state before observable behavioral changes occur. This approach complements traditional performance metrics and enhances the safety and design of human-automation interactions in conditionally automated vehicles.

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.

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.

• takeover transitions
• workload measurement
• hands on hands off engagement
• temporal
• automation
• stress driving

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
• Methodological Resource: measurement protocol