Fatigue and mental underload further pronounced in L3 conditionally automated driving: Results from an EEG experiment on a test track

Figalová, Nikol; Bieg, Hans-Joachim; Schulz, Michael; Pichen, Jürgen; Baumann, Martin; Chuang, Lewis L; Pollatos, Olga · 2023 · Crossref

DOI: 10.1145/3581754.3584133

archive: archived pipeline: cataloged

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates how supervising SAE Level 3 (L3) conditionally automated vehicles affects drivers’ mental workload and sleepiness compared to Level 2 (L2) and manual driving. The research addresses the concern that the shift from active driving to passive monitoring in L3 systems may lead to mental underload and increased fatigue, posing safety risks given that L3 drivers remain responsible for taking control when requested. To address this, the authors conducted an experiment on a real test track using an automated vehicle prototype, moving beyond the limitations of driving simulators which often employ unrealistic critical scenarios. The experimental design involved 23 adult participants with no prior automated vehicle experience. Each participant completed three counterbalanced 17-minute conditions: monitoring an L3 vehicle, supervising an L2 vehicle, and manual driving. The vehicle traveled in 2000-meter loops at 50 km/h on straight sections and 20 km/h in curves. Data collection included 32-channel EEG recordings (sampled at 1000 Hz) to measure oscillatory brain activity, specifically frontal theta (4-8 Hz) and alpha (8-12 Hz) power spectral density. Self-reported mental workload was assessed using the NASA-TLX scale, and sleepiness was measured using the Karolinska Sleepiness Scale (KSS). EEG data were preprocessed using the BeMoBil pipeline and analyzed via repeated measures ANOVA with Greenhouse-Geisser corrections. Results indicated that L3 driving was associated with significantly lower mental workload and higher sleepiness compared to both L2 and manual driving. Post-hoc tests revealed that NASA-TLX scores were significantly lower in L3 rides than in L2 (mean difference = 10.957, p=.020) and manual rides (mean difference = 14.043, p=.002). Similarly, KSS scores were significantly higher in L3 rides compared to manual rides (mean difference = 0.826, p=.039). EEG analysis confirmed these subjective reports: frontal theta power, an indicator of mental workload, was significantly higher in manual rides than in L3 rides (mean difference = 0.013, p=.010). Conversely, frontal alpha power, an indicator of fatigue and drowsiness, was significantly higher in L3 rides than in manual rides (mean difference = 0.028, p=.016). No significant differences were found between L2 and manual driving for these metrics. These findings suggest that the underload effect is more pronounced in L3 driving than in L2, where drivers must maintain constant monitoring. The increased alpha activity and self-reported sleepiness highlight that passive supervision in L3 systems can lead to significant drowsiness. The study concludes that these factors must be considered in the design of future automated vehicle interfaces. The authors recommend developing dynamic user interfaces that manage driver attention, adapt to the driving context, and include driving monitoring systems capable of detecting inattention and engaging the driver in a timely manner before a control transition is required.

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.

StageOutcomeToolModelPromptAttemptsCompleted
discover success Crossref 1 2026-08-09
archive success openalex 5 2026-08-09
extract success cached 4 2026-08-23
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 qwen3.8-27b-gittensor summ-v5 3 2026-08-23
tag success vector_similarity 10 2026-08-11
verify success 2 2026-08-09

Summary generated by qwen3.8-27b-gittensor on 2026-08-23; verification: pending re-verification.

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).