Automatic detection of mind wandering in a simulated driving task with behavioral measures
DOI: 10.1371/journal.pone.0207092
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 challenge of automatically detecting mind wandering (MW) during driving, a common cognitive state associated with performance degradation and increased accident risk. While MW is typically measured via subjective self-reports, the authors sought to develop an objective detection system using behavioral driving measures. The research was motivated by the need to minimize the negative impacts of "mindless driving" and the lack of prior studies attempting to detect MW in driving contexts using machine learning on behavioral data. The experiment involved 40 licensed drivers performing a car-following task in a driving simulator for 25 minutes. Participants maintained a 20-meter distance from a lead vehicle traveling at 80 km/h and reported their MW state upon hearing tone probes. The dataset comprised 957 labeled instances (MW-present or MW-absent). The researchers extracted four behavioral variables: offset from lane center, steering wheel ratio, foot operation, and distance from the lead vehicle. They computed global features (session-wide statistics) and local features (time-series statistics from 5s, 10s, or 15s windows preceding probes). Supervised machine learning models were built using two approaches: driver-independent (generalizing across participants) and driver-dependent (individual-specific). Participants were categorized by MW frequency into high/low and medium groups. The results indicated significant challenges in building effective driver-independent models. For participants with high or low MW frequencies, the best model used global features with SVM classification, achieving a kappa of 0.384 and 70% accuracy. However, for the medium MW group, performance was poor (kappa 0.124, 56.2% accuracy), and no model offered a significant improvement over others. In contrast, driver-dependent modeling yielded more promising results for some individuals within the medium MW group, suggesting that personalized models can effectively capture intra-individual behavioral variations. Statistical analysis revealed that few features significantly distinguished MW states for the medium group, highlighting the difficulty of generalization. The study concludes that developing robust MW detection systems must account for both inter-individual and intra-individual differences. While driver-independent models struggled with generalizability, particularly for moderate mind wanderers, driver-dependent approaches showed potential for specific users. These findings suggest that future safety systems should prioritize personalized calibration rather than universal algorithms. This work provides a foundational step toward automated monitoring of driver cognitive states, with implications for enhancing road safety and advancing psychological research on attention and behavior.
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 | canonical_url | — | — | 1 | 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.
- mind wandering
- cognitive
- drowsiness detection algorithms
- distraction detection algorithms
- situational awareness
- gaze based attention detection
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: validation psychometrics
- Theoretical Contribution: computational model