Evaluation of Eyes Off Road During L2 Activation on Uncontrolled Access Roadways
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 investigates driver eyes-off-road (EOR) behavior when using SAE Level 2 (L2) automated driving features on uncontrolled access roadways. The research addresses concerns that L2 systems, which handle lateral and longitudinal control but require driver supervision, may lead to increased distraction and crash risk. Specifically, the authors analyzed whether drivers exhibit higher EOR durations and frequencies when L2 systems are active compared to when they are inactive, and how intersection types influence this behavior. The analysis utilized naturalistic driving data from the Virginia Connected Corridor 50 Elite Vehicle Naturalistic Driving Study. The dataset included 771 driving epochs from 12 Tesla drivers equipped with adaptive cruise control and lane centering. Events were matched for criteria such as time of day and speed (20–55 mph) and categorized by L2 status (active vs. not active) and intersection type (no intersection, straight through, or turning). EOR behavior was operationalized using two definitions: EOR 1 (only forward glances considered on-road) and EOR 2 (all driving-related glances considered on-road). Metrics included total EOR time, mean EOR time, single longest glance, and number of off-road glances. Mixed-effect ANOVA models were employed to assess main effects and interactions. Results indicated that EOR behavior was significantly higher across all metrics when L2 systems were active. Drivers exhibited approximately two additional seconds of total EOR time and longer single longest glances when automation was engaged. Intersection type also influenced behavior; drivers looked away from the road less frequently and for shorter durations when navigating intersections compared to straight road segments. Significant interactions revealed that drivers adjusted their EOR behavior specifically when turning through intersections while L2 was active, but not when going straight. Ancillary analyses showed that EOR behavior increased at speeds below 37 mph and that hands-off-wheel behavior was three times more likely when L2 systems were active. Secondary task occurrence rates did not differ significantly by L2 status. The findings suggest that while L2 automation increases driver distraction, drivers remain somewhat attentive to complex maneuvers like turning. However, the increased EOR time and hands-off-wheel behavior indicate a potential safety risk, as prolonged off-road glances are associated with higher crash probabilities. The study highlights the need for further investigation into how drivers interact with automation in varying traffic contexts and underscores the importance of monitoring driver engagement to ensure the safety benefits of L2 systems are not negated by distraction.
Key finding
Drivers exhibit significantly increased eyes-off-road behavior, including longer total off-road time and more frequent glances, when Level 2 automated driving systems are active compared to when they are inactive.
Methodology
naturalistic
Sample size: 12
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: behavioral performance data, observational prevalence
- Methodological Resource: tool software