Naturalistic Study of Level 2 Driving Automation Functions
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 report details the Naturalistic Study of Level 2 Driving Automation Functions (L2 NDS), a project sponsored by the National Highway Traffic Safety Administration (NHTSA) and conducted by the Virginia Tech Transportation Institute. The study aimed to investigate real-world driver interactions with commercially available vehicles equipped with lateral and longitudinal driving automation features. Specifically, it sought to evaluate system performance, driver-system interaction, driver performance during alerts, and driver engagement, including a sub-study focused on drives exceeding two hours. The methodology involved a naturalistic driving study where 120 participants were recruited from the Washington, DC, region and balanced by age (25–39 and 40–54) and gender. Participants were assigned one of ten instrumented vehicles (including models from Audi, Infiniti, Mercedes-Benz, Tesla, and Volvo) for a four-week period. Each vehicle was equipped with a NextGen Data Acquisition System that continuously recorded video, audio, and parametric vehicle data. Drivers received training mimicking dealership experiences before driving the study vehicles in place of their personal cars. Over a 14-month period, participants drove 216,585 miles, with 70,384 miles driven with both lateral and longitudinal automation active. Data were sampled and annotated to analyze system use, non-driving tasks, and responses to Request to Intervene (RTI) alerts. Key findings indicated that drivers generally operated the automation systems in accordance with manufacturers’ intended use, activating features primarily at speeds above 40 mph and avoiding use in heavy traffic, non-interstate roads, or rainy conditions. Contrary to concerns about distraction, the prevalence of non-driving tasks did not increase when both automation features were active; common tasks included passenger interaction and monitoring the instrument panel. Regarding safety, 71 Safety-Critical Events (SCEs), comprising five minor crashes and 66 near-crashes, were observed. No statistical relationship was found between SCE rates and feature activation levels, and no RTIs were associated with any SCEs. Drivers responded to RTIs with an average reaction time of 0.94 seconds when hands were off the wheel. Some delayed responses were attributed to drivers intentionally testing system boundaries rather than distraction. Subjective feedback revealed that trust in longitudinal systems increased over time as drivers learned their limitations, while trust in lateral systems remained stable despite occasional performance issues. The study concludes that Level 2 automation features can be used safely in naturalistic settings, with driver behavior consistent with active supervision. The lack of increased non-driving task prevalence and the absence of a link between automation use and safety-critical events suggest that these systems do not inherently compromise safety when used as intended. However, the findings highlight the importance of managing driver expectations regarding system capabilities, particularly for lateral control. The results provide critical empirical data for refining human factors requirements and supporting the safe integration of driving automation technologies into commercial vehicles.
Key finding
No statistical relationship was observed between safety-critical event rates and feature activation level, and non-driving task prevalence did not increase when both lateral and longitudinal control features were active.
Methodology
naturalistic
Sample size: 120
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).
- Methodological Resource: dataset resource, tool software
- Theoretical Contribution: conceptual framework