Classification of Level 2 Driving Automation Events Observed on Public Roads – Part 2
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, published by the National Highway Traffic Safety Administration (NHTSA), documents the operational performance of SAE Level 2 driving automation systems on public roads. The study addresses the need to understand how partial driving automation systems function in real-world conditions, specifically focusing on system disengagements and driver interventions. The research utilized two passenger vehicles: a 2020 Hyundai Palisade SEL and a 2017 Tesla Model S 90D. Both vehicles were driven by professional drivers on three distinct test routes: a 108-mile highway route, a 32.4-mile rural route with single-lane roads and intersections, and a mixed route comprising highway, rural, and city roadways. Drivers maintained minimal contact with the steering wheel to keep the systems active, using remote triggers to record noteworthy events via synchronized cameras. The study classified observed events into three categories. Type I events involved the system suddenly terminating automation and transferring lateral control back to the driver. Type II events were subjectively noteworthy operations where the system remained active without issuing alerts. Type III events included driver-initiated manual overrides or unintended lane departures where the vehicle automatically corrected its position. The Hyundai Palisade’s system relied on forward-facing radar and a mono camera, while the Tesla Model S utilized eight cameras, ultrasonic sensors, and radar, along with "Navigate on Autopilot" features for automated lane changes and exits. Results indicated significant differences in event frequency between the two vehicles and across route types. For the Hyundai Palisade, Type I events averaged 9.1 per 100 miles on highways, 43.7 on rural routes, and 27.7 on mixed routes. Type III events were most frequent on rural routes (149.2 per 100 miles), with 946 of 1,085 such events being lane departures. In contrast, the Tesla Model S exhibited fewer Type I events (averaging 0.1 to 21.3 per 100 miles) but higher Type II event rates on highways (21.6 per 100 miles). The Tesla experienced 161 lane departure events out of 198 total Type III events. The data highlights that rural and mixed environments triggered more system disengagements and driver interventions than controlled-access highways, particularly for the Hyundai system. The findings provide empirical data on the limitations of current Level 2 automation systems in diverse driving environments. The high frequency of Type III events, especially lane departures on rural roads, underscores the necessity for continuous driver supervision and the potential challenges these systems face in complex, non-highway scenarios. This information supports NHTSA’s ongoing efforts to evaluate automated vehicle safety and informs the understanding of driver-system interaction in partial automation contexts.
Key finding
The Hyundai Palisade experienced significantly higher rates of Type I and Type III events compared to the Tesla Model S, particularly on rural and mixed routes, while the Tesla exhibited higher rates of Type II events on highways and rural roads.
Methodology
on_road
Sample size: 2
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
- Theoretical Contribution: conceptual framework