Real-World Use of Automated Driving Systems and their Safety Consequences: A Naturalistic Driving Data Analysis

Kim, Hyungil; Song, Miao; Doerzaph, Zachary R · 2020 · ROSA P / Safety through Disruption (Safe-D) University Transportation Center (UTC)

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 the real-world usage and safety consequences of Automated Driving Systems (ADS), specifically SAE Level 1 and Level 2 systems, to address gaps in understanding how drivers interact with these technologies on public roadways. While active safety systems have demonstrated crash reduction benefits, ADS capabilities are often misunderstood by drivers, leading to potential misuse, disuse, or abuse. The research aims to quantify the prevalence of ADS use during safety-critical events (SCEs), identify instances of system failure or unintended use, and assess driver perceptions of system usability and usefulness. The researchers analyzed data from the Virginia Connected Corridor Level 2 Naturalistic Driving Study (VCC L2 NDS), involving 50 participants who drove personally owned vehicles equipped with ADS for 12 months. The study focused on 44 participants with high-quality data, capturing 235 SCEs. ADS states and alerts were identified using machine learning algorithms validated by trained data reductionists reviewing video and kinematic data. Additionally, a post-study questionnaire collected subjective ratings on ADS usability and specific scenarios where systems failed to meet driver expectations. The analysis revealed that 20% of all SCEs (47 out of 235) involved ADS use, with longitudinal control features (e.g., Adaptive Cruise Control) used more frequently than lateral control features. In 57.4% of these ADS-involved SCEs, drivers engaged in unintended use, such as engaging in secondary tasks, driving with hands off the wheel, or using the system in adverse weather or non-highway conditions. In 12.8% of cases, the ADS failed to react to the situation or warn the driver. Subjective survey results indicated that drivers found ADS useful and easy to use, particularly longitudinal features. However, this positive perception correlated with increased comfort in engaging in secondary tasks while the system was active. Drivers frequently reported that ADS did not meet expectations in scenarios involving cutting-in vehicles, stopped leads, or blurred lane markings. The findings highlight a significant risk of unintended ADS use driven by driver overconfidence and misunderstanding of system limitations. The study concludes that current human-machine interfaces and training programs may be insufficient to prevent misuse. It recommends developing adaptive, multimodal interfaces and clearer owner’s manuals to manage driver expectations and mitigate safety risks associated with the gap between driver trust and actual system capabilities.

Key finding

Drivers misused automated driving systems in 57% of safety-critical events, and the systems failed to react or warn the driver in 13% of those events.

Methodology

naturalistic

Sample size: 50

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).

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.

• adas effectiveness
• automation surprise
• naturalistic crash near crash
• situational awareness
• driverless ads
• automation

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: observational prevalence
• Methodological Resource: dataset resource, tool software