A Dataset on Takeover during Distracted L2 Automated Driving
DOI: 10.1038/s41597-025-04781-8
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper introduces TD2D, a novel dataset designed to address the lack of publicly available data on driver takeover performance during Level 2 (L2) automated driving while engaged in secondary tasks. The motivation stems from the increasing prevalence of in-vehicle systems that enable non-driving activities, which can distract drivers and impair their ability to safely take control of the vehicle when automation fails. Existing datasets are insufficient because they either focus on manual driving contexts or lack specific takeover scenarios under varied distraction conditions. TD2D aims to facilitate research into how different types of secondary tasks impact driver response times and safety during critical takeover events. The study was conducted using a driving simulator modified from the CARLA software to simulate L2 automated driving scenarios. Fifty participants, balanced by gender and spanning five age groups, were recruited. Each driver completed 10 scenarios involving distinct secondary task conditions: one baseline (no task), three reference auditory tasks (0-back, 1-back, and 2-back cognitive loads), three naturalistic auditory tasks (audiobook listening, auditory gaming, and auditory texting), and three visual tasks (e-book reading, gaming, and texting). During each scenario, the automated vehicle drove at 50 km/h until a critical event occurred, such as a pedestrian jaywalking or a leading vehicle stopping suddenly, requiring the driver to take over control. Data collection included takeover performance metrics, physiological signals (ECG, heart rate, PPG, EDA) via wearable sensors, ocular data (gaze, fixation, pupil diameter) via eye-tracking headsets, and subjective workload assessments using the NASA-TLX questionnaire. The resulting dataset comprises 500 cases, providing multimodal data on driver behavior, physiological states, and cognitive workload across the 10 secondary task conditions. The dataset captures detailed takeover performance metrics, including reaction times and maneuvering behaviors, alongside synchronized physiological and ocular data. This comprehensive collection allows for both exploratory and hypothesis-driven analyses of how visual and auditory distractions, as well as varying cognitive loads, affect a driver’s ability to respond to automation failures. The authors highlight that TD2D is currently the only publicly available dataset that specifically provides takeover performance data in L2 automated driving contexts with diverse secondary task variables. The significance of this work lies in its potential to advance the development of safer automated driving systems. By providing rich, reproducible data on distracted driving during automation, the dataset enables researchers to better understand the mechanisms of driver distraction and its impact on safety-critical takeovers. This information is crucial for designing improved human-machine interfaces, developing effective monitoring systems for driver state, and establishing guidelines for secondary task engagement during automated driving. The authors anticipate that TD2D will serve as a valuable resource for the scientific community in mitigating risks associated with distracted driving in partially automated vehicles.
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.
- distraction detection algorithms
- temporal
- manual
- takeover transitions
- visual manual
- hands on hands off engagement
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: measurement protocol
- Theoretical Contribution: conceptual framework