Modeling Driver Behavior during Automated Vehicle Platooning Failures
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 research addresses the critical safety challenge of control transitions in automated vehicles (AVs), specifically focusing on driver behavior during automation failures. The study was motivated by the need for accurate driver process models that can predict human responses to AV takeovers, particularly in "silent failure" scenarios where automation disengages without warning. Such models are essential for designers to simulate safety outcomes and calibrate AV algorithms without relying solely on costly and risky on-road testing. The project aimed to identify influential factors in takeover performance and develop predictive models for driver decision-making, braking, and steering. The methodology comprised four phases: a comprehensive literature review, analysis of naturalistic driving data, a driving simulation experiment, and model development. The literature review identified key factors affecting takeover time and quality, such as time budget, secondary tasks, and alert modality, while highlighting gaps in modeling silent failures and gaze eccentricity. To address these gaps, the researchers analyzed 286 rear-end crash and near-crash events from the SHRP 2 naturalistic driving study using machine learning (Random Forest) to predict evasive maneuvers. Additionally, a driving simulation experiment examined driver responses to silent versus alerted failures in platooning scenarios. Finally, the team fitted models to predict post-takeover braking and steering control, comparing visual looming-based models against traditional reaction time and closed-loop models. The findings demonstrated that visual parameters, particularly visual looming (the ratio of angular size to its rate of change), are superior predictors of driver responses compared to traditional baseline models. Analysis of the SHRP 2 data revealed that the gaze eccentricity of the driver’s last glance plays a critical role in decision-making, distinguishing between "eyes-on-threat" and "eyes-off-threat" scenarios. The simulation results confirmed that silent failures significantly increase takeover times and reduce safety margins compared to alerted failures. Furthermore, the modeling analysis showed that evidence accumulation models based on visual looming captured driver braking and steering behaviors more accurately than conventional approaches. The significance of this work lies in providing a robust framework for AV software designers to simulate and improve human-AV interactions. By validating that visual looming and gaze behavior are primary drivers of emergency responses, the study offers a more accurate tool for predicting safety outcomes during control transitions. These models can help designers optimize takeover requests and automation algorithms to mitigate risks associated with silent failures and driver distraction. The research underscores the importance of integrating visual perception mechanisms into driver models to ensure safe interactions between humans and automated systems.
Key finding
Models based on visual looming captured driver responses better than traditional baseline reaction time and closed-loop models, and gaze eccentricity of the last glance plays a critical role in driver decision-making.
Methodology
mixed_methods
Sample size: 286
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.
- takeover transitions
- automation surprise
- situational awareness
- automation
- braking response
- driver post crash behavior
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
- Theoretical Contribution: computational model, theory or model