Learning from Drivers: A Case-Based Reasoning Framework for Takeover Control in Conditionally Automated Vehicles
DOI: 10.54941/ahfe1007861
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 addresses the safety and reliability challenges associated with takeover control (TOC) in conditionally automated vehicles (CAVs), specifically SAE Level 3 systems. In these vehicles, drivers may disengage from driving tasks but must resume control when the system reaches its operational limits. The authors identify a critical tension: while fixed regulatory time budgets (e.g., the 4-second minimum in UN Regulation No. 157) exist, they fail to account for the dynamic variability of real-world scenarios, driver states, and environmental conditions. Existing machine learning and deep learning models for predicting takeover performance often lack interpretability, generalize poorly to unseen scenarios, and do not explicitly reuse experiential knowledge. To address these limitations, the authors propose a Case-Based Reasoning (CBR) framework that models, evaluates, and improves decision-making during control transitions by learning from empirical data. The proposed framework implements a four-stage CBR cognitive cycle: retrieve, reuse, revise, and retain. Each takeover case is represented as a multi-dimensional feature vector comprising four categories of identifiers: human agent traits (gender, age, personality, trust), non-driving-related task characteristics (type, duration), environmental conditions (traffic density, scenario type), and takeover performance metrics (takeover time, reaction times, deceleration, steering stability, time to collision). When a takeover request is issued, the system retrieves the most similar historical cases from a repository using the Gower similarity distance, which handles mixed numerical and categorical data. The system selects the case with the best historical performance score to determine the initial time budget for the new scenario. If no sufficiently similar case exists, a conservative default of 4 seconds is applied. Following the takeover, the system enters the revision stage. It compares the observed takeover performance against predefined thresholds. If the reused time budget resulted in performance violations, the system revises the budget by adjusting it based on the difference between the observed and retrieved takeover times. This revised case is then retained in the knowledge repository, allowing the system to continuously learn and adapt. The framework ensures safety by constraining the minimum time budget to 0.67 seconds, corresponding to average human perception-reaction time. The significance of this work lies in its shift from static, fixed-time budgets to adaptive, interpretable decision-making. By leveraging past experiences, the CBR framework supports bidirectional learning between drivers and automated systems, personalizing takeover guidance based on individual driver capabilities and specific contextual factors. The authors conclude that this approach enhances safety and reliability in mixed-autonomy traffic environments. While the framework is currently under development and validation through simulation and experimental studies, it provides a foundational model for future Levels 4 and 5 vehicles to incorporate human-like reasoning in safety-critical decisions. Future work will focus on real-world validation and the integration of real-time physiological measures to further refine driver readiness assessment.
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 | pdftotext | — | — | 4 | 2026-08-10 |
| clean | success | clean | — | — | 2 | 2026-08-10 |
| chunk | success | chunk | — | — | 2 | 2026-08-10 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 2 | 2026-08-10 |
| promote | success | — | — | — | 1 | 2026-08-09 |
| summarize | success | llm | qwen3.6-27b-nvidia | summ-v5 | 2 | 2026-08-10 |
| tag | success | vector_similarity | — | — | 17 | 2026-08-11 |
| verify | partial | — | — | — | 2 | 2026-08-10 |
Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified_with_issues.
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).
- Empirical Findings: behavioral performance data
- Theoretical Contribution: conceptual framework, computational model