Adaptive Time Budgets for Safe and Comfortable Vehicle Control Transition in Conditionally Automated Driving

Liang, Kexin; Calvert, Simeon C.; Lint, Hans van · 2025 · ArXiv.org

DOI: 10.48550/arxiv.2511.05744

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the critical challenge of ensuring safe and comfortable vehicle control transitions in conditionally automated driving (CAD). When automation reaches its operational limits, drivers must resume manual control promptly. Existing systems often rely on fixed time budgets (e.g., 7–10 seconds), which fail to account for the significant variability in drivers’ required takeover times (ToT) and diverse traffic scenarios. This mismatch can lead to safety risks if the budget is too short or reduced efficiency and driver distraction if it is too long. The research specifically investigates the "takeover buffer"—the surplus time remaining after a driver consciously resumes control—to determine optimal durations that balance safety and comfort across varying conditions. To evaluate this, the authors conducted a driving simulator experiment at Delft University of Technology involving 57 participants. The study utilized a 3x3 repeated measures design, varying traffic density (0, 10, and 20 vehicles/km) and cognitive load via n-back tasks (n=0, 1, 2). Participants performed takeover maneuvers triggered by a fixed 7-second time budget, involving evasive lane changes to avoid collisions. Data collection included objective metrics (minimum time-to-collision, maximum deceleration, steering wheel angle) and subjective assessments (performance satisfaction, perceived time sufficiency, and perceived risk). The analysis focused on how the takeover buffer, decomposed into safety and comfort components, influenced these performance indicators. The results demonstrated that fixed time budgets are insufficient for ensuring consistent safety and comfort, particularly under high traffic density. Analysis of the takeover buffer revealed that buffers of approximately 5–6 seconds consistently yielded optimal outcomes. Subjectively, performance satisfaction plateaued at this range, perceived risk dropped to near zero, and drivers generally felt time was insufficient for shorter buffers. Objectively, longer buffers correlated with increased minimum time-to-collision and reduced maximum deceleration and steering angles, indicating smoother maneuvers. Notably, 5-second and 6-second buffers produced nearly identical performance effects, suggesting a saturation point where additional time does not significantly improve outcomes. Drivers also preferred relatively stable takeover buffers regardless of traffic density or cognitive task difficulty. Based on these findings, the paper proposes an adaptive time budget framework that dynamically allocates transition time by combining a predicted driver takeover time with a preferred, saturated takeover buffer (piece-wise function). This approach aims to provide sufficient time for safe maneuver execution while maintaining a psychological comfort margin. The study concludes that aligning time budgets with specific driver needs and situational demands can improve the reliability of control transitions, reduce crash risks, and enhance user experience in mixed-traffic environments. This framework serves as a foundational step toward more human-centered automated driving systems, though further validation in real-world contexts is recommended.

Key finding

Takeover buffers of approximately 5 to 6 seconds consistently lead to optimal safety and comfort outcomes, with drivers preferring stable buffer durations across varying traffic densities and cognitive tasks.

Methodology

simulator

Sample size: 57

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 author_sweep_intake on 2026-05-29.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified_with_issues.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• takeover transitions
• automation
• temporal
• automation surprise

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