Towards Safe and Comfortable Vehicle Control Transitions: A Systematic Review of Takeover Time, Time Budget, and Takeover Performance

2025-07-29

AI Summary

In conditional automated driving, driver takeover is required upon system failure, yet substantial inter-individual variability in takeover time renders fixed time budgets inadequate for simultaneously ensuring safety and comfort. To address this, we propose the concept of “takeover buffering” and a dual-classification framework, establishing a human–vehicle interaction theoretical model grounded in a task–capability interface. Through systematic literature review, causal factor analysis, and standardized measurement, we uncover the dynamic relationships among takeover time, time budget, and takeover performance. We further develop a unified conceptual framework and evaluation system to support the design of context-adaptive and personalized time budgeting mechanisms. The study outlines six future research directions, providing both theoretical foundations and practical pathways to enhance takeover reliability, mitigate risks, and optimize user experience.

Technology Category

• Cognitive Modeling & Cognitive Systems: Adaptive Behavior
• Humans and AI: Planning and Decision Support for Human-Machine Teams
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Studies of user behavior, including longitudinal effects of personalized systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Conditionally automated driving systems require human drivers to disengage from non-driving-related activities and resume vehicle control within limited time budgets when encountering scenarios beyond system capabilities. Ensuring safe and comfortable transitions is critical for reducing driving risks and improving user experience. However, takeovers involve complex human-vehicle interactions, resulting in substantial variability in drivers' responses, especially in takeover time, defined as the duration needed to regain control. This variability presents challenges in setting sufficient time budgets that are neither too short (risking safety and comfort) nor too long (reducing driver alertness and transition efficiency). Although previous research has examined the role of time budgets in influencing takeover time and performance, few studies have systematically addressed how to determine sufficient time budgets that adapt to diverse scenarios and driver needs. This review supports such efforts by examining the entire takeover sequence, including takeover time, time budget, and takeover performance. Specifically, we (i) synthesize causal factors influencing takeover time and propose a taxonomy of its determinants using the task-capability interface model; (ii) review existing work on fixed and adaptive time budgets, introducing the concept of the takeover buffer to describe the gap between takeover time and allocated time budget; (iii) present a second taxonomy to support standardized and context-sensitive measurement of takeover performance; (iv) propose a conceptual model describing the relationships among takeover time, time budget, and performance; and (v) outline a research agenda with six directions.

Problem

• Determining sufficient time budgets for safe vehicle control transitions
• Understanding variability in driver takeover time and performance
• Improving standardized measurement of takeover performance metrics

Innovation

• Taxonomy of takeover time determinants using task-capability model
• Concept of takeover buffer for adaptive time budgets
• Standardized context-sensitive takeover performance measurement