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

2025-11-07

AI Summary

In conditional automation, balancing driver takeover timeliness and comfort remains challenging. Method: This study proposes an adaptive time-budgeting framework that introduces a variable takeover buffer time mechanism, integrating a takeover time prediction model with a piecewise function to enable personalized allocation of temporal resources based on individual driver characteristics and dynamic traffic context. A driving simulator experiment—combined with an n-back cognitive load task—systematically evaluates the impact of varying buffer durations on safety metrics (e.g., takeover reaction time, trajectory deviation) and subjective experience. Results: A 5–6 second buffer duration achieves optimal trade-offs across safety, comfort, and robustness under varying traffic densities, significantly improving takeover success rate and user satisfaction. The framework provides a quantifiable, context-adaptive decision-support method for dynamic human–machine authority transition.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Economics, Online Markets and Human Computation: The sharing economy

Abstract

Conditionally automated driving requires drivers to resume vehicle control promptly when automation reaches its operational limits. Ensuring smooth vehicle control transitions is critical for the safety and efficiency of mixed-traffic transportation systems, where complex interactions and variable traffic behaviors pose additional challenges. This study addresses this challenge by introducing an adaptive time budget framework that provides drivers with sufficient time to complete takeovers both safely and comfortably across diverse scenarios. We focus in particular on the takeover buffer, that is, the extra time available after drivers consciously resume control to complete evasive maneuvers. A driving simulator experiment is conducted to evaluate the influence of different takeover buffer lengths on safety-related indicators (minimum time-to-collision, maximum deceleration, and steering wheel angle) and subjective assessments (perceived time sufficiency, perceived risk, and performance satisfaction). Results show that (i) takeover buffers of about 5-6 seconds consistently lead to optimal safety and comfort; and (ii) drivers prefer relatively stable takeover buffers across varying traffic densities and n-back tasks. This study introduces an adaptive time budget framework that dynamically allocates transition time by incorporating a predicted takeover time and a preferred takeover buffer (piecewise function). This can serve as an important first step toward providing drivers with sufficient time to resume vehicle control across diverse scenarios, which needs to be validated in more diverse and real-world driving contexts. By aligning the provided time budget with driver needs under specific circumstances, the adaptive framework can improve reliability of control transitions, facilitate human-centered automated driving, reduce crash risk, and maintain overall traffic efficiency.

Problem

• Developing adaptive time budgets for safe control transitions in automated driving
• Determining optimal takeover buffer lengths for driver safety and comfort
• Addressing variable traffic behaviors during vehicle control transition scenarios

Innovation

• Adaptive time budget framework for vehicle control transition
• Dynamic allocation of transition time using predicted takeover time
• Preferred takeover buffer of 5-6 seconds for safety