Switching from autopilot to the driver: A transient performance analysis

Venkita, Seshan Ramanathan; Willemsen, Dehlia; Alirezaei, Mohsen; Nijmeijer, Henk · 2019 · Crossref

DOI: 10.1177/0954407019878540

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 safety challenges associated with semi-autonomous vehicles, specifically focusing on the transient dynamics that occur when control switches from an autonomous system to a human driver. The primary research question is how to quantify the safety of this transition, particularly for vehicle-initiated, vehicle-to-driver (VI-V2D) take-overs, where the driver’s situational awareness may be low. The authors aim to develop a model-based "switching performance indicator" that provides a bound on the amplitude of transient dynamics using a priori system knowledge, thereby offering a systematic method to assess take-over safety beyond human-factors observations. The methodology employs a linear switched-system approach, modeling the vehicle’s lateral dynamics, the human driver, and the autonomous controller. The vehicle is represented by a single-track model, while the driver is modeled using a Pursuit–Compensatory cognitive controller combined with a quasi-linear neuromuscular dynamics model. The autonomous system is emulated by a path-tracking controller with independent longitudinal and lateral control. The study assumes constant longitudinal speed and focuses on the lateral control switch. By deriving the closed-loop dynamics for both the controller-vehicle and driver-vehicle systems, the authors establish the initial conditions for the driver system at the moment of switch, ensuring continuity in steering input and its derivatives. The core analytical contribution is the derivation of a bound on the transient response using the induced $L_\infty$ norm. This bound serves as the basis for the switching performance indicator, which checks whether critical states and outputs remain within acceptable limits during the transition. The indicator is validated through simulations of a lane-change maneuver at 100 km/h, comparing the transient responses of autonomous-only, manual-only, and switched scenarios. The results demonstrate that the indicator successfully captures the effects of various system parameters on take-over safety. Although the bound is slightly conservative, it accurately reflects the transient behavior, including peaks in lateral position, acceleration, and steering angle. The significance of this work lies in providing a rigorous, model-based tool for the design and evaluation of shared-control systems in future vehicles. By quantifying the safety-critical aspects of transient dynamics, such as amplitude and settling time, the proposed indicator allows engineers to assess and optimize the safety of control transitions before deployment. This approach bridges the gap between human-factors studies and systematic control theory, offering a predictive measure for ensuring safe handovers in semi-autonomous driving environments.

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.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

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

• takeover transitions
• automation
• steering pattern
• automation surprise
• manual
• lane positioning

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, conceptual framework