Haptics based multi-level collaborative steering control for automated driving

Nakade, Tomohiro; Fuchs, Robert; Bleuler, Hannes; Schiffmann, Jürg · 2023 · Crossref

DOI: 10.1038/s44172-022-00051-2

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 trust issues associated with over-reliance on automated driving systems, specifically focusing on the limitations of current haptic shared control (HSC) strategies. Existing systems often suffer from discontinuous operation, where driver intervention triggers an override that deactivates automation, or from poor integration of driver intent into trajectory planning. To resolve these issues, the authors propose a multi-level collaborative steering control framework designed for mass-produced vehicles. The strategy is built on three core functionalities: interaction, which enables haptic communication via admittance control; arbitration, which allocates control authority between the driver and automation based on estimated driver intent; and inclusion, which assimilates sustained driver deviations into the automated driving (AD) trajectory planning. This approach allows for continuous shared control without deactivation, supporting automation levels 0–4. The study validates the proposed framework through four experimental configurations: a virtual driver setup using an impedance-controlled motor to verify estimation algorithms; a human driver setup to test arbitration rules during slalom maneuvers; a static driving simulator to evaluate trajectory adaptation during double lane changes; and a physical test vehicle for proof-of-concept validation. The system estimates the driver’s target angle and impedance using an extended Kalman filter. Arbitration rules adjust the automation’s reaction torque based on four interaction types: cooperation, co-activity, collaboration, and competition. Inclusion logic adapts the AD trajectory when manual deviation is sufficiently large and persistent, effectively allowing driver-initiated rerouting. Results demonstrate that the system successfully estimates driver motor control parameters, though combined estimation exhibits some oscillatory behavior due to modeling errors. In arbitration tests, the system correctly adjusted automation impedance according to the interaction type; for instance, in collaboration mode, automation authority decreased as driver engagement increased, whereas in competition mode, automation resistance increased to oppose driver input. Trajectory adaptation tests showed that when inclusion was active, the AD trajectory shifted to accommodate driver-initiated lane changes, significantly reducing the sustained torque required from the driver compared to when adaptation was deactivated. On the test vehicle, the framework enabled smooth lane changes in collaboration mode with lower torque peaks than co-activity, while competition mode prevented trajectory deviation due to high automation resistance. The significance of this work lies in its integration of interaction, arbitration, and inclusion into a unified control framework, addressing the fragmentation of current ADAS functions. By enabling continuous, override-free shared control and assimilating driver intent into tactical planning, the system enhances driving safety and user acceptance. It provides a practical solution for partial automation that maintains the driver’s active role, thereby fostering trust and ensuring consistent coordination of vehicle actuators across varying automation levels.

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.

• steering pattern
• automation
• hands on hands off engagement
• control interfaces
• manual
• haptic feedback

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).

• Theoretical Contribution: computational model, conceptual framework