Model-based Evaluation of Driver Control Workloads in Haptic-based Driver Assistance Systems

Mbanisi, Kenechukwu C.; Kimpara, Hideyuki; Li, Zhi; Prokhorov, Danil; Gennert, Michael A. · 2022 · arXiv

archive: archived pipeline: cataloged verified

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

Summary

This paper addresses the challenge of objectively evaluating driver control workload and dynamic interactions in haptic-based driver assistance systems (ADS). Existing studies rely on subjective human experiments or instrumented vehicles, which often fail to capture internal workload variables or provide scalable evaluation methods. To overcome these limitations, the authors propose a novel simulation framework using a Human Model-based Active Driving System (HuMADS). This approach utilizes surrogate human driver models to simulate the biomechanical and dynamic interactions between a driver and an ADS during a vehicle overtaking task, enabling the quantification of hard-to-measure internal workload metrics. The study employs the HuMADS framework, built on the OpenSim biomechanics platform and SimBody physics solver. The system integrates a human driver model with an automated vehicle controller in a closed-loop feedback system. Two driver arm-steering impedance modes—tense and relaxed—were developed and validated against experimental data from prior literature. The simulation evaluated three haptic shared control conditions: No Conflict (aligned goals and trajectories), Conflict-I (mismatch in planned maneuver), and Conflict-II (mismatch in resulting trajectory). Performance and workload were assessed using metrics such as lateral position error, steering torques, reaction forces, and joint actuation effort (control stress and load quantity) for shoulder and elbow joints. Results indicate that No Conflict shared control scenarios improved driving performance and reduced control workloads compared to manual control. Specifically, relaxed drivers in No Conflict conditions achieved steering accuracy comparable to tense drivers in manual control but with lower effort. Conversely, Conflict scenarios led to unsafe maneuvers and increased workloads. Conflict-I resulted in the largest lateral position error (0.64 m) and highest control stress, while Conflict-II caused the ADS torque actuators to saturate at their 5 Nm limit, leading to high joint actuation efforts. The simulation successfully replicated experimental overtaking trajectories with maximum lateral errors under 0.35 m for manual modes, validating the model's accuracy. The significance of this work lies in demonstrating that surrogate human models can effectively simulate human-vehicle interactions to evaluate ADS designs before physical implementation. The findings confirm that haptic shared control reduces driver workload when system goals are aligned but increases physical stress and safety risks during control conflicts. This framework provides a tool for optimizing ADS tuning and conflict resolution strategies, ultimately informing the design of safer and more acceptable driver assistance systems.

Key finding

No-Conflict haptic shared control (relaxed driver) produced the lowest driver steering torque (1.04 Nm vs 2.47 Nm in manual control), lowest peak hand reaction force (4.2 N vs 30.1 N for Conflict-I), and lowest control stress/load quantity, while Conflict-I conditions saturated the ADS torque limit and produced the highest workload, confirming that aligned haptic guidance reduces driver effort whereas conflicting guidance increases it.

Methodology

modeling

Sample size: N/A (model-based simulation; driver models validated against Pick & Cole 2007 experimental human-subject steering data; vehicle inputs from Naranjo et al. 2008)

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 discover_arxiv on 2026-05-07 (3 acquisition events logged).

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

Topics

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

• automation
• manual
• control interfaces
• haptic feedback
• steering pattern
• teleoperation remote driving

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, theory or model, conceptual framework