Learning analysis of obstacle avoidance to beginners at trial times using NASA-TLX for drive assist system of welfare vehicle using Mixed Reality

MATSUNAGA, Nobutomo; NAKAMURA, Reo; TAKEUCHI, Yudai; OKAJIMA, Hiroshi · 2022 · Crossref

DOI: 10.1299/transjsme.21-00241

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 difficulty inexperienced users face when operating welfare vehicles, such as electric wheelchairs, in narrow indoor environments. While previous research proposed using Mixed Reality (MR) and virtual platoon control to assist driving by projecting a virtual vehicle onto a Head-Mounted Display (HMD), a significant limitation remained: instructors could not see the virtual vehicle, preventing them from providing timely, specific guidance. This paper proposes a training system that shares the MR space between the user and an instructor, allowing for real-time coaching to accelerate skill acquisition. The researchers developed a system using Microsoft HoloLens 2 and the STAVi welfare vehicle. The system automatically displays virtual obstacles, an ideal path, and steering start points (pin objects) on the HMD. Crucially, the user’s HMD view is shared with an instructor via Miracast, enabling the instructor to observe the virtual environment and provide immediate advice. An experiment involved 20 novice participants divided into two groups. Group A (n=10) used the proposed shared-space system with virtual objects and instructor guidance. Group B (n=10) used the conventional virtual platoon system without shared space or virtual objects, receiving only post-trial feedback. Participants performed five obstacle avoidance trials on a narrow course. Mental workload was quantified using the NASA Task Load Index (NASA-TLX) after each trial. The results demonstrated that Group A significantly outperformed Group B. Group A achieved an average of 3.0 successful completions per participant, compared to 1.2 for Group B, a difference that was statistically significant (p=0.0056). Furthermore, NASA-TLX analysis revealed that Group A’s Weighted Workload (WWL) scores decreased rapidly, dropping significantly between the first and second trials. By the fifth trial, Group A’s average WWL score was 43.27, significantly lower than Group B’s 53.73 (p=0.0065). Specifically, Group A showed marked reductions in mental demand, effort, and frustration, whereas Group B showed minimal improvement in workload metrics despite repeated trials. The study concludes that sharing the MR space with an instructor and utilizing virtual objects to highlight critical driving cues significantly accelerates the learning curve for beginners. The ability to provide timely, specific guidance reduces mental workload and improves driving proficiency more effectively than traditional methods. These findings suggest that MR-based training systems with shared visibility are valuable tools for welfare vehicle operation, potentially benefiting elderly or disabled users who require efficient skill acquisition.

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.

• in vehicle coaching

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: self report data
• Methodological Resource: tool software