Exploring Visual Augmentations for Improving the Operation of a Hydraulic Excavator using Expert Operation Replay

Abstract

Hydraulic excavators are widely used in construction work owing to their versatility. However, the general operation of these excavators is complex and novice operators require extensive training to operate them. In this study, we propose a virtual reality (VR)-based training system with three types of visual augmentations using pre-recorded expert operations to support the skill acquirement for handling a hydraulic excavator. To evaluate the effectiveness of the proposed visual augmentations in terms of skill improvement, we compared the scores of the trainees before and after training including combinations of visual augmentations. The results indicated that the display of the lever movement significantly improved the trajectory of the bucket tip, while ghost animation and slow motion did not show significant effects. Furthermore, by showing the lever input and excavator movement of the expert in slow motion, the task completion time increased because of the aftereffect. Our findings not only provide a design guideline for VR-based excavator operation training but can also be applied to augmented reality (AR)/mixed reality (MR) support systems for supporting practical excavator operations.