Virtual Reality Assembly of Physical Parts: The Impact of Interaction Interface Techniques on Usability and Performance

Abstract

In the last few years there has been increased interest in Virtual Reality technology application in the professional context. The technology is used in industry for training. There are several reported successful pilot applications in the industrial training of assembly of physical parts. However, our understanding of how interaction techniques affect performance on these tasks is still limited. In this research, we study the usability of three common hand-based interaction methods: controllers, camera-based hand tracking, and dataglove interaction in the virtual assembly of physical parts. We investigate how performance on the task is affected by the interaction method. We measure the usability, task time, task accuracy, and interaction errors in a study with 12 participants (n = 12) - with and without prior Virtual Reality experience. The results show that users rate the usability of controllers higher than camera based tracking, while datagloves are rated lower than camera tracked solutions. Users also complete the task faster, and with fewer interaction errors when using a controller, than when using a camera-tracking solution, or dataglove. However users execute actions more precisely when using a dataglove. Participants gender does not play a significant rule on the reported usability or task performance. However, as participants experience with Virtual Reality increases they can become less precise at executing actions with the controller, and more precise with the dataglove. The results of this study can help to design systems for the Virtual Reality assembly of physical parts, and to make decisions on which interaction method is suitable based on which aspects of the system are most important.