Locomotion in virtual reality (VR) is one of the biggest problems for large scale adoption of VR applications. Yet, to our knowledge, there are few studies conducted in-the-wild to understand performance metrics and general user preference for different mechanics. In this paper, we present the first steps towards an open framework to create a VR locomotion benchmark. As a viability study, we investigate how well the users move in VR when using three different locomotion mechanics. It was played in over 124 sessions across 10 countries in a period of three weeks. The included prototype locomotion mechanics are arm swing, walk-in-place and trackpad movement. We found that overall, users performed significantly faster using arm swing and trackpad when compared to walk-in-place. For subjective preference, arm swing was significantly more preferred over the other two methods. Finally for induced sickness, walkin-place was the overall most sickness-inducing locomotion method.

Paper link here.