Navigating Virtual Realms - Discovering the Impact of Head-Mounted Displays on Human Gait using Wearable Sensor Motion Capture

Abstract

This thesis outlines a preliminary investigation into the effects that wearing head-mounted displays (HMDs) have on human gait. Through the use of gait analysis, the research has two objectives: 1) to observe and quantify any changes to gait that may occur among individuals while wearing HMDs; and 2) to compare the data extracted from inertial measurement units (IMUs) to that of an optical motion capture (OMC) system. OMC and IMU data were collected on a pilot dataset of 20 healthy participants subject to trials under three testing conditions: Normal, using an optical seethrough Augmented Reality (AR) display, and using a video see-through Virtual Reality (VR) display. A study protocol was developed that outlined seven movement paths that tested the participants under various scenarios. With these paths repeated for each trial condition, analysis was done to identify any changes to gait. Observing the changes in the extracted gait metrics reflected participants slowing down and acting more cautiously while wearing the HMDs. However, post-hoc analyses on the results frequently revealed low statistical significance and a greater sample size was required. The observed changes included a decreased gait velocity and cadence (aka step frequency), as well as increases in stride time, reaction time, task completion time, and cumulative knee flexion throughout strides. An increase in the cumulative knee and hip flexions and marker/sensor heights were also observed during obstacle clearance. These changes in gait characteristics reflect how the participants felt a sense of instability and unfamiliarity, therefore slowing down and behaving more carefully in order to prevent collisions and/or accidents. These differences were observed in data derived from both the OMC and IMU systems, although the calculated magnitudes varied. Post-hoc analyses comparing the two datasets revealed minimal statistical significance in these differences. The observed decreases in IMU-derived gait metrics likely reflect the accumulation of sensor inaccuracies throughout data collection and processing. In both objectives, it was found that a larger study sample size was required to obtain more statistically significant results. Further development of the study would involve a larger study population, increased immersion in virtual environments, and the extraction of more comprehensive gait metrics to advance the understanding of the effects of extended reality (XR) technology. This would require enhanced resources, in terms of data collection and processing tools. Expanding knowledge on the effects of HMDs and XR technologies can allow for the manufacturing of safer devices, thus potentially mitigating the risk to users.