Shared Control for Telepresence Robot in Dynamic Virtual Reality Environment

Teleoperation in virtual reality (VR) environments, integrating human control and autonomous navigation, is a subject of interest for static and dynamic scenarios. Previous studies primarily focused on static scenes, leaving real-world complexities unaddressed. This research bridges the gap by investigating shared teleoperation in dynamic VR settings. Users navigate static and dynamic scenarios to reach goals, utilizing three modes: autonomous, teleoperation and shared control. In order to validate prior findings on efficacy of shared control, we designed two controlled environments: 1) static, and 2) dynamic. To validate this approach, we conducted a user study involving N = 22 participants in a Virtual Reality (VR) setting. Results show a disparity between subjective and objective measures between the static and dynamic scenarios. Users faced higher cognitive load and frustration with teleoperation in dynamic scenarios. Surprisingly, quantitatively, they performed better with full teleoperation. Shared control, indicated as the preferred control mode by participants, doesn't consistently deliver superior objective performance in dynamic VR scenarios. In this study, we conclude that unlike prior shared control studies, it is of utmost importance to validate the efficacy of shared control in both static and dynamic settings to avoid inconsistent performance in terms of subjective and objective measures.