Improving Immersive Telepresence Locomotion by Using a Virtual Environment as an Interface to a Physical Environment (VEIPE)
Eetu Laukka, Evan G. Center, Başak Sakçak, Steven M. LaValle, Timo Ojala, Matti Pouke · 2024
Immersive mobile robotic telepresence enables humans to feel present in a remote environment. These systems often use 360-degree panoramic cameras to stream video over a network to a head-mounted display (HMD) where the video feed is rendered to the user. This enables the user to freely look around in a remote environment. A drawback of using highly immersive technologies instead of a more traditional computer screen is that users often experience virtual reality (VR) sickness. Therefore, sometimes the users are only able to use these systems for brief durations. Moreover, the increase in bandwidth requirements of panoramic cameras and the time necessary to process the 360-degree panoramic view contributes to an often unacceptable amount of latency between the user’s actions and the observed reaction of the mobile robot, which can be referred to as perception-actuation loop. We present a novel method to mitigate these problems in immersive mobile robotic telepresence systems. We call this method virtual environment as an interface to a physical environment (VEIPE). In VEIPE, a digital twin of the remote environment is used to interface with the telepresence robot in the real remote environment. We present a study comparing teleportation through VEIPE as a locomotion method against a more traditional joystick-based continuous locomotion method for controlling a telepresence robot. Our results indicate that VEIPE induces less VR sickness compared to the joystick condition as measured by the simulator sickness questionnaire (SSQ) and users perform about 31 percent better in a simple navigation task. Furthermore, the users subjectively prefer teleportation through VEIPE over the joystick. We also present exploratory data about cognitive load measured with the NASA task-load-index (NASA-TLX) questionnaire, presence measured with the Slater-Usoh-Steed (SUS) questionnaire, and accumulated yaw in the navigation tasks.