A Virtual Reality Interface for Teleoperating Mobile Robots in Exploratory Tasks

Human-robot interaction (HRI) deals with the interaction between humans and robots, which is becoming increasingly important as automation progresses. The ongoing development in robotics is opening up more and more possible applications, whereas human control continues to pose a challenge. In particular, the remote control of robots by humans over distances requires special means. The visualization of the robot's environment represents a central challenge. Traditional methods using 2D screens to display content and mouse and keyboard to control the robot reach their limits. Overcoming this challenge requires immersive technologies such as Virtual Reality (VR), Augmented Reality (AR) or Mixed Reality (MR), which enable innovative approaches to visualization and control. This thesis presents a system for teleoperating a mobile robot in a real environment using VR. The basis for this is a real-time 3D reconstruction from the image data of a camera mounted on the robot. The system for visualizing the environment and controlling the robot in VR was developed using the Unity3D game engine. Two navigation metaphors were implemented to enable human operators to control a physical robot in the virtual representation of the real environment. A user study was conducted to assess and compare the performance, usability, and intuitiveness of both metaphors. The task of the user study was to navigate the robot and to investigate the environment to find the target point. The results showed significant differences in objective performance, but almost no deviation in subjective perception. Both metaphors proved to be suitable for the teleoperation of a mobile robot in VR.