Trust-building Robot training Using perspective Switching in XR
Modern manufacturing increasingly relies on robot-assisted quality inspection with sensors, cameras, and AI to detect defects quickly and reliably, reducing human error and boosting productivity. Despite these advantages, undetected faults can trigger costly recalls, making trust in such systems critical for successful adoption and day-to-day operation. Training that explains how these systems work and what their limitations are is therefore essential to cultivate appropriate trust while avoiding overtrust.
TRUST is a collaboration between the University of Bamberg, Mindport GmbH, and Bamberg’s AI Experience Center for SMEs.
Technology
The TRUST project explores virtual reality (VR) as a training medium to build trust in robot-assisted quality inspection and counter overtrust. VR uniquely allows users to switch perspectives and execute tasks from a robot’s point of view – an experience impossible to realize in the real world. This perspective-switching is used as an innovative didactic method to deepen cognitive understanding of quality inspection features (e.g., how sensors and AI algorithms detect deviations) and to foster affective trust, while simultaneously making system limitations explicit to mitigate overtrust.
The project will design, implement, and evaluate this VR training at the University of Bamberg’s AI Experience Center, involving both business professionals and students. Empirical insights will illuminate how robot perspective-taking in VR shapes key psychological processes in human–robot interaction. Ultimately, TRUST aims to demonstrate how targeted XR trainings can enhance the practical deployment and use of robot-assisted quality inspection by improving users’ understanding, confidence, and decision-making.
