Proxemic-aware Augmented Reality For Human-Robot Interaction

This study introduces a novel proxemic-aware augmented reality (AR) system to mitigate information overload in AR-enabled human-robot interaction (HRI). The system leverages human-robot proxemics to automatically adjust what and how much visual content needs to be presented. Therefore, the operator can perceive the relevant data through AR interfaces without being overwhelmed by excessive information exposure. We propose a task-specific model for evaluating human-robot proxemic (HRP), where the system can identify HRP levels based on raw features, such as distance and orientation. Based on HRP levels, we design a set of visual elements for presenting robots’ information at various levels of detail. To demonstrate the functionality of the system, we present a series of proof-of-concept applications showing that our system can assist the operator in a wide range of HRI tasks. The user study proves that the proxemic-aware AR system can reduce mental loading, increase visual clarity, and improve interaction efficiency in HRI.