Deploying Multi-Modal Communication Using Augmented Reality in a Shared Workspace

Abstract

Robots are no longer working isolated in safety fences and Human-Robot Collaboration (HRC) is becoming one of the most promising topic of research to improve the efficiency in many application scenarios. Sharing the same workspace, both human and robot should clearly understand the intentions and motions of each other, in order to enable an efficient and effective interaction. In this work we propose an AR-based system to show the robot planned motion and target to the worker. We focused on representing this information in an intuitive way for inexperienced users. We introduced a multi-modal communication feedback in order to enable the user to agree with or change the robot plan using gestures and speech. The effectiveness of the system has been evaluated with test cases performed by a group of testers with no robotic experience. The results showed that the system helped the user to better understand the robot intentions and planned motion, improving the ergonomics and trust in the interaction. Furthermore, the evaluation included the rating of the different input modalities provided, in order to compare the different ways of communication proposed.