Visual perception of robot movements – How much information is required?

Abstract

Human-robot interactions are steadily increasing in all areas of life. In this context, a common motion learning process of human-robot dyads has not been studied so far. The observation of movement characteristics plays a crucial role in the assessment and learning of movements in human-human dyads. But what visual information of a robot movement can be perceived and predicted by humans? The following study examines the perception and prediction of robot putt movements by humans with different visual stimuli. Relevant clues could be identified for the specific movement. Ultimately, with sufficient visual information, humans are able to correctly predict the outcome of a robot putt movement.