Distraction by a human or a robot: Effects of perceptual load and action type

Abstract

This study investigates how humans process and attend to robot actions compared to human actions under varying cognitive demands. Using a perceptual load paradigm, we examined whether robots capture attention similarly to humans and how this is influenced by the nature of their actions. Participants performed a letter detection task while being presented with task-irrelevant videos of either human or robot agents performing communicative or noncommunicative actions. Results demonstrated that both humans and robots captured attention through their actions, particularly when these actions were communicative. Under high perceptual load conditions, human distractors caused more interference than robot distractors, suggesting that agent identity becomes particularly important when cognitive resources are limited. These findings provide insights into how humans process robot actions in attentiondemanding situations and may have important implications for designing robot behaviors in operational contexts.