Researchers at Toyohashi University of Technology have examined how the movement patterns of autonomous mobile robots affect human emotional responses during close encounters, finding that unpredictable behavior can prolong discomfort and physiological arousal. The study focused on passing scenarios in a virtual reality environment and measured both subjective emotional ratings and physiological indicators associated with stress.
The research was conducted by teams from the university’s Visual Perception and Cognition Laboratory and Cognitive Neurotechnology Unit. Participants navigated a virtual corridor while a robot approached from the opposite direction, allowing the researchers to isolate the emotional effects of different robot motion patterns without physical risk. Emotional responses were assessed using self-reported valence and arousal ratings, alongside palm skin conductance measurements, which are commonly used as an indicator of sympathetic nervous system activity.
In an initial experiment, the robot moved along a straight, continuous path and approached participants from varying directions. The direction of approach did not produce measurable differences in emotional or physiological responses. Over repeated encounters, however, arousal levels and skin conductance responses declined, indicating habituation. The researchers attributed this effect to the simplicity and predictability of the robot’s movement.
A second experiment introduced uncertainty into the robot’s behavior. In this condition, the robot either stopped and restarted unexpectedly, remained stationary, or continued moving smoothly. Encounters involving the stop-and-go motion produced higher reported discomfort and elevated arousal and skin conductance responses compared with the other behaviors. Unlike the predictable movement condition, these responses did not diminish over repeated trials, suggesting that participants did not habituate to the unpredictable behavior.
The researchers interpret this sustained response as evidence that uncertainty in robot motion disrupts human expectations and maintains a heightened state of alertness. The findings are consistent with psychological research indicating that unpredictable stimuli can provoke anxiety or aversive reactions. According to the authors, predictability may therefore be a key consideration in designing robot behaviors for environments shared with humans.
The study was led by Yuta Matsubara, a graduate of Toyohashi University of Technology’s master’s programs in computer science and extended reality. He stated that understanding emotional reactions to robots is necessary as autonomous systems become more common in daily settings, and that insights from controlled virtual experiments can inform future design choices.
The results were published in the peer-reviewed journal International Journal of Social Robotics. The research team plans to extend the work to more complex virtual environments, including curved pathways, variable corridor widths, and scenarios involving multiple robots or pedestrians, with the aim of developing behavioral design guidelines that account for human emotional responses.
