Prediction of Emotions Caused by Robotic Tactile Stimulation

Abstract

Nonverbal communication via touching affects human emotions. This study aims to investigate the effect of robotic touch on human emotions. Russell's circumplex model that explains human emotions along orthogonal arousal and valence axes was employed for the investigation, and three principal motions （grasp, tap, and stroke） were implemented to investigate the human emotional response to touch. The results showed that human emotions were affected by the touching speed and pressing force of a robotic touching unit. To cause desirable emotions, motion parameters were designed based on experimental results and employed to examine the emotional response to the motions. The arousal responses could be predicted by a superposition of the responses of the three principal motions.