Abstract
Human affective touch is known to be beneficial for social-emotional interactions and has a therapeutic effect. For touch initiated by robotic entities, richer affective affordance is a critical enabler to unlock its potential in social-emotional interactions and especially in care and therapeutic applications. Simulating the attributes of particular types of human affective touch to inform robotic touch design can be a beneficial step. Inspired by the scientific finding on CT-optimal affective touch - a gentle skin stroking at velocities of 1-10 cm/s evidenced to be pleasant and calming, we developed a proof-of-concept haptic rendering system - S-CAT, using pneumatic silicone soft robotic material to simulate the attributes (velocity, temperature and applied normal force) of CT-optimal affective touch. To investigate whether the affective touch performed by the S-CAT system elicits psychological effects comparable to CT-optimal, manual affective touch, we conducted an experimental study comparing the effects of CT-optimal versus non-CT-optimal stimulation velocities in each of three types of stimulation modes (S-CAT device, skin-to-skin manual stroking, hairbrush manual stroking), and across them. Our measures included subjective ratings of touch pleasantness and intensity, neurophysiological responses (EEG), and qualitative comments. Our results showed that velocity modulated subjective and neurophysiological responses in each and across these three stimulation modes, and that CT-optimal stimulations from S-CAT system and manual method received similar ratings and verbal comments on pleasantness, suggesting that the S-CAT touch can have comparable effects to manual stroking. We discuss the design insights learned and the design space that this study opens up to support well-being and healthcare.