Engaging Older Adults With a Persona-Switching Baby-Type Robot: A Study of Acceptance, Physical Interaction Styles, and Behavior Recognition

The development of robots for elderly care support is advancing. Care robots such as PARO aim to foster user attachment and contribute to mental well-being, particularly through physical interactions including touching and soothing. We focused on a “dress-up” approach inspired by commercial pet robots, which enables easy changes in appearance and personality, to further promote attachment to such care robots. We developed a baby-type care robot that allows users to change its persona (appearance and personality) by swapping facial masks. This paper presents a three-part investigation based on this robot. First, we evaluated the acceptance of this approach among older adults, and found that older adults accepted the developed persona-switching system from the perspectives of perceived enjoyment and intention to use. Second, we analyzed the effects of different appearances on the styles of physical interaction, and found the facial mask effects when the robot wore the baby face mask, causing smoother user movements compared to the male/female masks. Finally, we established a technology to accurately recognize 10 physical interactions by achieving more than 80% accuracy on average.