Partner Agent Showing Continuous and Preceding Daily Activities for Users’ Behavior Modification

Abstract

In this research, we aimed to support improvement in users’ daily lives by behavior modification for people who cannot self-manage their daily activities, such as bed-making, cleaning, tidying, and sleeping. We focused on the preceding behaviors of others who get along with users to encourage them to act on daily matter. Our proposed system adopts an anthropomorphic animation agent that shows its own activities in daily life to users to stimulate their incentives to follow actions of the agent as a familiar and ambient presence. We conducted a series of four-day experiments to investigate whether the preceding behaviors of the agent were repeated and continuously affected users’ spontaneous actions. From the results, we suggest that the proposed system has a possibility to induce daily activities of users that will ultimately become spontaneous.