An Emotional Model Based on Location-Dependent Memory for Partner Robots

Abstract

With user expectations for user-friendly robots growing, the question of what makes a robot “user-friendly” continues to be debated. With the human need to add an “emotional” aspect to robots – “humanoid” shape such as Asimo and “fuzzy-feely” appeal such as Paro – we propose an emotional model based on location-dependent memory for partner robots. Focusing on the functions of emotion in social interaction, our proposed model is based on emotions, feelings, and mood and “episodic” memory related to changes in feeling. We propose map building and behavior control based on the emotional model. Experimental results demonstrate the feasibility of the emotional model and related behavior based on location-dependent memory.