Fuzzy based emotion generation mechanism for an emoticon robot

Abstract

This paper discusses the architecture in which the emotion of a robot is designed and generated. The proposed architecture integrates long term affective characteristic and short term interactive changes into the emotion generation mechanism of a robot. Emotion of a robot is generated by a fuzzy logic from three inputs: the robot's personality, the ambient environment and the interaction with human. The virtual personality of this robot is expressed via an LED-face using emoticon facial expressions. The robots with different type of personalities were tested in various environments. Test subjects can identify different robot's personality from the interaction.