Robot Partner Development Using Emotional Model Based on Sensor Network

This paper discusses the development of robot partner that can perform not only static conversation, but also can perform emotion expression by facial expression, gesture, and word expression using emotional model based on sensor network, therefore it can interact naturally with a person. Generally, the robot has sensors equipped inside it, however to express emotion the equipped sensors are not enough to grasp the necessary input information about the surrounding environmental situation. Therefore we propose a sensor network applied to the robot partner for estimating the environment states as input data, after that the acquired data is processed using emotional model to gain the desired emotional expression. In this paper, first we explain the concept of informationally structured space, robot partner that we are developing, and sensor network. Next, we explain the development of emotional model that consists of data acquisition from sensor network as an input and the model output such as face, gesture, and word expression. Finally, we conduct several experiments based on the proposed method, and discuss the ability of emotional model to develop robot partner that can perform emotional expression.