A multi-modal interaction robot based on emotion estimation method using physiological signals applied for elderly^*

Abstract

In recent years, robots that estimate emotions in real time have been proposed and are expected to be used in nursing care and at home. Especially in the nursing care field, there is a demand for both reduction of care burden and maintenance, and improvement of quality. Among emotion estimation technologies, those that incorporate methods using physiological signals can constantly acquire real-time human emotional reactions using physiological signals such as EEG and HRV from wearable sensors. We believe that if this technology can be used for mental care of the elderly, it will be possible to reduce the burden and improve the quality of nursing care. For this purpose, we prototyped a robot that responds to emotion of humans in real time, and making facial expressions, body movements, and speeches to improve or maintain the emotional state of user, and we conducted experiments to apply this robot to three elderlies. As the result of experiments, it was confirmed that the emotional states of them improved during the interaction with the robot and their ratios of voluntary utterance, when they could hear the robot, ranged between 58% to 81%.