並列タイトル等生命維持にかかわる生理現象を介した人間 : ロボットのコミュニケーションと身体情動モデルの設計
一般注記In this dissertation, we focus on physiological phenomena of robots as the expressive modality of their inner states and discuss the effectiveness of a robot expressing physiological phenomena, which are indispensable for living. We designed a body-emotion model showing the relationship between a) emotion as the inner state of the robot and b) physiological phenomena as physical changes, and we discuss the communication between humans and robots through involuntary physiological expression based on the model. In recent years, various robots for use in mental health care and communication support in medical/nursing care have been developed. The purpose of these systems is to enable communication between a robot and patients by an active approach of the robot through sound and body movement. In contrast to conventional approaches, our research is based on involuntary emotional expression through physiological phenomena of the robot. Physiological phenomena including breathing, heartbeat, and body temperature are essential functions for life activities, and these are closely related to the inner state of humans because physiological phenomena are caused by the emotional reaction of the limbic system transmitted via the autonomic nervous system. In human-robot communication through physical contact, we consider that physiological phenomena are one of the most important nonverbal modalities of the inner state as involuntary expressions. First, we focused on the robots' expression of physiological phenomena, proposed the body-emotion model (BEM), which concerns the relationship between the inner state of robots and their involuntary physical reactions. We proposed a stuffed-toy robot system: BREAR―which has a mechanical structure to express the breathing, heartbeat, temperature and bodily movement. The result of experiment showed that a heartbeat, breathing and body temperature can express the robot's living state and that the breathing speed is highly related to the robot's emotion of arousal. We reviewed the experimental results and emotional generation mechanisms and discussed the design of the robot based on BEM. Based on our verification results, we determined that the design of the BEM-which involves the perception of the external situation, the matching with the memory, the change of the autonomic nervous parameter and the representation of the physiological phenomena - that is based on the relationship between the autonomic nervous system and emotional arousal is effective. Second, we discussed indirect communication between humans and robots through physiological phenomena - which consist of the breathing, heartbeats and body temperature - that express robots' emotions. We set a situation with joint attention from the robot and user on emotional content and evaluated whether both the user's emotional response to the content and the user's impression of relationship between the user and the robot were changed by the physiological expressions of the robot. The results suggest that the physiological expression of the robot makes the user's own emotions in the experience more excited or suppressed and that the robot's expression increases impressions of closeness and sensitivity. Last, we discussed the future perspective of human-robot communication by physiological phenomena. Regarding the representation of the robots' sense of life, it is thought that the user's recognition that the robot is alive improves not only the moral effect on the understanding of the finiteness of life but also the attachment to the robot in long-term communication. Regarding the emotional expression mechanism based on life, it is expected that the robot can display a complicated internal state close to that of humans by combining intentionally expressed emotions and involuntary emotional expressions. If a robot can express a combination of realistic voluntary expressions, such as facial expressions and body movements, in combination with real involuntary expressions by using the real intentions and lying, it can be said that the robot has a more complicated internal state than that of a pet. By using a robot expressing a living state through physiological phenomena, it can be expected that the effect of mental care will exceed that of animal therapy, and we expect to provide care and welfare support in place of human beings.
コレクション（個別）国立国会図書館デジタルコレクション > デジタル化資料 > 博士論文
受理日（W3CDTF）2019-05-06T10:27:56+09:00
連携機関・データベース国立国会図書館 : 国立国会図書館デジタルコレクション