In this study, we develop and evaluate an asymmetrical intention arbitration technique that enables an operator using a CA to arbitrate an operation intention while maintaining the operator's independence without feeling uncomfortable about the fusion with the operation of a companion operation agent when an operator operates the CA and a companion operation agent (human and automatic program) enters the CA. In particular, by substituting the gaze and gesture of a person who is difficult to achieve a specific skill, such as a visually impaired person or a hearing impaired person, an accompanying operation agent can help an individual participate in a society with diversity.
These manipulations need to be done in real time, and intervention in non-verbal manipulations may be the primary focus.
In addition, in the operation of a CA in which the task is difficult to be limited, a companion operation agent is required to respond in recognition of the operator's operation intention, so it is necessary for a human to voluntarily intervene while automating the reflexive action by a program.

What is required on the companion operation agent side is an interface with a low cognitive load, and an interface provided with a certain amount of physical input and output for performing physical work at the same time is required. In this study, we implement and evaluate them.




  • 大澤 博隆 リーダー
  • 岨野 太一 研究員
  • 峯岸 朋弥 博士課程1年