We are working on standardization for the social implementation of new robots, ELSI strategies, development of demonstration sites, construction of technology for the cooperative operation of multiple AI robots, and development of simulators for system integration. In addition, we are working on a framework for sharing the data set developed in R&D item 3 with R&D item 1, as well as on control and operation methods for the adaptable AI robots in R&D item 2, to ensure close collaboration in anticipation of demonstration experiments.
An organizational development framework for ELSI/RRI, which includes concept design, analysis, ethical compliance, and risk assessment of the robot to be developed, was established in accordance with IEEE SA 7000. In addition, to integrate people, robots, and AI, we developed a support tool for developing social implementation scenarios based on the Service Robot Design Matrix (SRDM) by identifying relationships among stakeholders and pain points extracted from them.
Furthermore, as an elemental technology for self-organizing socially implemented AI, we constructed a living laboratory for measuring daily life functions in which various sensors are placed. This living lab is a space where users can understand "what they can do" and "what they cannot do" and "what they want to be able to do" through activities in the lab. In other words, it is a space where they can estimate their cognitive/physical abilities and extract necessary support while having fun through activities in the living lab. It is an epoch-making space that can realize micro-analysis (movement and eye movement measurement, muscle activity estimation), meso-analysis (object manipulation and human behavior analysis), and macro-analysis (task execution evaluation).
Based on the analysis results in the living lab, we also conducted a cooperative planning simulation of AI robots and constructed a IoT based control framework for multiple robots/sensors using a common OS that enables the coordination of robots, sensors, interfaces, and other devices. We conducted demonstration experiments using this framework in the living lab and rehabilitation room at the National Center for Geriatrics and Gerontology.
In addition, to construct a privacy-conscious activity dataset for nursing homes, the actual activities of people are represented by a virtual environment and avatars. In addition, we are developing elemental technologies to measure biometric and physiological information such as eye gaze and heart rate, and to provide the data set together with synthetic voice and daily life sounds.