Progress Report

Adaptable AI-enabled Robots to Create a Vibrant Society[3] Social implementation of a collective of coevolution AI robots

Progress until FY2024

1. Outline of the project

This research theme aims to develop key technologies for AI robots to become a ubiquitous, safe, and reliable part of daily life. Our primary focus is on creating a self-organizing social implementation AI that adapts to people and environments. This AI will autonomously determine which robot provides what support, when, where, and to whom, using advanced simulation technologies.
To ensure successful societal integration, we are also establishing demonstration environments (living labs) and crafting social implementation scenarios. Our efforts include designing development concepts with an eye toward standardization, building robust safety evaluation criteria and risk assessment methods, and addressing critical ELSI (Ethical, Legal, and Social Issues).
To accelerate real-world validation in nursing care, this theme will concentrate on Nimbus Sensing & Computing and fostering exciting social experiences, as shown in our integrated research vision.

Figure

2. Outcome so far

This research developed "Nimbus Computing," a social implementation AI simulator for the coordinated operation of our Nimbus Robots. This simulator enables us to determine which robot provides what support to whom, and with what parameters, establishing a framework for seamless collaboration among multiple robots, welfare devices, and caregivers. This significantly contributes to providing seamless services with adaptive AI robots.
We're also vigorously pushing for the social implementation of our rehabilitation systems and Nimbus Robot technologies. We launched the Moonshot Challenge House concept with partners, starting efforts to introduce robots and training systems into homes. Collaborating with a nationwide elder care company, we created a vision for future care facilities. These initiatives are accelerating discussions with care providers, local governments, and robot developers to realize care facilities incorporating Moonshot's achievements.
To speed up AI robot social integration, we're creating new development concepts, formulating application scenarios, setting up verification facilities, and exploring shared data management based on ELSI (Ethical, Legal, and Social Issues). We're also preparing for the international standardization of AI robot ethical evaluations. A significant achievement here is the IEEE Standards Association's approval to establish the AI robot ethical standardization (P7017) working group, initiating global discussions. This is crucial for enhancing AI robot social acceptance and ensuring safe, confident use.
Furthermore, as an exciting initiative involving AI robots, we are promoting Smarter Inclusive Dance, an activity designed for enjoyment by everyone, regardless of disability. This initiative is being advanced through interdisciplinary and international collaboration, and its innovativeness has been highly praised. It has received the Robot Award and has also been featured as a case study for comprehensive knowledge utilization by the Cabinet Office.

Photo

3. Future plans

We're driving the development and validation of an AI robot group collaborative operation simulator, designed to automatically determine roles within AI robot groups. This aims to revolutionize eldercare by enabling advanced coordination among care robots.
To further accelerate societal integration, we're also focused on establishing safety evaluation standards for the Robotic Nimbus and developing ethical standardization for AI robots used in healthcare. These combined efforts will expedite the adoption of AI robots in practical care settings.