We asked Waseda University professor Ogata Tetsuya (尾形 哲也) whether Japan can really win with “Physical AI.” The interview explores how Physical AI should be defined, as well as Japan’s current position and its prospects going forward.
Japan once led the world in robotics, but the United States and China are now ahead in the development race for Physical AI. Ogata Tetsuya (尾形 哲也), chair of AIRoA and a professor at Waseda University, discusses what Japan needs to make a comeback.
Ogata defines Physical AI as “operating robots using generative AI methods.” He explains that end-to-end learning, moving beyond the framework of machine learning reliant on conventional dictionary- or rule-based programs, could enable full-body movements and highly challenging actions.
He expressed the view that applying training based on the large volumes of data used by generative AI to robot operation in the physical world is expected to accelerate future development and expand the range of new applications.