Hitachi and FANUC to Launch Physical AI Joint Business in FY2027

The Nikkei reported on September 30, 2026, that Hitachi, Ltd. and FANUC signed a strategic partnership to establish a joint business aimed at implementing physical AI, with the venture set to begin in FY2027. The report focused on the companies’ collaboration on physical AI for manufacturing.

Hitachi, Ltd. and FANUC each bring implementation capabilities for manufacturing and expertise in robot control, respectively. Their move to establish a joint business represents an effort to take physical AI from the experimental stage to deployment. If the venture proceeds as announced from FY2027, a key question will be how far the companies can integrate shop-floor data, robot control and operations design.

The focus of this development is that Hitachi, Ltd. and FANUC are advancing physical AI for manufacturing beyond one-off demonstrations and into a joint business. The venture is scheduled to begin in FY2027. At least as of the announcement, the companies have set up a framework premised on implementation and commercialization, moving beyond research and development. However, the excerpt from the primary source does not specify the processes targeted, the division of responsibilities, the functions to be offered or the scale of investment. At this point, the available information is limited to the joint business and its start date. In our related coverage, FANUC showcased a demo of an AI welding agent, pointing to an approach that reads drawings and uses them to inform work decisions. Reports on FANUC and Google’s AI welding robot and FANUC and Google’s AI welding robot, as reported by The Nikkei have also repeatedly covered the link between drawing recognition and task automation. The new joint business appears to extend that work by connecting robot-side control and control data with the implementation of AI for use on the shop floor. Hitachi’s efforts have also been reported, including its development of the “brain” for AI robots for manufacturing lines with a German startup. The companies had already signaled work on core functions for AI robots operating on manufacturing lines. This latest partnership can be seen as a move to bring those individual projects together in a joint business covering shop-floor data, control and deployment processes. However, the excerpt alone does not establish which processes the two companies will start with or what scope they will jointly cover. From an industry perspective, physical AI requires more than a learning model: it must connect robot control, shop-floor data, actual tasks such as welding and material handling, and operations after deployment. Hitachi brings implementation capabilities for manufacturing sites, while FANUC has expertise in industrial robot control and equipment. Their combination could link AI inference with equipment-side control. The practical questions, however, are which tasks the venture will target, what its revenue model will be, and how it will collect shop-floor data and use it for retraining. The next points to watch are the target processes, product launch timing, industries where the technology will be deployed, the terms for collecting and using data, and the division of responsibilities between the companies. The excerpt from the primary source does not provide this information, so the original report must be consulted for details.

In Japan, collaboration between Hitachi, Ltd. and FANUC, which support automation on manufacturing floors, could strengthen the connection between equipment-side control technology and on-site implementation. If AI adoption advances in areas such as welding and manufacturing lines, the effects could extend to companies involved in industrial robots, control equipment and the handling of shop-floor data.