Mitsubishi Electric to showcase physical AI at CEATEC 2026, alongside lights-out factories and next-generation data centers

Mitsubishi Electric announced that it will exhibit as the Mitsubishi Electric Group at CEATEC 2026, which will be held at Makuhari Messe from Oct. 13 to 16, 2026. The company will be in the General Exhibits area in Hall 8, at booth 8H412. The exhibition will center on “circular digital engineering,” which combines on-site expertise with AI, and physical AI that understands the physical world and operates safely.

Mitsubishi Electric is presenting physical AI not simply as a showcase of robots, but as an operational model that turns knowledge from manufacturing sites into data and feeds it back into control. Based on the announcement, companies that need to address factory automation, OT security and data center operations will need to consider not only equipment performance but also data infrastructure and operational design.

What is new in this exhibition is that Mitsubishi Electric has not made physical AI a standalone theme. Instead, it has brought together autonomous screw-fastening control, shelf placement by humanoid robots, OT security, lights-out factories and data center solutions as part of a single flow. In other words, the company is showing on the same exhibition floor not only demonstrations of perception, decision-making and action, but also the computing infrastructure, control, visualization and operations that support them. The focus is not on robots alone, but on a system for turning on-site expertise into data and reusing it. Our related article, Mitsubishi Electric to exhibit at CEATEC 2026, covered the advance announcement. The latest announcement provides more detail about the exhibits. While the earlier announcement focused on the plan to exhibit, the latest lists screw-fastening, humanoid robots, PUE 1.1 and a group of AI agents, clarifying what the company intends to showcase. However, readers should consult the original source for details; publicly available information does not yet reveal the implementation level of each demonstration or the scope of its application in mass production. From an industrial perspective, this exhibition is significant because it connects factory automation not only to robot adoption, but also to learning work know-how, visualizing on-site data, OT security and energy-efficient data center operations. The approach assumes a cycle in which data collected at manufacturing sites is fed back into control and simulation, placing greater emphasis on operational design than on one-off AI demonstrations. Still, it will be necessary to confirm which processes will have what level of autonomy, what size of facility the PUE 1.1 solution is designed for, and whether CVC collaborations have reached the implementation stage.

For Japan’s manufacturing sector, it is significant that assembly tasks such as screw-fastening, identification of sounds at worksites, OT security and data center operations with a PUE of 1.1 are presented in the same context. A key issue will be how to incorporate the digitization and visualization of on-site expertise into upgrades of existing factories, creating potential areas of connection for domestic companies with control equipment, sensors and factory software.