Mitsubishi Electric to Demonstrate Machining System Linking Humanoid Robots and CNC Machine Tools at JIMTOF2026

The system builds processes based on natural-language instructions and assigns tasks to machines and robots, with practical use targeted for 2030. Mitsubishi Electric will demonstrate it at JIMTOF2026, held Oct. 26-31, 2026, at Tokyo Big Sight.

Mitsubishi Electric has presented a system that coordinates machine tools and humanoid robots, starting with natural-language instructions. If it can reduce the need for individual control designs and production-line modifications, manufacturers using existing equipment will have a clearer picture of the conditions for adoption.

The key point is that the system does not position the humanoid merely as an aid for transport or parts supply; it systematizes coordination with CNC machine tools. According to Mitsubishi Electric, an AI machining task planner builds processes based on natural-language instructions and assigns roles to machines and robots. This goes a step beyond an approach that assumes existing automation equipment and assigns only peripheral tasks to humanoid robots. However, practical use in 2030 is a target, and the system is currently at the demonstration stage. In our related article, Mitsubishi Electric showcases screw-tightening AI at CEATEC, Mitsubishi Electric demonstrated autonomous control of screw-tightening work. The latest announcement can be seen as an extension of that effort, moving from automating a single process to building processes and dividing work among equipment. Separately from the business restructuring outlined in Mitsubishi Electric and Hon Hai agree to collaborate, the company is also deepening its control technologies for manufacturing sites. By applying physical AI to its Maisart technology, Mitsubishi Electric may be seeking to handle machine and robot control under a common design approach. The industry implication is a shift away from competition over robot performance alone and toward how machine tools, control systems and task planning are integrated. In this system, natural language is the input, the AI machining task planner processes it, and CNC machine tools and humanoids coordinate as the output. If this works, the focus will be on how much it can reduce the development of machine-specific communications functions and control sequences that have traditionally been required. However, it is not yet clear how broadly tasks can be registered, how exceptions will be handled, how reliably problems will be resolved, or how much work can be covered by registered tasks. The JIMTOF2026 demonstration is expected to be the first public opportunity to assess the system’s maturity.

The system consists of three elements: an “AI machining task planner,” “CNC machine tools equipped with physical AI” and “general-purpose humanoid robots.” The planner analyzes natural-language instructions and generates a work plan by combining the necessary tasks from a list of tasks registered in advance. These include setup, cleaning, tidying, finished-product removal, transport, inspection and troubleshooting. Mitsubishi Electric said it implemented newly developed physical AI using its Maisart AI technology brand in the planner and on the CNC side. Because machines and robots exchange operating information in natural language as they carry out machining, the system can work with different robot models and is said to reduce the burden of individual system integration.

For Japanese machine-tool and control-equipment makers, competition may increasingly center not just on supplying robots but on the control layer linking CNC machine tools and humanoid robots. Mitsubishi Electric’s emphasis on Maisart offers an example of how Japan’s strengths in control and embedded technologies can be connected to automation at manufacturing sites.