FMC3 Robotics Advances Two Factory Projects Using Autonomous Mobile Robots

FMC³ Robotics, a Bavarian company that builds AI brains for robots, says it is advancing two industrial proof-of-concept projects with autonomous mobile robots. One is at a consumer electronics manufacturing plant. The other is at a battery cell production site. 

Both aim at work that's repetitive, physically demanding and hard to automate flexibly with traditional methods. The electronics project focuses on mobile material handling, the kind of job where bins have to travel between stations. At the battery site, robots process battery electrode sheets, which are thin and delicate to handle. 

Why these jobs? Conventional automation does well when the product, the material and the layout stay put. Factories rarely do. FMC³ says changing products, variable materials and existing equipment make standard automation difficult to adapt. Its answer is an industrial embodied AI agent platform, pitched as one brain for any robot. 

The platform ties task capture, data preparation, simulation, training, validation, deployment and feedback from running operations into one traceable loop. A cloud-based, domain-specific world model works alongside a language model trained on industrial material, so a robot can plan long tasks from plain language instead of code. Complex jobs get chopped into reusable building blocks the company calls atomic skills. Picking up an unfamiliar part, for example, runs through four steps: perception, decision, execution and tactile-force feedback. 

FMC³ plans to show the approach at electronica 2026 in Munich, Nov. 10 to 13, in the Future Mobility Pavilion. The display will include the two proof-of-concepts, two robot platforms and its Doceo Data System. The company has said its wheeled robots are already working in industrial settings on intralogistics, material sorting and machine tending. 

Partnerships are part of the plan. FMC³ works with Vector Informatik, which brings embedded systems and data security expertise, PEM Motion, a production engineering specialist, and Arm. In September, FMC³, Vector and PEM Motion signed a memorandum of understanding to build a standardized data infrastructure for industrial embodied AI, and they intend to set up a joint venture in Germany. FMC³'s Dr. Joachim Fetzer has said industrial embodied AI is as much a data and systems challenge as a model one. 

The projects also mark a shift for the company, which says it is moving from core technology research toward proof-of-concept validation, commercial deployment and industrial-scale production. Factory trials are where that gets tested, because production lines don't forgive slow or unreliable robots. 

For now, these are early projects. The announcement doesn't name the plants involved or give results, so how well the robots perform against traditional automation remains an open question.