Physical AI Beckhoff Controls Robots and Machines via Voice Command

Related Vendors

FAULHABER Drive Systems

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

SIMTOS (Seoul International Machine Tool Show)

Automation specialist Beckhoff connects artificial intelligence directly with machine control. At the Hannover Messe, the developers will be demonstrating how users will be able to control industrial systems simply by speaking.

At the Hannover Messe in April, the focus will be on physical AI. Automation technology manufacturer Beckhoff will be presenting in Hall 27 how large language models control real motion sequences in the factory via standardized interfaces. Instead of rigidly processing programmed codes line by line, machines will in future be able to recognize the context of their environment and react autonomously to complex requirements. The direct link between generative AI and classic, deterministic control technology is making precisely this change in the manufacturing industry possible.

The pre-show press conference featured a compact demonstration cell in which the Xplanar planar motor drive system worked together with the Twincat Coagent AI tool. Users entered simple voice commands via an audio interface. The hovering transport modules then set themselves in motion and independently initiated the required sequence. This application illustrates how man and machine will interact on the shop floor in the near future: In future, anyone who wants to operate or adjust a system will no longer need in-depth programming knowledge.

Industrial Robot Plays Chess Against Trade Fair Visitors

When the Hannover Messe opens its doors from April 20 to 24, 2026, Beckhoff will extend this principle to a fully integrated industrial application. The modular industrial robot Atro will take center stage at the trade fair stand. Visitors will be able to observe on site how the system learns and takes over tasks when the operator simply speaks to it.

The technical backbone of this communication is the Model Context Protocol (MCP), i.e. the interface between the LLM and the control software. This standard allows the control system to act as an intelligent agent: it translates human speech into specific machine commands, plans the paths of the robot arms in the room and also performs ongoing diagnostic tasks. To illustrate the concept for the trade fair visitors, the robot will compete in real chess games against the guests at the stand.

AI Leaves the Chat Window

Software tools such as the TwinCat CoAgent and the TwinCat Machine Learning Creator form the basis for this development. They support plant manufacturers throughout the entire life cycle of a machine, from automated code generation during the design phase in engineering through to error analysis when the plant is already in operation.

The managing founder of the company, Hans Beckhoff, attaches great historical importance to this technological development: "Technical inventions have always determined the development of society. Artificial intelligence and physical AI will take on this role with the same significance as the steam engine and electricity once did."(mc)