The joining of artificial intelligence assistants to the factory hall

The basic device To grind a steel ball was the same since about 1900, but the manufacturers were steadily automated around it. Today, this process is driven by a vector belt, and most of the time, it is automatic. The most urgent task for humans is to know when things are going well – and so can be delivered soon Amnesty International.

The Schaeffler factory in Hamburg begins with steel wire that is cut and pressed in rough balls. These balls are harden in a series of ovens, then three accurate mills are placed until they are spherical in the tenth of micron. The result is one of the most diverse components in the modern industry, which allows the decreased spirit joints in everything from the lattice to car engines.

This level of accuracy requires a continuous test – but when defects appear, their tracking can make a mystery. The test may show a defect that occurs at a stage on the assembly line, but it may not be clear. Perhaps the torque on a tightening tool may be, or the newly replaced grinding wheel affects quality. Tracking the problem means comparing data across multiple parts of industrial equipment, none of which is designed with this.

This may also be a function of machines. Last year, Schaeffler became one of the first user to Microsoft Factory Undersecretary, a new product backed by large language models designed specifically for manufacturers. The tool, similar to Chatbot, can help track the causes of defects, stop working, or excess energy consumption. The result is something like ChatGPT for factories, with the use of Openai models in the back interface thanks to the company’s partnership with azure of Microsoft.

Kathleen Mitford, Vice President of Microsoft Marketing Company in the Global Industry, describes the project as a “thinking agent who works at the top of manufacturing data.” As a result, Metford says, “The agent is able to understand the questions and translate them accurately and accurately in exchange for the standard data models.” So the factory worker may ask a question like “What causes a higher level of defects?” The model will be able to answer using data from all over the manufacturing process.

The agent was deeply combined into the products of the current institutions of Microsoft, especially Microsoft Fabric, and his data analysis system. This means that Schaeffler, which runs hundreds of plants on Microsoft, is able to train its agent on data from all over the world.

Stefan Sochik, Vice President of Chefler, says that the scope of data analysis is the real power of the system. “The main benefit is not Chatbot itself, although it helps,” he says. “It is a mixture of OT Data Data platform in the back interface, and Chatbot depends on that data.”

Despite the name, this is not Agency AI: it does not contain goals, and its powers are limited to answering any questions posed by the user. You can prepare the agent to implement basic orders through Microsoft Studio from Microsoft, but the goal is to have the agent to make his own decisions. This is primarily AI as a tool to access data.