At the same time, more and more companies are developing assistance systems that support people in their work or make it easier to learn. The challenge is called: AI plug and play, because a prototype is all well and good, but integrating AI-based quality inspection into an existing line presents challenges for many people in positions of responsibility. At HANNOVER MESSE, you will find the answers to your questions about AI as an optimization approach or as an assistance system.

AI and skilled workers

There is a shortage of machine operators in industry. Production managers are looking for employees and/or increasingly relying on process automation or the support of intelligent assistance systems. Some machine builders already offer their customers pre-trained basic models, which the production manager plays on the controller, and support the machine operator. With greater experience, users continue to sharpen the model on the Edge, training it for their needs, learning and improving their processes. However, such a basic model can only work if customers are also willing to share trained models further.

AI experts talk about federated learning in this context. Thus, customers no longer give the data to the machine builder, but only the model trained with its data. The machine builder cannot view the data used. It collects the models from many customers and keeps refining its model. Pre-trained models make AI plug and play possible. Some automation companies such as Beckhoff are already working on this.