Self-Learning Production Control using Algorithms of Artificial Language

The aim of iProd is the development and validation of a solution concept which, within the scope of Industrie 4.0, exploits the technical possibilities of digitization. The digital image of a production is the basis for the holistic analysis and evaluation of the automatically recorded data. On the basis of the examined patterns in the production system, exact predictions of the future system behavior can be made while taking account of influencing disturbing factors and accompanying deviations.

Manufacturing companies in Germany are facing an increasingly turbulent environment. Growing product complexity and shorter product life cycles require robust production processes. This includes coordination of production and assembly with regard to unplanned deviations.

The iProd research project therefore includes the development and validation of an Industrie 4.0-ready production control solution addressing the problem of deviations and disturbances in production. The main goal is the improvement of logistics performance.

The basis for this is a digital image of the real-world production environment, taking into account the system state. This is made possible through the pervasive use of information technology and sensors on the shopfloor resulting from a number of Industrie 4.0 initiatives: In the high-resolution data now available, patterns can be identified using artificial intelligence and reproducibly maintained with the help of machine learning techniques, similar to the functioning of the human brain.

Taking into account influencing disturbing factors, this allows exact predictions of future system behavior.

The data is collected via an online platform. Within the scope of a production control system, the deviations thus assessed are reduced or compensated by interventions. Through the use of artificial intelligence, this process shall be able to run in a highly automated manner.

The solution to be developed is tested and validated together with industrial and practice partners in application cases. The technology is targeted at both manufacturing companies as well as platform and IT system providers of order processing solutions for manufacturing companies.