New research association FORinFPRO: 2 million euros for the development of AI-based and self-adapting manufacturing processes

How can we develop self-adaptive control systems for machines, plants, and process chains that are able to learn from past process steps in order to adapt to future process requirements? This is the central research question of the new Bavarian research consortium FORinFPRO. The solution aims not only to contribute to higher product quality, but also enhance robustness with increased resource efficiency using recycled materials in various processes. This will be investigated for the specific demands in manufacturing lightweight components. To produce complex components from composite materials, innovative manufacturing processes will be enhanced regarding sensing, process technology, modelling, and control to meet the challenges posed by future self-adaptive processes.

Researchers from the fields of materials and process engineering as well as AI and control technology from eight research institutions and seven companies are cooperating to achieve this. The technical infrastructure of the "AI Production Network Augsburg" and the vision and dynamism of the newly founded AI Chair at the TU Nuremberg will accelerate for the realisation of the FORinFPRO project.

The research objective is to develop a methodology that allows not only the local control of specific sub-processes, but realises  complex process chains with global control at the overall process level. The aim is to create robust and self-adaptive controlled processes that can actively compensate for material fluctuations on the input side. New interfaces between the manufacturing processes and their control systems will be created based on the analysis, sensory monitoring and modelling of the sub-processes. The project is also investigating how algorithms with AI can be used to continuously learn from process data. The result is a general closed-loop control concept for process chains that meets the requirements of the industrial project partners and will enable learning manufacturing processes of tomorrow.

Source: University of Augsburg, Chair of Control Engineering in Computer Engineering