Ml4Pro2 - Machine learning for production and its products
Through machine learning (ML), knowledge can be generated from data in order to generate added value at all levels of the company processes. Products such as mobile platforms, robots or vehicles use data to optimize their behavior.
However, production systems are also increasingly using it to react more flexibly to new market developments and customer needs and to produce products optimally using the available resources. The local use of ML procedures close to the origin of the data is particularly promising.
The aim of the project is to make ML available for intelligent products and production processes. For this purpose, the latest ML methods are to be integrated into products and production chains. It is also about raising the awareness of companies to use ML for agile business models. The main topics are hybrid learning processes, the integration of expert knowledge, the interpretability of data, learning on data streams and cognitive edge computing. The ML methods are considered cross-application based on three industrial use cases: condition monitoring, process optimization and improvement of product quality. Results and procedures are made available to other companies on an ML platform. This platform includes, for example, reference implementations, methods for data preprocessing and data visualization as well as application knowledge about typical processes when using the ML methods.