An essential step in all manufacturing processes is the metal surface coating technology, since in particular the obtained corrosion and wear protection has a significant influence on the durability of components. In addition, many other functionalities such as tribological properties, appearance, hardness, ductility, thermal load-bearing capacity or conductivity etc. can be precisely adjusted with surface coatings. Electrochemical or plating coating is one of the most effective and low-cost surface engineering processes; it has the largest market share of all coating processes worldwide.
However, unlike in mainly mechanical manufacturing processes, there is still no suitable and sufficiently cost-effective measurement technology in industrial surface technology, especially for the coating process itself; even critical process parameters can often only be monitored offline (corresponding process monitoring is known from semiconductor manufacturing, but the costs for this are in industrial electroplating technology are too high). The product quality of coated components can therefore only be determined when the end product is already available. At this point, it may no longer be possible to determine and trace which parameter variation(s) during the production process are responsible for the quality loss. This is particularly problematic in the case of safety-relevant components, e.g. fastening elements, especially if the defect, e.g. material embrittlement caused by hydrogen which results from the electroplating process, is only revealed during technical use.
In this project, an AI-based measuring system solution is being developed for industrial electroplating, which for the first time can cost-effectively and with sufficient accuracy provide all process parameters relevant for the digitalisation of electroplating production processes. The innovative measurement system is intended to link system-adapted, cost-effective and industry-suitable in-situ analysis of the process baths with AI-based evaluation of measurement data from process and plant control, status parameters of process units and other relevant data. Intelligent machine learning algorithms should enable individual adaptation of the measurement system to the respective coating processes.
The key point is thus the replacement of a large part of the otherwise required, very expensive chemical analytics by AI-based data evaluation. The planned system solution should be fully integrable as a cyber-physical production system (see figure) and have the necessary interfaces.