Revolution in production: intelligent machines in use

What is the aim of this project and who is involved?

The aim is to use AI to optimise our production processes, increase quality and reduce costs. In this context, we are working together with IT and colleagues from the plant on the implementation of this innovative technology. This measure contributes directly to POLYTEC's overall strategy by focussing on data-supported work and thereby improving the efficiency and scalability of our processes.

What progress has already been made?

We have already achieved some successes. At our site, we have significantly improved the data quality in the material master and in the product hierarchies. In addition, component images have been stored in booking terminals in order to precisely identify sources of error directly on the component. The introduction of a viscosity measurement app enables the digital recording of paint test data. This means that data can be used efficiently both in the model and by the process technicians. Another exciting feature of the model is that it can predict what influence the weather may have on painting results or dust and dirt inclusions in the component.

How has cooperation improved?

The use of AI makes it much easier to systematically identify the causes of problems or faults (known as root cause analysis) in process engineering. AI also improves understanding of the painting process and collaboration between different departments, particularly in planning and painting. More efficient planning of component painting leads to a smoother process and fewer colour changes per round. Which in turn improves the overall painting process.

What impact does the project have on customer satisfaction and internal processes?

The digitalisation of measurement and testing processes as well as data expansion gives us greater clarity and comprehensibility in our processes. Process analyses are partially automated, which leads to faster and more efficient root cause analyses. The feedback from the AI not only improves our production planning; the measures implemented as a result also lead to better painting results and therefore greater customer satisfaction.

Which technologies were used and could be of interest for other projects?

This project is a pioneer in the application of machine learning for process optimisation at POLYTEC. Process mining in particular has proven to be extremely useful and can be used in many areas for data-driven process optimisation.

What obstacles did you have to overcome during the project?

The quality of our data was a key hurdle. Data is like a food source for an AI. Without accurate data, the AI could not make accurate predictions. However, through extensive data analysis, preparation and the implementation of additional safety logic, we were able to overcome this challenge.