Brief description:

This best practice story dealt with the optimization of a precision process, the plastic welding of a welding system of the company PolyMerge, as automatically as possible or at least prompted by remote engineering.

In addition to the definition of necessary sensor technology, one aspect of this was the collection of data from the production processes and the design of secure data communication. The industrial edge gateway ARENDAR was used to collect and pre-process the data and was further developed by adapting interfaces as required. The collected machine data was then provided either by the ARENDAR itself or by the CONTACT Elements software. Here, initial visualizations were possible. Appropriately prepared data could then be further processed for the purpose of data analysis. Together with the analysis software RapidMiner Studio, data connection options were developed that enable predictive analyses. For the analyses, the calculation of desired parameters based on the collected data from production, the evaluation of the resulting weld seam and a systematic analysis of the data to identify conclusions about the cause of the defect as well as possibly the initiation of a warning system to avoid rejects were discussed and conceptualized. The integration should thereby favor the creation of analysis processes that draw on real-time data as well as historical data.

Results:

  • Merging of the acquisition and management concepts of machine data from the best practice story
  • Linking the data collection and management concepts to visualization and evaluation tools
  • Development of a holistic retrofitting approach from the retrofitting of an edge device such as the ARENDAR, to the setup of data provision and visualization using CONTACT Elements, to the integration of predictive analysis results using RapidMiner Studio

Involved consortium partners:

Arendar IT Security GmbH filled the role of success story owner. There was also a close exchange with the related success story 4, which was led by Polymerge GmbH. Polymerge presented the use case in which a plant with a complex precision process was digitized. It was expected that the evaluation of the data would improve the quality of the production process.
Arendar was responsible for the acquisition of the data in the plants in the test laboratory and at the pilot customer as well as the preprocessing and provision of the data for evaluation.

Contact Software GmbH provided CONTACT Elements, its integrated PLM & IoT solution, which allowed access to usage data as well as development data of a system in one platform. The data managed here could be made available both to the user and to other systems, such as systems for data analysis. Trigger and workflow functionalities also made it possible to trigger business processes or initiate specific measures, such as service, maintenance or engineering change processes.

The Institute of Production Systems (IPS) contributed its expertise in the further processing and analysis of the data and was significantly integrated in linking the methods of data analysis with the internal domain knowledge in the area of production systems, i.e. which data were required for which considerations and how they had to be analyzed in a meaningful way.

With the ice.NET platform, PDTec AG provided its many years of expertise to the project, especially in the areas of data and information management, data interfaces, and data networking.

RapidMiner GmbH offers an open and extensible data analysis software that enables companies to benefit from artificial intelligence. Built for analytics teams, RapidMiner Studio unifies all steps of the data science lifecycle from data preparation to the use of machine learning to the application of predictive models. Over 540,000 analysts are already using RapidMiner Studio to incorporate knowledge from data analytics into decision-making processes.