Skip to content. | Skip to navigation

Personal tools
You are here: Home Theses Celsius or Fahrenheit? Extracting Ontologies from OPC UA for a Plastic Processing Use-Case


Prof. Dr. S. Decker
RWTH Aachen
Informatik 5
Ahornstr. 55
D-52056 Aachen
Tel +49/241/8021501
Fax +49/241/8022321

How to find us

Annual Reports





Celsius or Fahrenheit? Extracting Ontologies from OPC UA for a Plastic Processing Use-Case

Thesis type
  • Bachelor
Student Max Kocher
Status Finished
Submitted in 2021

Analyzing a live data integration of an injection molding machine at the IKV, implementing an ontology extraction feature to the existing software (Java, .net or Python) and adding semantics to the process. Demonstrate the results in a demo scenario


Scientific Question

In the context of Industry 4.0, reliable communication is as important as data analysis. For example, the NASA probe Mars Climate Orbiter was lost in 1999 due to a simple unit error. In general, just numerical values (e.g. "22.4") make the analysis of collected data difficult.

In order to increase the informative value of data for data scientists, we research common vocabularies the cluster of excellence “Internet of Production” at RWTH Aachen University. Ontologies properly map semantic relationships, but must be expensively created and maintained. Standards already exist in the engineering sciences, but their full potential is not being exploited. 

The aim of this master thesis is to extend a data stream of real machine data with proper metadata (e.g. units) and to extract an ontology from existing communication standards like OPC UA. Finally, commonalities between production systems will be automatically identified and the information flows of different machines will be analyzed prototypically using Machine Learning.



Scientific Methodology

• Analyzing the live data integration (OPC UA) of an injection molding machine at the IKV

• Implementing an ontology extraction feature into an existing OPC UA Connector

• Construction of an ontology and verification on a real injection molding machine



Objective and Expected Results

• Extraction and analysis of information models from a CPPS in injection molding

• Conversion of an OPC UA information model into an ontology

• Upgrade of the current data integration with semantics, including evaluation

• Prototypical data analysis via Machine Learning on semantic data


The results of this work are an important step towards interoperability between production machines in the context of Industry 4.0 and include software for the conversion of OPC UA information models, as well as a concept and implementation of a comparison of different ontologies with subsequent evaluation. Data Scientists are supported in their work and the interoperability of machines is promoted.


For more information, see the following attachment:

Interoperable Machine Learning for the Internet of Production Using Semantics.pdf — PDF document, 148Kb

Related projects

Document Actions