Fraunhofer-Institut für Produktionstechnologie IPT

In the context of production processes, many process and status variables are acquired via different
sensors. Due to varying data quality, data preprocessing (DPP) is the most time-consuming phase and an
essential step for the successful use of machine learning (ML). Based on an existing benchmarking
consisting of hundreds of production use cases, the influence of data leakage regarding ML is to be
quantified, interactions between individual DPP methods are to be proven and the benchmarking is to be
optimized as well as validated based on real production use cases.

Your tasks

  • Familiarization with the state of the art on DPP methods
  • Selection and implementation of ML algorithms
  • Benchmarking of data preprocessing pipelines and analysis of the influence of data preprocessing on the result of ML algorithms
  • Evaluation and documentation of the results

What we expect from you

  • You are studying mechanical engineering, computer science, industrial engineering, or a comparable subject
  • Initial experience in the field of machine learning is an advantage
  • A high degree of initiative and team spirit
  • Very good language skills in German and English

What you can expect from us

  • Pioneering role in benchmarking methods of data preprocessing in production.
  • An excellent equipment of machines and devices
  • Collaboration in an industry-oriented research project and in a dedicated team

Notice: You can write your project work or thesis with us immediately and in direct cooperation with
the Chair of Metrology and Quality Management. According to the guidelines of the RWTH Aachen
University, the internship is preferably to be completed at a manufacturing company and can only take
place with us in exceptional cases.

We look forward to receiving your application
Maik Frye
Fraunhofer Institute for Production Technology IPT
Steinbachstraße 17, 52074 Aachen

Um sich für diesen Job zu bewerben, sende deine Unterlagen per E-Mail an