Capturing Synthetic Knowledge through Simulations and explicit Knowledge in Suitable Data Structures

SPP 2422: Working group 1

Aims

The aim of the working group is to identify suitable categories and formats so that existing domain knowledge from practice or the laboratory on the one hand and synthetic process data from simulations on the other can be structured and collected.

Work Program

  • Research/collection of design, production and heuristic knowledge for the design of active surfaces of forming tools/effective systems for forming.
  • Methods for recording, structuring and processing numerically generated process knowledge (process simulations, analytical models, etc.)
  • Preparation, transformation and integration of the acquired/existing domain knowledge in/for machine learning processes
  • Capturing process noise in simulations
  • Use simulations for "non-measurables"
  • Explainability of simulation vs. experiment (matching of e.g. image data to simulations)
  • Interface to data science (style guide for synth. data)
  • Quality / goodness of synth. data
  • Weighting: Is a high proportion of synth. Is a high proportion of synth. data always expedient (e.g. for image data)?
  • Form-foundation model

Contact

Univ.-Prof. Dr.-Ing. Thomas Bergs

RWTH Aachen University

Laboratory for Machine Tools and Production Engineering (WZL)

Campus-Boulevard 30

52074 Aachen

E-mail: t.bergs@wzl.rwth-aachen.de

Website: https://www.wzl.rwth-aachen.de/cms/~sijq/WZL/?lidx=1

Prof. Dr.-Ing. Dipl.-Wirtsch.-Ing. Peter Groche

Technical University of Darmstadt

Institute for Production Engineering and Forming Machines

Otto-Berndt-Straße 2

64287 Darmstadt

E-mail: peter.groche@ptu.tu-darmstadt.de

Website: https://www.ptu.tu-darmstadt.de/institut_3/index.de.jsp

To the top of the page