TR73

  • Increase font size
  • Default font size
  • Decrease font size


B1 - Self-learning Engineering Assistance System



Simultaneous Development of a Self-learning Engineering Assistance System

Project Status: finished

Last Update: 12.05.2021



Members


B1-pic-third phase

The processes of sheet metal forming, which are being researched within the framework of the Collaborative Research Centre Transregio 73 (SFB/TR73), offer new possibilities for adapting components to increased requirements due to their increased design freedom. The combination of solid and sheet metal forming offers enormous potential for product development, and suitable computer-aided tools must be made available to exploit this potential.


Subproject B1 is dedicated to this overall goal with the development of a self-learning assistance system (SLASSY). SLASSY enables the knowledge-based support of the design engineer in the development of sheet metal formed components with complex secondary form elements. The necessary design-relevant manufacturing knowledge is collected simultaneously with the manufacturing process development by using machine learning methods. It is then stored in the multidimensional knowledge base in the form of meta or prediction models. The architecture of SLASSY supports the synthesis of a component from main and secondary form elements as well as its knowledge-based analysis.
A major goal of the third phase is the use of spatially resolved meta models for the prediction of local component and tool properties. For example, to directly predict the degree of deformation on the component surface.

In addition, manipulated variables for the expansion of the design space under consideration of two or more target variables are investigated and the genetic multi-objective optimization algorithm developed in phase two is extended by corresponding prediction variables. In this way it is possible to provide the product developer with concrete information about the extent to which he has to adjust or modify his production parameters and machines in order to achieve the selected result.

Finally, the multi-dimensional knowledge base of SLASSY is supplemented by the corresponding building blocks to cope with local metamodels, new multi-objective optimization methods and the influence of the process sequence of sheet metal massive forming.


Working Groups


Publications

    2021

    • Sauer, C.; Breitsprecher, T.; Küstner, C.; Schleich, B.; Wartzack, S.: Architecure and self-learning within an assistance system for design for manufacturing within sheet-bulk metal forming . In: Kenji Suzuki (Edt.): 21(2021)3, Basel, Suisse: MDPI AI, submitted
    • Hinz, L.; Metzner, S.; Müller, P.; Schulte, R.; Besserer, H.; Wackenrohr, S.; Sauer, C.; Kästner, M.; Hausotte, T.; Hübner, S.; Nürnberger, F.; Schleich, B.; Behrens, B.; Wartzack, S.; Merklein, M.; Reithmeier, E.: Fringe Projection Profilometry in Production Metrology: A Multi-Scale Comparison in Sheet-Bulk Metal Forming. In: Steve Vanlanduit (Edt.): 21(2021)7, Basel, Suisse: MDPI Sensors, published
    • Sauer, C.; Schleich, B.; Wartzack, S.: Simultaneous Development of a Self-learning Engineering Assistance System. In: Merklein M., Tekkaya A.E., Behrens BA. (Edt.): Lecture Notes in Production Engineering, (2021), Cham: Springer, pp. 127-146

    2020

    • Küstner, C.: Assistenzsystem zur Unterstützung der datengetriebenen Produktentwicklung. In: Jörg Franke, Nico Hanenkamp, Marion Merklein, Michael Schmidt, Sandro Wartzack (Edt.): Dissertation, C. Küstner, 1(2020)353, Erlangen: FAU University Press, pp. 219
    • Bickel, S.; Sauer, C.; Schleich, B.; Wartzack, S.: Comparing CAD part models for geometrical similarity: A concept using machine learning algorithms. In: Procedia CIRP, (2020), in print
    • Sauer, C.; Schleich, B.; Wartzack, S.: Meta-model based generation of solution spaces in sheet-bulk metal forming Beitrag in einer Fachzeitschrift. In: Procedia CIRP, (2020)91, pp. 905-910

    2019

    • Sauer, C.; Schleich, B.; Wartzack, S.: Einsatz von Graphdatenbanken für das Produktdatenmanagement im Kontext von Industrie 4.0. In: Ralph H. Stelzer, Jens Krzywinski (Edt.): Entwickeln Entwerfen Erleben in Produktentwicklung und Design 2019, (2019), Dresden: Thelem Universitätsverlag & Buchhandlung GmbH & Co. KG, pp. 393-406

    2018

    • Sauer, C.; Schleich, B.; Wartzack, S.: Deep learning in sheet-bulk metal forming part design. In: DS92: Proceedings of the DESIGN 2018 15th International Design Conference, (2018), pp. 2999-3010
    • Sauer, C.; Kügler, P.; Kestel, P.; Graf, M.; Göbel, K.; Niessen, C.; Schleich, B.; Wartzack, S.: Ein ontologiebasierter Ansatz zur Wissensrepräsentation für die smarte Produktentwicklung. In: Rieg, F.; Brökel, K.; Scharr, G.; Grote K.; Müller, N.; Lohrengel, A.; Nagarajah, A.; Corves, B. (Edt.): Tagungsband 16. Gemeinsames Kolloquium Konstruktionstechnik, (2018), pp. 294-305
    • Sauer, C.; Dworschak, F.; Schleich, B.; Wartzack, S.: Einsatz von Process-Mining zur Erweiterung der semantischen Produktbeschreibung. In: Krause, D.; Paetzold, K.; Wartzack S. (Edt.): Design for X. Beiträge zum 29. DfX-Symposium, (2018), pp. 227-238
    • Breitsprecher, T.: Entwicklung eines selbstlernenden Assistenzsystems zur automatischen Akquisition von konstruktionsrelevantem Fertigungswissen. In: Dissertation, T. Breitsprecher, 1(2018)449, VDI Verlag, pp. 192

    2017

    • Wartzack, S.; Sauer, C.; Küstner, C.: What does Design for Production mean? – From Design Guidelines to Self-learning Engineering Workbenches. In: Meyer A, Schirmeyer R, Vajna S (Edt.): Proceedings of the 11th International Workshop on Integrated Design Engineering, (2017), Magdeburg: Universität Magdeburg Lehrstuhl für Maschinenbauinformatik, pp. 93-102
    • Sauer, C.; Küstner, C.; Wartzack, S.: Einsatz von Deep Learning zur ortsaufgelösten Beschreibung von Bauteileigenschaften. In: Krause, D.; Paetzold, K.; Wartzack, S. (Edt.): Design for X. Beiträge zum 28. DfX-Symposium, (2017), Hamburg: TuTech Verlag, pp. 49-60

    Presentations

      2020

      • 05.05.2020: Sauer, C.: Meta-model based generation of solution spaces in sheet-bulk metal forming, Skukuza Rest Camp

      2019

      • 28.06.2019: Sauer, C.: Einsatz von Graphdatenbanken für das Produktdatenmanagement im Kontext von Industrie 4.0, Dresden

      2018

      • 24.05.2018: Sauer, C.: Deep learning in sheet-bulk metal forming part design, Dubrovnik
      • 26.09.2018: Sauer, C.: Einsatz von Process-Mining zur Erweiterung der semantischen Produktbeschreibung, Tutzing
      • 11.10.2018: Sauer, C.: Ein ontologiebasierter Ansatz zur Wissensrepräsentation für die smarte Produktentwicklung, Bayreuth

      2017

      • 07.04.2017: Wartzack, S.: What does Design for Production mean? – From Design Guidelines to Self-learning Engineering Workbenches, Magdeburg
      • 04.10.2017: Sauer, C.: Einsatz von Deep Learning zur ortsaufgelösten Beschreibung von Bauteileigenschaften, Bamberg