Exploiting generative models for performance predictions of 3D car designs

Sneha Saha, Thiago Rios, Leandro Minku, Bas Vas Stein, Patricia Wollstadt, Xin Yao, Thomas Back, Bernhard Sendhoff, Stefan Menzel

Research output: Chapter in Book/Report/Conference proceedingConference contribution

163 Downloads (Pure)


In automotive digital development, engineers utilize multiple virtual prototyping tools to design and assess the performance of 3D shapes. However, accurate performance simulations are computationally expensive and time-consuming, which may be prohibitive for design optimization tasks. To address this challenge, we envision a 3D design assistance system for design exploration with performance assessment in the automotive domain. Recent advances in deep learning methods for learning geometric data are a promising step towards realizing such systems. Deep learning-based (variational) autoencoder models have been used for learning and compressing 3D data allowing engineers to generate low-dimensional representations of 3D designs. Finding representations in a data-driven fashion results in representations that are agnostic to downstream tasks performed on these representations and are believed to capture relevant design features. In this paper, we evaluate whether such data-driven representations contain relevant information about the input data and whether representations are meaningful in performance prediction tasks for the input data. We use machine learning-based surrogate models to predict the performances of car shapes based on the low-dimensional representation learned by 3D point cloud (variational) autoencoders. Furthermore, we exploit the stochastic nature of the representation learned by variational autoencoders to augment the training data for our surrogate models, since the limited amount of data is usually a challenge for surrogate modeling in engineering. We demonstrate that augmenting training with generated shapes improves prediction accuracy. In sum, we find that geometric deep learning approaches offer powerful tools to support the engineering design process.
Original languageEnglish
Title of host publication2021 IEEE Symposium Series on Computational Intelligence (SSCI)
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Number of pages9
ISBN (Electronic)9781728190488
ISBN (Print)9781728190495 (PoD)
Publication statusPublished - 24 Jan 2022
EventIEEE Symposium Series on Computational Intelligence (IEEE SSCI 2021) - Orlando, United States
Duration: 5 Dec 20217 Dec 2021

Publication series

NameIEEE Symposium Series on Computational Intelligence
ISSN (Electronic)2770-0097


ConferenceIEEE Symposium Series on Computational Intelligence (IEEE SSCI 2021)
Abbreviated titleIEEE SSCI 2021
Country/TerritoryUnited States

Bibliographical note

Funding Information:
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement number 766186 (ECOLE).

Publisher Copyright:
© 2021 IEEE.


  • representation learning
  • 3D point clouds
  • auto encoder
  • variational autoencoder
  • regression analysis
  • Auto encoder
  • Representation learning
  • Variational autoencoder
  • Regression analysis

ASJC Scopus subject areas

  • Artificial Intelligence
  • Decision Sciences (miscellaneous)
  • Control and Optimization
  • Safety, Risk, Reliability and Quality
  • Computer Science Applications


Dive into the research topics of 'Exploiting generative models for performance predictions of 3D car designs'. Together they form a unique fingerprint.

Cite this