Data-Driven Engineering Design: Intelligent data-driven design for digital manufacturing

Thumbnail

Event details

Date 13.02.2019
Hour 15:1516:15
Speaker Bernhard Thomaszewski, Department of Computer Science and Operations Research, Université de Montréal, Canada
Location
Category Conferences - Seminars

The design complexity and flexibility enabled by Digital Manufacturing is disrupting established production processes. However, as manufacturing technology keeps evolving, the gap between "what we can produce" and "what we can design" is growing steadily. In this talk, I will describe an architecture for intelligent design software that leverages data and computation to augment human design capabilities. This approach relies on three key components:

  1. Tailored computational models that manage complexity by exploiting problem-specific physical and geometric insights,
  2. Data-driven abstractions that condense simulation results in reduced-dimensional representations for fast, optimization-driven design, and
  3. Intelligent design interfaces that enable users to interact with simulation and optimization data in order to explore complex design spaces in intuitive and efficient ways.
I will illustrate these building blocks on a set of challenging inverse design problems ranging from mechanical metamaterials to 3D-printed textiles and self-deploying surfaces.

Bio of speaker: Bernhard Thomaszewski is an Assistant Professor in the Department of Computer Science and Operations Research at the University of Montreal. Bernhard received his MSc (2005) and Ph.D. (2010) from the University of Tuebingen. He was a post-doctoral researcher at ETH Zurich, where he also held an adjunct lecturer position until 2016. Before joining the University of Montreal, Bernhard was a Research Scientist at Disney Research Zurich, where he was heading the group on Computational Design and Digital Fabrication.  Bernhard's research lies at the intersection of visual computing, computational mechanics, material science, and digital manufacturing. His work aims to leverage data and computation to create powerful design software for complex mechanical systems including compliant mechanisms, mechanical metamaterials, self-deploying surfaces, and personalized robots. 

Practical information

  • Expert
  • Free

Contact

  • Pedro Reis, Flexible Structures Laboratory (fleXLab), IGM-STI-EPFL

Tags

Data-Driven Engineering Design Mechanical Engineering Machine Learning

Share