Human-Inspired Generative Design

Dr Richard Ahlfeld of Monolith AI gave a presentation on 'Human-Inspired Generative Design' to the NAFEMS' Optimisation Working Group's (OWG) Special Interest Group. The presentation recording has been made available to NAFEMS members.

Generative design is a design exploration process where engineers can input their design goals into the generative design software, along with parameters such as performance, materials, manufacturing methods, and cost constraints and the software then explores all the possible permutations of a solution and suggests a list of optimal design.

The quality of solutions suggested by generative design software has improved dramatically over the last years for structural analysis, but not in fluid or aerodynamic analysis. In this talk, we will show how deep learning tools borrowed from image processing can allow engineers to build their own generative designs solutions that also work for fluid dynamics problems.