Thank you for taking the time to read this.
I am a graduate student doing research using Fusion 360's Generative Design feature.
I have a few questions for my thesis.
I'm curious about the mathematical representation of Generative Design in Fusion 360.
Here is an example of a paper where a GAN model is represented mathematically.
Reference(OH, Sangeun, et al. Design automation by integrating generative adversarial networks and topology optimization. In: International design engineering technical conferences and computers and information in engineering conference. American Society of Mechanical Engineers, 2018. p. V02AT03A008.)
Hi @sap051342C394,
Thanks for your question! I'm not sure I understand exactly what you want to know, but I'll take a stab at it.
Generative design uses various optimization strategies based largely upon the topology optimization literature. The exact algorithms and approaches we use are proprietary, but in the topology optimization literature it is typical to discretize the optimization domain and represent the shape either explicitly (i.e., density) or implicitly (i.e., signed distance field). Additionally, we use AI in portions of our pipeline, but which ones and how they are used is proprietary.
You can find hints about which classes of optimization algorithms we use from various other thread postings (link, link, link).
All of that indirectly hints at the answer to your question. Unfortunately I'm not allowed to directly say what we use under the hood in generative design. Best of luck to you!
Ben
Can't find what you're looking for? Ask the community or share your knowledge.