Skip to Content
Uncategorized

Software Designs Products by Simulating Evolution

New CAD software takes input from designers, then “evolves” new designs on its own.
November 7, 2014

Software that can “evolve” novel component designs could help designers and engineers by automating part of the creative process.

3-D printed lunar lander
Brackets for a lunar lander designed by Autodesk for the company Moon Express using Dreamcatcher.

Autodesk developed the computer-aided design software, called Dreamcatcher, over the past seven years. The California-based company already makes 3-D software that’s widely used in architecture, engineering, animation, and other industries. But Dreamcatcher takes a novel approach known as “generative design.” Among several approaches, it uses algorithms that mimic the process of evolution to produce new designs after starting with a list of parameters chosen by the user.

Dreamcatcher shows how even creative industries could be changed by automation (see “How Technology Is Destroying Jobs”). “Designers worry it will take their job away, and [they] doubt that it can create beauty,” acknowledges Autodesk’s senior director of design research, Mark Davis.

Dreamcatcher’s simulated evolution process begins with the software offering a set of potential solutions to a problem posed by the user. The software then “breeds” a new generation of solutions by recombining the best of those designs, as chosen by the user, and adding some random variation. The software can repeat this process to produce thousands of designs (see “Unnatural Selection”).

When a product designer or engineer feeds design goals—along with parameters such as materials, performance criteria, and cost constraints—into Dreamcatcher, the results may look nothing like conventional designs. Asked to create a chair, for instance, Dreamcatcher may produce legs made of a confluence of interwoven scaffolding rather than the usual solid ones.

3-D fabricated chairs
Three chair designs produced by the Autodesk software.

“Generative design software has been mostly on the fringe for 30 years,” says Hod Lipson, an expert in 3-D fabrication at Cornell University. “Hollywood animators, gaming developers, and computer artists made beautiful, elaborate effects with it,” he says, adding that it is likely to become more common because it complements the flexibility offered by 3-D printing.

Designers at Autodesk used Dreamcatcher to design a unique carbon fiber swingarm—a component that provides rear suspension—for a super-fast electric motorcycle, the LS-218, made by Lightning Motorcycles of San Carlos, California.

Dreamcatcher was fed files of a normal swingarm that had been produced by earlier versions of CAD software. The designers also fed in the width of the tire with clearances, the forces that would come with the weight of a rider, the available materials, the forces from wind and gravity, and size limits. Dreamcatcher kicked out a design for a swingarm made of hollow intersecting triangles sheathed in carbon fiber.

“You can make things that are structurally stronger from an engineering perspective” using generative design, says John Maeda, a prominent designer and a partner at the venture capital firm Kleiner Perkins Caufield & Byers. “Generative design is an exuberant field of discovery with many good things coming out of it.”

Lightning Motorcycles expects the swing arm to improve the LS-218’s acceleration, performance on curves, and predictability. If it performs well in road tests and races, bikes with the carbon fiber swingarm will be shipped next year.

Dreamcatcher has so far been used to make just a few other functional components including a bracket for a spacecraft and a bike frame. But Autodesk hopes that industries from architecture to construction will adopt the generative software when it’s commercialized.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.