(Bloomberg Businessweek) -- Project Discover, from Autodesk Research design studio the Living, uses a computer to create thousands of building designs based on a project’s goals and constraints. The computer then simulates the designs’ long-term durability to narrow down the best options.
The Benefit
This so-called generative design process can quickly determine the best trade-off between weight and structural strength or reconfigure buildings based on workers’ preferences for seating and other elements.
Innovator
● David Benjamin● Age: 45● Head of the Living group at Autodesk Research in New York
Background
Benjamin received a master’s in architecture from Columbia before founding the Living as a design firm specializing in the latest software. Autodesk acquired the company in 2014, while the Living was refining a cabin partition design for Airbus SE planes.
Incentives
Artificial intelligence design assistance can automatically adapt previous designs for new projects, saving hours of drafting work.
Augmentation
“In the best-case scenario,” Benjamin says, generative design enables “a kind of co-design between human and computer that could not be possible by human alone or computer alone.”
Office Fab
As part of Project Discover, the Living designed a three-floor, 300-person Autodesk office space in Toronto according to worker preferences. It will open this year. Autodesk is considering the approach for other offices, and the group has started work on an apartment building in Vietnam.
The Verdict
Will generative design replace human architects and designers?
“It won’t. It can’t,” says Robert Woodbury, professor of architecture at Canada’s Simon Fraser University. “However, at a low level of design—for example, simple house plans—generative design may well have a disruptive effect through its ability to generate many thousands or millions of suitably functional designs.”
To contact the author of this story: Michael Belfiore in New York at michael@michaelbelfiore.com.
To contact the editor responsible for this story: Eric Gelman at egelman3@bloomberg.net, Jeff Muskus
©2017 Bloomberg L.P.
