Project Dreamcatcher allows designers to specify functional requirements like material type, manufacturing method, performance criteria, and cost restrictions, and based on these design requirements, Dreamcatcher searches a procedurally synthesized design space to evaluate a vast number of generated designs for satisfying the design requirements. The resulting design alternatives are presented back to the designer along with the performance data of each solution. The designer can then select the alternative based on the performance data or other factors such as aesthetics. In other words, instead of modeling a design by hand, analyzing it, failing, and iteratively updating the design by hand until it passes the analysis, Project Dreamcatcher generates designs and shows them to the designer who can pick one. Autodesk is working to make the computer a true partner in the design process — expanding its role from sophisticated documentation tool that also provides analysis, simulation, and collaboration capabilities.

Recently our CTO, Jeff Kowlaksi, described generative design in this video:

Dream design is alive in the lab.