Whether it's a facial or hip-joint implant specifically created for one patient or an efficient radiator constructed with minimal material, generative design (using a computer to generate thousands of possibilities given requirements defined by a human) can produce the best design in less time than even an army of human designers.

Companies use generative design capabilities to create not just good, or even great, designs, but the most optimized designs. Once design parameters are defined for a specific project, the software goes to work, using evolutionary algorithms to generate solutions by mimicking the way nature accepts or rejects designs. Incorporating the advantages of additive manufacturing and employing design targets like weight reduction and load requirements, the software creates lattice structures that use a precise amount of material, exactly where needed. Many of the resulting designs can only be fabricated using 3D printing.

Infinite seats

Using software that mirrors the design process of Nature, a generative-design generated-chair design can be of equal strength but dramatically reduced weight — requiring less material. The Living, part of Autodesk, has been working on research and development projects in the fields of architecture, art, industrial design, aerospace, computer science, engineering, manufacturing and synthetic biology to create new types of buildings, public installations, prototypes, and architectural environments.

Bio Computation and Next Generation Aerospace

Airbus used cloud computers to generate and evaluate thousands of possible structures, designed using synthetic biology and principles from architecture, for their concept plane that depicts travel in 2050.

Autodesk continues to offer desktop applications, packaged into collections but also offers cloud-based services that leverage the connectivity of the internet and the ability to apply more than one computer to a problem — particularly in the areas of analysis, simulation, and collaboration.

Imagining a City of Evolutionarily Optimized Buildings

Instead of the traditional design/analyze/update approach, what would happen if a designer simply provided constraints and allowed infinite computing resources in the cloud generate thousands of possibilities and evaluated them? Using http://www.autodesk.com/products/revit-family/overview - Revit Architecture and Project Vasari/Autodesk Green Building Studio (now http://insight360.autodesk.com - Insight 360 ) application programming interfaces (APIs), IDEA Studio residents developed a plug-in to automatically generate, analyze, visualize, and rank various design configurations.

Imagine you're an architect and your client asks you to design a building. He has two requirements:

• Minimize energy use.
• Maximize rentable floor space.

That sounds totally reasonable, right? So as an architect with lots of CAD prowess, you design a building in pursuit of those two goals. You run an analysis. The results show that your rentable floor space is too small. So you increase the size of the building. Now your projected energy costs are too high. So you iterate by making adjustments to your building design until both criteria are met.

But what if you could just give those two requirements to the computer and let it generate the right answer? What if instead of generating one correct answer, it could generate thousands, and let you pick one?

Perhaps you pick one based on aesthetics or how it might fit in with the other buidlings in the neighborhood.

Yes, generative design can do that. It shifts the job of architect from defining the solution to carefully defining the requirements so that the computer can generate the solution. As the architect, you're in charge of the problem space instead of the solution space.