Up until recently, most of the things we bought were designed by humans.
Cars, computers, cans, chairs and cathedrals.
Everything could be traced back to an individual or team who set out to design products with the ideal the combination of form and function.
However, recently software has begun to find new ways to design and produce objects which would previously have been impossible for a human mind to develop.
Welcome to the world of Generative Design.
As shown in the video here on this article, designers can now tell special software what parameters and outcomes a design should deliver, and the software can then run millions (or billions) of iterative simulations to find designs which meet those specifications.
These systems are extremely effective at finding unique designs which have combination of strength with reduced weight, by only adding material where it is necessary.
Oftentimes, these designs bear striking similarities to designs produced in nature through thousands of iterations of evolution through natural selection, such as designs looking like animal skeletons.
For example, see this metal component for a motorcycle, designed by the generative systems from Autodesk software:
One previous drawback of such complex designs was that they could be nearly impossible to manufacture with traditional forming techniques. However, with recent innovations in additive manufacturing and 3D printing, it is now possible to produce these new designs in a matter of hours.
With future versions of generative design software surely being built on machine learning and other Artificial Neural Networks, it might even become possible for Artificial Intelligence software to solve design challenges in ways which a human would never have approached the problem.
Software has already shown its ability to be creative, and this is only likely to accelerate.
So the next time you look at a unique piece of architecture, an airplane or ambulance, it may have in fact been a computer which designed it.
Latest posts by Nick Skillicorn (see all)
- Balancing your innovation portfolio: Does the 70-20-10 rule still apply? - November 28, 2022
- Why Facebook needed to let go of 13% of their workforce - November 21, 2022
- If you want someone to make a decision, give them fewer choices - November 14, 2022
- Nine pregnant women cannot deliver a baby in one month - November 10, 2022