Generative design may make most designers unemployed

generative design

Imagine a piece of software that can design things for you according to your instructions. 

So, say, you want to design a chair. You could tell the software that you want the chair to be made of wood and be able to withstand a certain amount of weight, and maybe set a few other parameters.

And then the software gets to work and the next thing you know, it’s conjured up a chair design based on your instructions. And not just one iteration of the design, but maybe a few dozen.

That is more or less exactly what Autodesk’s new version of Netfabb – being prepared for launch next year – will be able to do. In fact, this process – where software designs objects – is being increasingly integrated into industrial design applications.

The technique has been around for a decade or so, and is referred to as “generative design”, but it’s only now that the computing power and artificial intelligence systems are available to do it properly.

Even 2D image- and video-editing software is quite memory-hungry, and anything dealing with 3D images even more so.

Moreover, industrial design is mostly done by scientists and engineers and the applications they use are loaded with real-world physics, so everything they design conforms to the laws of physics, right down to the molecular level – and that requires a lot of data processing.

Stephen Chadwick, managing director of Dassault Systèmes for the north European region, or EuroNorth, says some might think generative design makes it possible to remove humans from the design process completely.

“In terms of generative design, the ability for computing power to create the perfect design you could question,” says Chadwick in answer to a question from EM360º.

“You may hear that it’s possible to remove the human element from the design process, and the example I would give here would be – and it’s something we’ve been using extensively – is the design of brackets which are used to hold aeroplane engines in place.

“You could accurately model the stresses and strains the physical components at the molecular level of a particular manufacturing material.

“You could then choose the right manufacturing material based on your observations. And that material would then be organically designed automatically in generative design to enable to be the optimum weight, lightest possible, and strongest shape possible to function and do the job you originally wanted it to.

“You could say that that’s removed the human completely. But actually what would really happen is that the human is now involved in much more intellectual tasks and artistic elements of the design.

“So if we were to take a car or a shelf, or whatever object or product, the human is still needed to make it an object of art, or at least artistic, whereas the computer will just design it to be functional.

“The human will create it to be appealing to other humans, who are the ones who spend their money based on decisions which often relate to their artistic tastes.

“They might accept that under the bonnet, so to speak, the design meets certain requirements and standards, but the human experience – visual and other senses – is what is often the ultimate decisive factor when purchasing something.”

Dassault Systèmes’ 3D design software platform and suite is one of the world’s most widely used, and Chadwick claims that almost every aeroplane and road-going vehicle began some aspect of its life in the company’s applications, the most well-known of which is perhaps Catia.

The other big software company in this space is Autodesk, which, as mentioned above, is releasing Netfabb next year complete with the generative design feature included.

Netfabb is currently available in beta, and is being used by Autodesk partners such as Stanley Black & Decker, the maker of power tools and other products.

Greg Fallon, Autodesk vice president of simulation, explains Netfabb’s generative design features.

“Generative design technology takes goals set by a designer or engineer,” says Fallon. “For example, size, weight, strength, style, materials, cost, and any number of other criteria, and then uses cloud computing to create a massive number of design solutions.

“Using intelligent algorithms based on machine learning and advanced simulation, it produces smart design solutions that can be difficult for today’s designers and engineers to discover and model efficiently.

“The designer or engineer then identifies and adapts the right solution as desired. This process leads to major reductions in cost, development time, material consumption and product weight and gives our manufacturing customers the ability to design and engineer in brand new ways.

“Autodesk Generative Design is not just topology or lattice optimization alone – it’s a massive step beyond that.

“While optimization focuses on refining a known solution without any notion of manufacturability, generative design helps the engineer explore a whole cadre of functional and manufacturing design options.

“With Autodesk Generative Design, a designer or engineer can not only discover a new solution, they can then bring it to life using additive manufacturing tools.”