Autodesk Is Set to Jumpstart its Big Vision for Generative Design

Autodesk has announced that it has released a beta of its Generative Design tool inside Netfabb 2018 and that a full roll out of the technology will occur later this year. Although details about Netfabb’s Generative Design components are still a bit vague, Autodesk is presenting the design world with what it believes is a completely new paradigm for engineering. Are they right? They’re certainly talking a big game.

Let’s see what they’re up to.

Over the past few years, I’ve been visiting Autodesk shows where presenters have teased a generative design tool called Project Dreamcatcher. For those who haven’t heard of it yet, Generative Design (also known as topology optimization) is a wundertool that, according to Greg Fallon, Autodesk’s VP of Simulation, “takes goals set by a designer or engineer, e.g. 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.”

To that end, Generative Design can be seen as the next level of CAD, where an engineer’s ingenuity is paired with the power of parallel computing and simulation algorithms to create products that are optimized and, frankly, may have been unimaginable in the past.

Jeff Kowalski, Autodesk’s CTO, actually characterizes Generative Design in more comically hyperbolic terms, saying that it is the fulfillment of the promise implicit in the phrase “computer aided design.” Whereas, CAD systems have been tools that allow designers to record the geometry, and some of the function, that their designs will have, the computer hasn’t actually been involved in the design process.

With Generative Design, however, that paradigm has shifted. Now, computers and software can be relied upon to help craft an engineer’s vision and make it better than it may have been otherwise. Kowalski likens it to Tony Stark’s J.A.R.V.I.S. system -- even though Dreamcatcher (Autodesk's topology optimization software) isn’t a talking, strong-A.I. assistant. Still, “Just as J.A.R.V.I.S. augments Tony Stark’s engineering genius, Autodesk Generative Design gives our customers superpowers,” Kowalski said.

Although I’ve described Autodesk Generative Design as an optimization tool, it’s actually much more powerful than that. The software gives engineers the ability to define a project’s constraints (weight, material, heat dissipation, etc.) and create a litany of design solutions that can then be reviewed and further refined.

Take, for example, how Black and Decker’s Breakthrough Innovation group used topology optimization. Although the team was only formed recently, they’ve already been tasked with transforming the way that electric line workers do their job, day in and day out.

One of the most common tools used by an electric line worker is a hydraulic crimper. Every day a line worker will lift a 15.4-pound tool above his or her head to secure a wire and the grid itself. As you can imagine, this type of repeated activity can wreak havoc on your joints and cause a tremendous amount of discomfort. To lessen the degree of work-related stress involved, Black and Decker’s team wanted to lessen the weight of their crimping tool while maintaining its strength.

The answer for the Breakthrough Innovation team lay with topology optimization tools. Working with Autodesk’s consulting team, the Black and Decker group defined all of the criteria that would drive the optimization of its design. After giving Autodesk this foundation, the company’s algorithms poured through simulations creating numerous design concepts that Black and Decker engineers could consider.

In the end, Autodesk’s software created a design that had many of the quirky hallmarks of a topology optimization experiment (oddly shaped ribs, a screen-like lattice, bumps and dimples in odd places), but it also produced a tool that was 3 lbs lighter—no small feat. What’s more, since the generative tools being used on the Black and Decker project were running with Netfabb, Autodesk’s additive manufacturing suite, the Breakthrough team could quickly send its part off to be manufactured.

“The generative design capabilities we can access with Netfabb are almost magical.” Frank DeSantis, Vice President of the Breakthrough Innovation group, said. “It’s not brute force engineering. It’s elegant. You define a problem and you get a solution set unlike anything you’d predict. The results of the wire crimper project ensure we’re going to be applying the incredible combination of generative design and additive manufacturing that Netfabb offers to an array of other products we have in development.”

What Does Generative Design Mean for the Future of Engineering?

Generative design will clearly have a big impact in product innovation in the future. As design time tables become more compressed, engineers will have to leverage the power of the cloud to create design options that meet expanding demands.

In addition to speeding up innovation and transforming its impact on product revolutions, advanced algorithms could play a big role in accelerating additive manufacturing innovation. In fact, the two technologies might form a positive feedback system where both gain from the innovation taking place in the field of the other.

How so?

Well, it should be immediately obvious that generative designs lack a degree of manufacturability when it comes to many traditional industrial production methods. Sure, eccentric geometries might be easier to injection mold, but milling the fine skeletal components that lend a product its strength and innovative identity can’t be machined very easily. For components that have finer critical structures additive manufacturing might be the only effective manufacturing method.

As optimized designs become even more intricate and complex additive manufacturing systems will have to be built to meet the delicate post-processing needs of these finer parts. What’s more, as generative design is unleashed on more materials, additive machines, particularly those that print in metal, will need to be designed to use a wider range of alloys to meet design requirements.

But generative design’s isn't just about optimizing a shape or lightweighting a product. The concept extends well beyond the physical nature of a component and Fallon wanted to make that point clear. Fallon insisted that generative design wasn't limited to form synthesis and optimization saying that the technology could be extended across the product design and manufacturing landscape.

While Autodesk has stressed the point that generative design does have a topology optimization component, its only one of the types of solvers that will be integrated into its new software. In its most grand incarnation, Generative Design will be a platform where any question can be asked of the program and, if a solver is present to crunch the numbers, the software will deliver a series of results.

To say that Fallon's idea for his new software is ambitious might be an understatement. 

One example of this wider vision was Greg’s idea that  users could soon begin to think of their product as a system. Not only would they be working with the physical object, but they’d be optimizing a product for manufacturing and even the supply chain all from a core Generative Design interface. What’s more, if a factory needed to be reconfigured to launch a new project, this new tools could be harnessed to develop the plan for that action.

Greg even mentioned that, in the future, Generative Design will even be able to create the g-code that will mill a product.

Finally, and on the complete other end of the design ecosystem, generative design could make it easier for makers to create disruptive designs that shake up niche industries. As was said before, we're possibly looking at a tool that can augment the abilities of an engineer to deliver a truly remarkable product.

With the right idea, a product can be optimized for any number of traits making it more valuable to its end user. And that extends to customization as well. With a Generative Design tool, product engineers can quickly take the input they receive from a client and create a design that’s bespoke and fits their individual design criteria. That alone could corner a market.

So, in the end, the technology is still in its infancy (and is admitted amorphous and difficult to grasp) but Autodesk has a grand vision for how generative design will transform engineering. Fallon sees Autodesk Generative Design being rolled out across its product line touching everything from AutoCAD to BIM360 and beyond. In fact, this new product will be more of a design tool than the software you’re using today. As Fallon put it, “those products (AutoCAD, BIM360, Fusion360, etc.) will be the tools that document what’s created by Generative Design.” So, Generative Design might end up being the node from which all design springs.