Industryweek 31159 Capture 1 0
Industryweek 31159 Capture 1 0
Industryweek 31159 Capture 1 0
Industryweek 31159 Capture 1 0
Industryweek 31159 Capture 1 0

(Re)Designing the Future with Generative Design

July 24, 2018
Sponsored by Autodesk

In May, General Motors took its first step into a new world of industrial design that has the potential transform the industry.

The auto giant has now officially become the first North American automaker to adopt Autodesk's new generative design solution which it will use to tap into the combined strength of artificial intelligence, metal additive, and subtractive manufacturing expertise.

For now, GM is taking a small step into the future—dipping its toes into the full generative design capabilities with a redesigned seat bracket.

It's an initial proof-of-concept project, sure, but the results are already making the proof hard to ignore.

The new part—an unworldly, twisting swirl of metal that no human would have ever dreamed up alone—combines the 8 individual pieces of the original design into a single build that weighs 40% less than the previous construct, but is also 20% stronger.

This is exactly what the hype around generative design has been preparing us for over the last few months; it's exactly the kind of mind-blowing improvements that should be impossible through one system change. Yet here it is. And it's only the start.

This means that this tiny piece could be the beginning of a whole new design revolution in this revolution-rich tech era.

So What is Generative Design?

On the face of it, generative design is a fairly simple concept: Engineers or designers, plug in the requirements of a new part and the system spits out endless iterations of design solutions that fit. Simple enough, right?

Well, no. It goes much, much deeper than that.

To Rob Cohee—Autodesk's senior manager of manufacturing technical marketing and evangelism —the technology transcends comparisons with any normal design software tool. The better comparison, he says, is with Watson—IBM’s AI rock star.

"Watson is a supercomputer that does things that no other computers have ever been able to do and processes information at a rate that is just absolutely staggering," he says.

Autodesk's generative design solution, he argues, does the same thing for manufacturing design engineers.

"Generative design empowers [engineering organizations] to explore many more alternatives to create a wider body of data that they can then utilize to make absolutely sure that they've explored a solution domain to make the best decision for their business," he says.  

This data-based approach, he explains, allows these companies to move designs forward into production and take something to market at speeds that would be unimaginable with even whole teams of engineers.

In this sense, the technology goes far beyond just spitting out new designs and into a fundamentally new process for designing products.

(Re)Designing the Future

Many early adopters of the software, Cohee notes, have been tempted to describe it as a simulation tool or a topology optimization system, a part consolidation software or finite element analysis technology.

All of that is true, of course, but "it does all of these things at the same time," Cohee says. "It considers not one improved design alternative, but rather dozens, hundreds, or thousands of potential design alternatives."

Essentially, it flips the role between software and engineer in the design process. In the past, the creative work all existed in the engineer's head and the CAD system simply recorded those ideas. “As an industry, we’ve suggested CAD is computer “aided” design. Yet what does CAD software do that you didn’t explicitly tell it to do?“ Cohee suggests. Generative design places the creativity into virtual hands of the system—engineers enter the requirements of a piece (where it attaches, tolerances, materials, etc.)—and the system uses its vast resources to iterate its way through any number of possible solutions.

And, unlike with traditional design engineering, it doesn't do this in a vacuum. The designs can be simulated and tested for performance in real world applications to help narrow down the right options for the design.

"Let's say, a new design is holding up for strength and it's maintaining tolerance levels of detail for things like displacement, but what if we apply that load tens or hundreds of thousands of times?" Cohee says. "We can then incorporate fatigue into the problem statement as well. We can incorporate things like at what point do we exceed the yield and begin to buckle, how much of that buckling is acceptable as part of the design?"

In the process, the system also solves one of the most frustrating elements of design: manufacturability.

"It's funny how many times I've had somebody with welding gloves tap me on the shoulder of my white collared shirt saying, ‘Hey I can't make this,'" Cohee recalls. "That's a problem, and not only for your laundry. It's a problem when manufacturability and the process that you have on the shop floor don't make their way into how to inform the product development process."

Along with the performance requirements of a part, he says, manufacturability—from the shop capabilities to machine access to materials—can be integrated into the design process to help guarantee that you're left with a design that will work and that you can actually make.

The system, Cohee says, helps demonstrate that part is not only strong enough to hold up and maintain its tolerance during its lifetime of abuse, "but that it's also 100% manufacturable in a 3, 3+2, or 5-axis subtractive manufacturing process."

AI Generation

Looking at the full picture, Cohee's point about the Watson comparison really begins to set in. It seems to be exactly where the technology is heading—connecting what would otherwise be completely disparate decisions in a design-to-manufacturing process and meshing them together to produce designs and manufacturing instruction to get better parts made faster.

And that is really the point.

"As more and more of our simulation and manufacturing capabilities are worked into generative design, it (being process aware) should reduce that back and forth between manufacturability, strength, aesthetic, and design intent," Cohee says. "When you incorporate all those things into generative design, you can get your products to market a heck of a lot faster that way."

Click here to learn more about generative design! 

Popular Sponsored Recommendations

Empowering the Modern Workforce: The Power of Connected Worker Technologies

March 1, 2024
Explore real-world strategies to boost worker safety, collaboration, training, and productivity in manufacturing. Emphasizing Industry 4.0, we'll discuss digitalization and automation...

3 Best Practices to Create a Product-Centric Competitive Advantage with PRO.FILE PLM

Jan. 25, 2024
Gain insight on best practices and strategies you need to accelerate engineering change management and reduce time to market. Register now for your opportunity to accelerate your...

Transformative Capabilities for XaaS Models in Manufacturing

Feb. 14, 2024
The manufacturing sector is undergoing a pivotal shift toward "servitization," or enhancing product offerings with services and embracing a subscription model. This transition...

Shifting Your Business from Products to Service-Based Business Models: Generating Predictable Revenues

Oct. 27, 2023
Executive summary on a recent IndustryWeek-hosted webinar sponsored by SAP

Voice your opinion!

To join the conversation, and become an exclusive member of IndustryWeek, create an account today!