Pic of the Week: Faster Lightweighting Through AI
An AI model can help designers get to the best lightweight solution faster — by presenting a limited number of optimal solutions.
Share
Industry moves fast.
We make sense of what’s next.
Manufacturing Connected is a new platform from Additive Manufacturing and our sister publications. Bringing you big-picture insights that cross technology and industry boundaries.
GET CONNECTED
In the past part designers have struggled with tools that aren’t up to the task of developing the complex designs that additive manufacturing can achieve. But with the rise of generative design, the pendulum may have swung too far in the opposite direction: With so many possibilities at the engineer’s fingertips, how does he or she zero-in on exactly the right one?
Today’s Pic offers one potential solution. AI-enabled software company Vixiv has spent the last several years 3D printing lattices, crushing them and feeding the data into an AI model, all to build a tool that can better enable the use of these lightweighting structures in 3D printed parts.
Crucially, Vixiv’s software does not ask the designer to set parameters at a granular level, nor does it present the entire field of possibilities. Instead, the cloud-based AI engine fits different lattice designs into the volume based on shape, material and application requirements set by the user (strength or stiffness, for example).
In just minutes, the software provides 5 to 20 iterations to choose from, all of which can solve the given problem. (But if the software doesn’t have a lattice that works, it won’t hallucinate — this is one AI that isn’t afraid to admit when it doesn’t know the answer.)
“It’s basically saying, I can get to your strength requirements in six different ways,” says Aaron Chow, CEO. “Which do you prefer?”
From there, it is up to the user to select the final design using their own domain knowledge, cost sensitivities and other preferences. While lightweighting is a primary objective for nearly all lattice applications, a user might opt for a slightly heavier part because it is cheaper or easier to manufacture for instance.
With this solution, designers don’t need to understand the intricacies of lattice cell design, nor do they need to work through dozens of iterations.
“It can cut 5 hours of design work to 10 minutes,” Chow says. “This is a way to support that leap to additive for everyone that doesn’t have a team of PhDs with extensive design experience.”
Vixiv entered public release on October 1, offering its AI model tuned for Multi Jet Fusion (MJF) in nylon 12. Data from early users will help to inform future iterations of the tool, including models for metallics and other 3D printing processes.
Related Content
How Additive Can Complement Subtractive Manufacturing
Think of additive and subtractive as complementary processes, rather than being in opposition.
Read MoreLoudspeaker Enabled by Polymer 3D Printing: Pic of the Week
Node Audio’s Atom loudspeakers incorporate multiple 3D printed components to deliver mechanical and acoustic performance.
Read More3D Printing Enables a Larger "Sweet Spot" Relative to This Golf Club's Size: The Cool Parts Show #77
Cobra Puma Golf's Limit3D iron uses 3D printed stainless lattices to remove weight from the interior so that it can be reintroduced with tungsten inserts that lower the center of gravity. The combination offers a smaller, sleeker club head with the forgiveness of a larger one.
Read More3D Printed Tennis Racket Serves a Collaborative Win
The collaboration between All Design Lab and Protolabs led to the creation of a 3D printed tennis racket via direct metal laser sintering, called Hìtëkw. This project not only pushed the limits of additive in terms of design capabilities but revealed how important active communication between customer and manufacturer can be for a project’s success.
Read MoreRead Next
Pic of the Week: Cooling, Compacted
This week, we’re looking at a 3D printed water-charge air cooler that weighs just 1.4 kg — down from the 16 kg of a traditional air-to-air setup.
Read MoreLightweighting Without the Lift? Vixiv Launches Software for Instantaneous Lattice Design
Artificial intelligence plus months of sample testing enable a tool for lightweighting via additive manufacturing that does not use generative design and does not need high computation demand. Available for Multi Jet Fusion now, with more AM methods to come.
Read MoreCrushing Lattices to Feed An AI Model
Why deliberately destroy 3D printed lattice structures? Software developer Vixiv is doing it to inform its AI, making lattice designs faster and less complicated to access.
Read More

