When Michael and I started thinking about what eventually became Hyperganic in 2017, we knew that the complexity of the objects we’d create would approach the complexity of natural objects. If you eliminate the factor of manual labor, which is present in today’s Computer Aided Design (CAD) tools and progress to Algorithmic Engineering, things become way more intricate. If the computer just aids you in creating a design, you, as the human, still hold the pencil that creates the drawing. If the magic brush is held by computer code, even simple algorithms can produce very complex and functional results.

So the question was always, how do we evaluate these resulting objects? In traditional engineering, most things are over-engineered — humans like to stay on the safe side. The cost of reengineering an object, if it doesn’t perform properly, is often prohibitive.

Staying on the safe side is the opposite of innovation, though. You need to move past the breaking point and then step back a little. That’s how you find out what’s possible. In traditional engineering, this process is painful and takes weeks, months, sometimes even years. So people do it only if it is absolutely necessary.

In Algorithmic Engineering, when a computer code creates functional physical objects, a new iteration is only a few minutes away, even for complex parts.

Additive Manufacturing allows us to quickly produce modified parts and put them to a physical, real test. And that’s what our customers have been doing, and doing quite successfully. It’s much faster than traditional iteration cycles.

But couldn’t you move the whole cycle into the computer? This is where simulation and physics come into play. But it’s not that simple.

Conventional simulation tools may be refined, but they struggle with the complex objects that come out of our algorithms. Often they could not even load the simplest of the models that we provided. Also, they make you install complex software that’s usually driven by a human and requires a process called “meshing”, a laborious task of manual preparation that often takes longer than the execution of the simulation itself.

In one of the last conferences before the COVID pandemic, I sat together with entrepreneur Sebastian Thrun. He looked at me after hearing what we are doing and said, “You absolutely gotta own physics.” I said that it’s not so simple, but I wrote it down and it stayed on my mind.

Last year, I gave my usual keynote talk at Xpreneurs, the great German incubator program. After my talk, one founder in the program, Nina Korshunova, came to me and told me about what she was doing with her just-founded startup, DirectFEM.

It was a match made in heaven.

What Nina and her team created, was the perfect physics engine for Hyperganic. It draws on more that 15 years of work at the Chair for Computation in Engineering at the Technical University of Munich.

DirectFEM’s quasi-mesh-less simulation kernel works directly with our data model: voxels, more or less 3D pixels. It basically works on the particles that a 3D printer outputs, just like we do for the engineering side.

With DirectFEM’s simulation technology, we can now directly physically evaluate every design that comes out of Hyperganic Core. As a result, we can build iterative self-optimizing feedback loops that don’t require external evaluation and complex handshaking with off-the-shelf software.

It was clear that DirectFEM should be our first acquisition for Hyperganic. I am super happy that the entire DirectFEM team decided to come on board and to bring their technology to Hyperganic Core. And just as promised, we have already expanded their team to accelerate things even further.

We are now working hard to make Hyperganic Physics a key part of Core 3 at launch day. It’s a game changer. And, yes, thank you, Sebastian, for reminding me of the importance of a physics engine in our tech.

Thoughts? Connect with me on Twitter or LinkedIn and let’s discuss.

About Lin Kayser
Lin is the co-founder and CEO of the Hyperganic Group. His entrepreneurial journey stretches back to the early 1990s and covers areas as diverse as industrial control systems and transforming Hollywood from analog to digital.

This is his personal blog which contains many posts that pre-date Hyperganic. His views are his own.