Interview first published in the May/June 2021 issue of 3D ADEPT Mag, then online in January 2022. Roles, products and outlook reflect 2021.
Tech companies that emerge out of anonymity often do it with a bang: through a big launch event to capture the public’s attention or by announcing a seed round but I strongly believe the key in sparking interest lies in the way a company explains how it can transform industries. The latest company that decided to turn stealth mode off is Hyperganic and I sat down with CEO & Founder Lin Kayser to understand the transformation they ambition to bring to the AM industry.
Headquartered in Munich, Germany, Hyperganic, in simple terms, develops and provides a software platform for advanced design and engineering of parts, structures and entire machines of extreme complexity. What makes them outstanding? Artificial intelligence (AI) and algorithms. As per Kayser’s words, they “create the design of physical objects using AI and algorithms.” “We mass produce them in digital factories that have AM at their heart. Our goal is to radically accelerate innovation and it all starts at the design level with engineers who want to create new objects.
The truth is, there is a lot of manual work involved in the work of designers. They have to match things in their head and put them on their screen. The process is like building every single façade of the building you want to build. At Hyperganic, we take the thought process of the engineer which requires to answer the question “how do I build something?”, we put it into algorithms and then, the algorithms can automatically recreate parts that otherwise the design engineer would create. The algorithms bring him to the next level to help him produce the things he would like to see and that’s the key enabler of AM.”, he adds.
Algorithmic design translates engineering rules into computational procedures. It does not necessarily involve a machine-learning model; “algorithms” and “AI” are therefore not interchangeable terms.
According to Kayser, a big number of manufacturers do not see the utilization of AM beyond prototyping or small batch runs. Therefore, when it comes to mass production, they often think of another manufacturing process. The Hyperganic team wants to enable this mass production and wants to do so especially for the most complex parts that may absolutely require AM. In Kayser’s opinion, such type of parts are so “tailored to AM, so advanced that not only does AM make sense in terms of functionalities, but also financially and for mass production”.
How it started…
In this 2021 interview, Kayser looked back on an entrepreneurial career that began around thirty years earlier. His previous company worked on image processing for cinema. He cites An Inconvenient Truth, associated with Al Gore, among the influences that led him to reconsider his contribution to environmental challenges.
He recounts selling his company in 2011 before exploring other directions.
“For me, innovation was something that was happening fast but only in very specific fields like computing or IT but when we look at physical objects, nothing has changed that much in decades. I can’t help but think that was one of the issues to address, in order to be more sustainable, to move forward faster and solve some of the issues we were facing at a global level. Fortunately, in 2012, I stumbled across 3D printing and I was hooked”, he recalls.
The thing is, the software guy in him knew that coding and software development was one of the most creative things someone with his background could do, because “you imagine something in your head and you give life to it. But it’s hard to do it physically because software is virtual.” “When I discovered 3D printing [hobby 3D printers], I realized that it is a machine that can essentially print codes”, he notes.
The absence of a usable digital model can hold back a manufacturing project. A 44% figure repeated in an article about ZVerse illustrated this problem; without a detailed scope and method here, it should not be treated as a representative measure of the whole industry.
“We are designing these models in a traditional way of creating CAD models for physical objects. We have been doing that for the last 15 years or so. I started asking myself if we could use computer codes to build physical objects. Instead of modelling something visually, can we use AI and algorithms to create physical objects, structures and maybe entire machines? That was in 2014 and that’s how the idea for Hyperganic was born”, he explains.
Together with Michael Gallo, CTO of his previous company and current CTO at Hyperganic, they sat down, discussed this idea, and hesitated. They finally started developing a software platform based on a volumetric representation of geometry; in other terms, “a workflow that will create various variations of different parts in order to automate the work of the design engineer so that designing becomes finding the right object in these variants”. This means that in the operating systems they create, one could automatically design objects of a similar type. Therefore, the final look of the end-part will depend on the environment and the parameters the design engineer would have fed into the algorithm. That was in 2017. Hyperganic was officially founded.
A voxel is a digital volume element, comparable to a three-dimensional pixel; it does not represent a molecule. Kayser describes this geometric-kernel choice in his account of Hyperganic Core.
How it’s going…
During the 2021 interview, Hyperganic reported developing its customer base in Europe, the US, China and Southeast Asia. The $7.8 million funding round discussed in this feature refers to early 2021, to support the team and platform.
German funds HV Capital and VSquared Ventures led this first VC investment round for Hyperganic, and co-investors included US-based Converge, industrial partner Swarovski and PC pioneer Hermann Hauser, co-founder of ARM.
At the heart of the developments of the company, there are the Hyperganic Core and Hyperganic Print Framework platforms.
In the 2021 presentation, Hyperganic Core was the voxel-based geometry kernel for algorithmic design. Digital detail does not guarantee equivalent physical printing resolution. Hyperganic Print Framework addressed build preparation and process control. Access modes and functions discussed in the interview are not a description of the current offering.
Design, simulation, build preparation, production management and quality control are distinct software functions. One supplier may cover several, as Kayser explains here.
“We do much more than just the “design” part. We can ensure mesh repair, slicing, support generation, nesting, stacking, voxel-level process control and even multi-material support”, Kayser explains.
How does Hyperganic platform do what it does…
The company presented an approach intended for several process and material families. Actual compatibility nevertheless depends on machine interfaces, formats and the available level of control.
“We work with polymer 3D printers, metal AM technologies, resin-based systems. We even support manufacturers who print with concrete. We cover the full spectrum of materials on the market”, the CEO outlines.
Discussed applications include aerospace, consumer goods and medical products. Customisation, complex structures and optimisation can motivate this approach; neither voxels nor complexity automatically make a part useful, manufacturable or qualified.
Examples presented at the time include a bicycle helmet and a 3D printed rocket-engine geometry. These illustrate design and manufacturing; the article does not provide helmet certification or engine flight-qualification evidence.
As far as the rocket engine is concerned, the team decided to explore an entirely algorithmic approach for its creation.
“We decided to give the algorithms just one geometric starting point, a spline curve that describes the inside of the combustion chamber, stored in an Excel spreadsheet. There were no CAD files, no existing models, just data, stored in a spreadsheet, plus the algorithms that interpret the data to generate the functional parts of the engine. With this minimal starting point, we built the algorithm that generates the actual geometry. The engine is created top to bottom, laying out channels that transport the cryogenic oxidizer first down around the chamber, cooling the inner wall that is exposed to the burning fuel, and then back up again in an outer layer of channels, for later injection into the chamber itself.
We used Hyperganic‘s capability to precisely control the process parameters for every point in space to modify the structure of the metal itself. The inside of the chamber is printed very densely, using high laser power, whereas the outside becomes almost porous to keep the weight down”, a company’s report reads.
This passage describes the company’s experimental approach. Laser power and porosity alone cannot predict component performance: simulation, material testing and functional validation remain necessary for the intended application.
The approach programs a geometry-generation process rather than manually drawing each variant. Choosing rules, constraints and evaluation criteria remains engineering work.
“Traditionally you start with an existing solution and make incremental changes. In Hyperganic engineering, you regenerate objects from scratch all of the time. You generate hundreds or thousands of variants that can be evaluated. In traditional engineering, redesigns are costly and need to be avoided; in Hyperganic engineering, it is the norm and the key to coming up with a global optimum”, the report highlights.
Producing many variants does not, on its own, prove that a global optimum has been reached. That depends on the explored space, objective, constraints and verification method.
What’s next for the company?
The commercial objective presented in 2021 was to expand platform use beyond individual engineers to third parties wishing to develop their own applications.
Moving forward, Kayser strongly believes that AM will not stay a niche forever. On the contrary, for him the technology will become one of the most important fields in manufacturing in a few years. “By bringing algorithms into this game, we create a new type of manufacturing that has not been done before. We are providing that missing piece that will lift AM to a new level”, he concludes.
This exclusive feature has first been published in the May/June edition of 3D ADEPT Mag.
THE 3D ADEPT WEEKLY NEWSLETTER
The industry moves fast. Stay ahead.
Every week, the latest additive manufacturing news, in-depth features and real-world applications, straight to your inbox. Plus, be the first to know when a new issue of 3D ADEPT Mag is out.
Free · One email a week · Unsubscribe anytime
Clicking loads our email provider’s sign-up form, including its analytics and Google’s anti-bot protection. You’re only subscribed if you complete the sign-up. Read our privacy policy.