Hello and welcome to Additive Insight, the original additive manufacturing podcast and your source for news, interviews, and comments on the latest 3D printing and AM intelligence brought to you by TCT Magazine. I'm your host, Sam Davies, and today I'm joined by Hyliion CEO, Thomas Healy.
Hyliion is a publicly listed manufacturer of clean modular power generation systems, best known for the Karno generator technology it acquired off of GE Aerospace in 2022. When Hyliion made that investment, it also inherited a significant metal additive manufacturing capacity that facilitated the early development of the Karno technology and is now deployed to manufacture commercial products like the Karno 200 kW power module.
Throughout our discussion, Healy shares more on the capabilities of Karno, the company's AM workflow, and why Hyliion wouldn't have a product without additive. We also discuss how the company has scaled to a fleet of three dozen metal AM systems, and how we expect Hyliion's application of the technology to evolve.
If you like what you hear, don't forget to subscribe wherever you get your podcasts. You can also head over to TCTMagazine.com to subscribe to the TCT membership platform and receive more additive insight via our print and digital offerings.
Thomas, welcome to the Additive Insight podcast. We'll be talking today about how additive manufacturing is being used in your mission to deliver more efficient power generation technologies, but first, I was wondering if you could take us back to the start. So, Hyliion was founded over a decade ago. Can you tell us the story of how and why the company was established?
Absolutely. So, this is your stereotypical startup journey — it's been a roller coaster, it's been a lot of fun and exciting, but it's come with twists and turns as well. But just to level set, we're now a company focused on the power generation space. So we're really using additive manufacturing to enable a really new type of power gen. So how the company started, back when I was in college at Carnegie Mellon Pittsburgh, I had this wild idea of let's go electrify semi trucks, bring these big 18-wheelers, move them off of diesel, move them onto electric, and that was the origin of the company.
We took it public on the New York Stock Exchange in 2020 on that premise, and then fast forward to 2021, GE approached us with this really new innovative tech they had. It's called the Carnot power module, and as I was just mentioning, it's like a new form of power gen, a new way to make electricity. So fast forward a year later, we actually ended up acquiring that whole division out of GE, brought it under Hyliion, and then in 2023 we saw that the EV space was not heading in a good direction, right?
There were a lot of our peers that were going bankrupt. It was tough to raise capital. The pace at which fleets were adopting new electric vehicles was slowing down. The costs were higher. The mandates were softening. So it was just this perfect storm that we decided let's get out of the EV space and why don't we go all in on power gen, and that has been a fantastic decision as you look at the amount of power data centers need, the military itself — we're working on various different programs — they need a significant amount of power with unique attributes, and then just even commercial buildings. Like if you can provide a solution that someone can now make electricity cheaper than they can buy from the grid, why would you not do that?
Okay. So as you mentioned, 2022 was a key moment in the company's history, acquiring Carnot from GE. Can you tell us exactly what you were acquiring at that time? How mature was that technology? And I guess what had GE developed up until that point in 2022?
Yeah. So, it was at a phase where prototypes had been made, but nothing that was ready for going to market, commercial application. So, developing a new type of power generation technology — like envision going and developing a new turbine for a jet engine, or a new fuel cell. These type of endeavors take many, many years, if not in the order of magnitude of like a decade to develop. The Carnot is no different.
So, the team within GE actually started working on this in the 2017-ish timeframe, and when we made the acquisition in 2022, it was at a phase where it had already gone through, I think it was three different iterations of prototypes. And then when we made the acquisition, we even then brought it through another iteration or two that then is now what we're bringing to market as our commercial product. So, what it is, is it's a 200 kW design system that's called a linear heat generator. We'll go dive into this in a lot more detail, but the foundation is it actually uses heat as its fuel source in order to produce electricity.
Now, how do we make heat? You can make it by reacting natural gas, diesel, hydrogen, propane — truly a variety of fuels that we can feed into our system, make heat, and then we'll be providing electricity through the power module.
Okay. So, you mentioned that a couple of iterations since 2022 when you acquired the technology. So the commercial technology as it exists today, what would you say are the key capabilities of the technology? What does it enable? What does it allow your users to do?
So the general premise is you can now make your own electricity on site as opposed to being reliant on the grid. So if you think about a commercial building, most facilities have a big air conditioning unit either sitting on the roof or sitting outside to produce all the cold air for the building. Our vision is you do the same exact thing for power gen. You have a box that's sitting next to the air conditioning unit, and it's sitting there getting fed with whatever fuel you have available to you, and then it will produce all the electricity that that facility needs.
And what this enables is, for data centers where the grid doesn't have the capacity that data centers need, now they can move to on-site power generation and actually move faster than they could have before. If you look at a shopping mall, it enables them to now actually make their own electricity cheaper than buying grid electricity.
And then if you look at applications like the military, we're the power plant on board the new unmanned military ship. So their nickname is the ghost ship. So there's no crew on board. It's this fully autonomous vessel. Well, if you don't have a crew on board, you don't have the ability to do maintenance on engines, right? So now what the Navy needed was a solution that was very low maintenance, could be deployed for months and months on end, potentially even years on end, with zero maintenance. And that's one attribute of our technology — it's inherently a very low maintenance solution.
I understand that it's also a scalable technology. So you mentioned that the first product you have is the Karno 200 kW power module, but I believe that you're looking ahead to a larger product. So, talk to us about that kind of roadmap and the progress you're making there.
Yeah. So, the 200 kW is perfect for — think of a Home Depot. One of those boxes could power an entire Home Depot. A Walmart Supercenter, two of the 200 kW boxes could power an entire Supercenter. Now, the reason that they need more power, they have all those fridges for the food in the facility, so that draws more power. But if you think about data centers, we're talking more in the order of magnitude of megawatts of power as opposed to kilowatts, or potentially even gigawatts of power.
And so, what we've also designed is a more modular system that is either 2.4 or 3.2 megawatts of power. So, if you think about a 30 megawatt data center, they would deploy 10 of these 3.2 megawatt boxes, roughly, outside their facility, and that will produce all the power that they would need at that site.
One unique thing though that I'm sure we'll discuss is inside that 3.2 megawatt box is actually just a series of these 200 kW systems strung together. So, 16 of them strung together in order to make that 3.2 megawatts. And then you string those 3.2 megawatt boxes together.
So, the reason that we're doing that — we get asked this all the time, is why not just make a larger four-shaft system that would make 3 megawatts itself? Well, the reality is that we've designed the system to really match the production machines, the additive manufacturing equipment, so that we're optimizing the size of these parts to fit on the print beds of these machines. And so, if we make them just a millimeter bigger, now our production throughput is going to significantly take a step back.
So, are you then constrained in terms of size of these systems by what AM can do as opposed to the challenge of scaling the technology into a larger form factor?
That's correct. So, we've designed the parts to really match that 250 by 250 mm build plate size. And so, our parts are 249 mm in diameter. And so, that's where if we went bigger, now you're no longer able to use the M2 GE Additive machine. Or we also use the M line machine — it's a 500 by 500 mm build plate. We can fit four parts on that. If you go to the parts being a little bigger, now you can only fit three, then you can only fit two, right?
And that's where we looked at it and said, yeah, the real optimized size for additive manufacturing is this 250 mm. Let's just stick with that, and then as people need more power, you just string them together. Think of it like a battery pack in a car. You know, you think of a Tesla — the whole floorboard is filled with thousands of these small little batteries that are all strung together. That's our same philosophy for power generation: you string a bunch of these Carnot cores together at 200 kW in increments, and that gives you the power you need.
Yep. So, additive manufacturing — I assume when you acquired the Carnot technology, that's kind of what they were doing. They were using AM for these power generation systems. Why is AM the kind of most suitable manufacturing method for these products? And I guess how important is AM to what you're doing here?
We don't have a technology if it's not for additive manufacturing. So, that is how critical this is to our business. So, if you think about our technology, it's based on Stirling engine tech, which is this concept that was designed and developed 200 years ago by Dr. Stirling, and he had this concept of if you trap a gas — so think of like a balloon, you fill a balloon with helium, right? Well, if you heat that balloon up, it would grow. Now, if you cool it down, it shrinks. Same thing with your car tires in the winter — why do you need to put more air in them? It's because it got cold outside and it shrunk.
So, we do this heating and shrinking 20 times a second inside this chamber inside our engine, and that allows us when it expands to actually move a piston, and then that produces electricity, and then you shrink it back down and it compresses the gas and pulls it back. And so, that's how our system works, right? And that's a Stirling engine — been around for 200 years.
No one has really brought this to market in a commercial application. Organizations like NASA have used it for powering satellites, things like that, but it hasn't really been commercialized because it's been so difficult to manufacture — specifically, it's so difficult to manufacture the heat exchangers that go into the Stirling engine.
And so, what we looked at, we said, "All right, Stirlings have always for 200 years been kind of known as this euphoria of power generation. Well, there's been a massive breakthrough in manufacturing over the last couple of decades thanks to 3D printing or additive manufacturing. Why don't we try to make those really complex heat exchangers with additive as opposed to conventional practices, and now that's enabled us to really unlock these euphoria benefits of a Stirling engine."
How do you approach the design of these systems? Has that been an in-house capability or has that been a collaboration with third parties?
It's really been an in-house effort. So, we have a rockstar engineering team. It's led by our CTO Josh Mook, who was one of the founders, really the founding member behind the Carnot technology within GE. And then he had assembled a team around him at GE of brilliant aerospace engineers, additive manufacturing engineers. And then as we made that acquisition out of GE, that team came over as well.
So, brilliant group, deep history in both additive as well as aerospace. And they've taken the reins within Hyliion to really go design all these components themselves. I would say we've got probably one of the strongest — obviously I'm very biased, I love our team — but we've got one of the strongest, if not the strongest, designers for additive manufacturing out there.
And we're also able to then go push the bounds of what these machines are able to do, right? Because the engineer that's actually designing the components to go into our Carnot power module is also an engineer that has experience on how these machines work, and can really push the thought process as to what can be created through additive.
