Hyliion Holdings Corporation - Ordinary Shares - Class A Stock

Hello, everyone. Thank you for joining us and welcome to the Hyliion Holdings Second Quarter 2026 Earnings Conference Call. [Operator Instructions]

I will now hand the conference over to Greg Standley, Chief Accounting Officer. Greg, please go ahead.

Thank you, and good morning, everyone. Welcome to Hyliion Holdings' Second Quarter 2026 Earnings Conference Call. Joining us today are Thomas Healy, Chief Executive Officer; and Jon Panzer, Chief Financial Officer. A slide presentation accompanying today's call is available on Hyliion's Investor Relations website at investors.hyliion.com.

Please note that during today's call, we will be making certain forward-looking statements regarding the company's business outlook. Forward-looking statements are predictions, projections, and other statements about anticipated events that are based on current expectations and assumptions as such, are subject to risks and uncertainties. Many factors could cause actual results to differ materially from forward-looking statements made on this call.

Factors that may cause such differences are discussed in our presentation and press release, as well as our filings with the Securities and Exchange Commission. You are cautioned not to place undue reliance on forward-looking statements, and we undertake no duty to update this information except as required by applicable law.

With that, I'll turn the call over to Thomas.

Hello, and thank you for joining us for Hyliion's second quarter 2026 earnings call. This was a strong quarter for Hyliion, and we have a lot to cover on today's call. I'll organize my remarks around 3 topics: military progress, product progression towards initial customer site deployments along with customer demand, and speed improvements in additive manufacturing.

First, a few highlights. We were awarded a $41.7 million contract with the U.S. Navy, our largest military contract to date, and we still expect to secure additional military contracts before year-end. We are raising our full-year revenue guidance from $10 million to $15 million. Finally, we have identified additive manufacturing speed improvements that we believe have the potential to increase print speed and throughput by up to 3x. I will cover our progress against the 2026 milestones we laid out at the start of the year, update our 3-year outlook, and then turn the call over to Jon for the financial results.

Starting with the military, last month we announced a $41.7 million contract with the U.S. Navy to scale the KARNO Power Module into multi-megawatt systems. Under the contract, we expect to deliver 2 power modules, 1 rated above 2 megawatts and the other rated above 3 megawatts. Both are built on the same modular 800-kilowatt architecture we are developing today for data centers and the USX-1 Defiant autonomous Navy ship, which means we are scaling an existing building block rather than creating a new system.

These larger modules expand the range of Navy applications from larger vessels to powering military bases. The contract also includes funding to further advance our additive manufacturing capabilities. Coming into the year, we set a goal of securing $40 million to $50 million in new military contracts. This Navy award on its own achieves that goal. However, we expect to close additional military awards this year, including one from a different service branch for approximately $7 million.

We expect these additional awards to bring us close to $50 million in new military contracts for 2026. The performance periods on these contracts run 2 to 3 years, depending on the specific contract and our pace of execution. We expect additional military contracts next year and in the years to follow. Some will be cost-plus development contracts similar to what we are executing on today and increasingly, others will be for delivery of complete commercial power systems.

The result is a steadily growing base of military and government contracts that support revenue growth in future years. This quarter, we engaged Abdul Subhani as a strategic advisor for military opportunities. Abdul serves as Civilian Aide to the Secretary of the Army for Texas, Distinguished Chair of Innovation and Senior Advisor to the Superintendent of the United States Military Academy at West Point, and U.S. Technology Advisor to the British Military Academy. Abdul is assisting us with building relationships with new contacts throughout the military, particularly within the U.S. Army, to broaden awareness of our KARNO technology and build on our contract pipeline.

Our engagements across all branches of the military increased significantly this past quarter. We held numerous meetings with senior-level officers to discuss the potential of our power system. We were also invited to participate in Senator McCormick's Defense and Innovation Summit, where President Trump's Energy Secretary, Chris Wright, and Secretary of War, Pete Hegseth, all participated. Overall, it was a terrific quarter for defense progress with more engagements and business opportunities expected in the quarters ahead.

Shifting to product development, I will cover where we stand with our development work and deployment of early adopter units. Overall, commercial interest remains strong as we near initial customer site deployments. Ahead of moving units to customer sites, we are completing a block of design enhancements that capture improvements from what we have learned from initial operations. The block consists of changes spanning airflow and cooling improvements, upgraded insulation blankets, an improved piston, and software and controls enhancements.

Together, we are expecting to increase system durability, performance, and power. We have recently completed updating our first commercial customer unit with these upgrades, and that will be the first system deployed to a customer site. We are building a number of power modules in parallel, including the 800-kilowatt Navy system, which is now in assembly. We are accumulating run hours across multiple systems simultaneously and performing extended duration runs to build operating hours and validate durability.

We are also expanding our manufacturing capacity, including moving key assembly operations from Cincinnati to Austin. We continue to expect to complete the remaining early adopter units, approximately 10 KARNO Cores in total, this year. We expect our first customer site deployment to begin over the next quarter, with multiple units going to that initial site. We are planning a data center deployment to follow, and the balance of units are Navy assets, including the 800-kilowatt system for the USX-1 Defiant autonomous Navy ship.

We also recently announced the successful operation of multiple KARNO Power Modules functioning together as a single scalable power unit. This matters for customers who will run more than one system at a site, and it completes another of the 2026 milestones we set out at the start of this year. We also continue to make progress towards our 200-kilowatt target and expect to reach that power level by year-end. We are testing improvements across several components with our primary focus on continuing to iterate the design of the regenerator, which we believe is the remaining key enabler to achieve full power.

Shifting to customers, our 2 largest segments of opportunities are data centers and the military in that order. Data centers represent our largest area of customer interest, driven by AI-related power demand and the sheer size of the market. Over time, we expect a significant portion of our production capacity to be directed towards this segment. We have quantified part of that interest through non-binding LOIs representing approximately 750 KARNO Cores, which are subject to the execution of definitive purchase agreements.

Most of the interest we are hearing from customers is not yet reflected in LOIs or purchase contracts. And while we have not announced new data center LOIs since our last update, we are engaged with a growing number of companies discussing interest in deployments ranging from tens to hundreds of megawatts. We are seeing 3 buckets of data center projects: low tens of megawatts, around 100 megawatts, and gigawatt scale. For the small and medium sites, our discussions center on the KARNO Power Module as the primary power solution, given its efficiency, scalability, and compact footprint.

For gigawatt scale builds, a combined cycle gas turbine will likely provide baseload power, and we are being considered to handle 10% to 20% of the overall need, specifically to handle transient loads. Across all 3, there is a strong interest in our system's native 800-volt DC capability, which is the architecture data centers are moving towards.

On the military side, interest spans autonomous vessels, base power, and forward operating installations driven by mobility, fuel flexibility, and low maintenance requirements. These opportunities include both R&D programs like our current Navy work and over time deliveries of production systems.

Turning to 2027 deployments, we will focus on military deliveries against contracts we already hold, which generates near-term revenue. For data center customers, we plan to deploy 200-kilowatt power modules at their AI test facilities so they can experience the technology firsthand, which we believe is the fastest path to larger volume orders of our multi-megawatt system that we plan to have initially ready in 2028. These early deployments may be structured as outright sales, loaned units, or power purchase agreements.

With this sequencing, we believe it will enable greater revenue in the near term. It will move commercialization of the 200-kilowatt power module into 2027, but we believe it allows us to start deploying units sooner in the data center space and to build a substantially larger long-term customer pipeline. Now switching to scaling and manufacturing. As we discussed previously, one of the key milestones we set for 2026 was to work on improving the speed and throughput of the additive manufacturing process.

I am pleased to report that we are making significant progress. We have determined how to take advantage of the full laser power and capabilities available in the printers, implementing software enhancements that optimize how parts are printed, and are continuing to refine part design, including evaluating material changes where appropriate. Together, we believe these initiatives have the potential to increase overall printer speed and manufacturing throughput by up to 3x, depending on the part.

We have already begun demonstrating some of these speed improvements in prints we are making today. Separately, we recently signed a beta machine agreement with Colibrium Additive, a GE Aerospace company and the manufacturer of the additive printers we use at Hyliion. Through this collaboration, we are working together on the next generation of additive manufacturing systems. Up to this point, we have not yet sized our production capacity expectations because we were still assessing the capabilities of our fleet of 30 printers spanning different generations, sizes, and capabilities.

Now that we have a better assessment of print speed capabilities, we have determined that our existing base of installed printers will be able to produce up to 15 megawatts per year of KARNO Core capacity. With the optimization work I described, we are now in a better position to share high-level expectations on the relationship between KARNO output and printer investment for printers we plan to purchase going forward.

Based on our current manufacturing roadmap, we believe the advancements we are working on will significantly improve the capital efficiency of scaling KARNO production once fully implemented. We currently estimate that approximately $1.5 million worth of investment in printers and related manufacturing equipment can support approximately 1 megawatt of annual KARNO Power Module production capacity.

Based on our current pricing expectations, that 1 megawatt of annual capacity represents approximately $2.5 million to $3 million worth of annual revenue. We believe this ratio of investment to output will provide an attractive return on investment once we reach efficient production volumes. It is still too early to determine exactly when we will achieve that level of production volume, but we now have an additive platform capable of throughput that justifies further investment.

As a note of caution, the approximately 3x improvement in printer speed that I discussed earlier is still being validated and will require additional testing. However, we have gained enough confidence in the improvements we are seeing that we now believe there is an opportunity to accelerate additional printer investment into 2027 that we had previously expected to make in 2028. We look forward to providing additional updates on printer throughput improvements and our manufacturing investment plan in the coming quarters.

Finally, I would like to remind everyone that buyers of the KARNO Power Module are eligible for a 30% investment tax credit on both the purchase price of the system and qualifying investments required for its installation. Under current law, that credit is available to customers through 2036.

Turning to the 2026 performance milestones that we laid out for the year, we checked off 3 more this quarter: new military contracts, printer speed enhancements, and demonstrating multi-power module operations. We are just over halfway through the year and have completed half of the milestones we set. By our next earnings call, we expect to check off at least 2 more, including surpassing our $10 million revenue milestone and completing our initial customer site deployment, with 2 additional milestones expected by year-end.

As I mentioned before, we will be prioritizing military opportunities and delivering 200-kilowatt systems to data center test facilities, which will move commercialization of the 200-kilowatt system into 2027. We believe this plan will drive higher near-term revenue and larger long-term volume opportunities. On our 3-year outlook, the overall trajectory remains the same, although the sequencing of our deployments has changed some.

Our 2027 expectations include commercialization and ramp-up of 200-kilowatt deliveries, the development of a multi-megawatt KARNO system for the U.S. Navy, and the assumption of printer acquisitions that will drive future production growth. For 2028, we plan to accelerate system deliveries and capital investment in growth assets while delivering our first multi-megawatt KARNO system to data center customers.

With that, I'll turn the call over to Jon to walk through the financial results for the quarter.

Thank you, Thomas, and good morning, everyone. In the second quarter, we recorded revenue of $4.9 million from Research and Development services. This compares with revenue of $1.5 million in the second quarter of 2025 and $2.8 million in the first quarter of this year. Revenue growth this year is primarily attributable to increased production of components for the 800-kilowatt power module we are building for the Office of Naval Research, including the 4 KARNO Cores that will power the system and other Power Module components.

Cost of revenue was $4.6 million, resulting in gross profit of $366,000. Operating expenses for the second quarter were $15.7 million, approximately flat compared to the second quarter of 2025. R&D spending in the quarter was $9.5 million, down 6% from $10.1 million a year ago. The year-over-year decrease primarily reflects a shift in spending towards revenue-generating services for the Navy, with the associated costs of that work reflected in cost of revenue. SG&A expenses were $6.4 million in the quarter, up approximately $0.5 million or 8% compared to the second quarter of 2025.

On the powertrain exit and termination expense line, we recorded a credit of $258,000 related to asset sales. Our net loss for the second quarter was $13.9 million, compared with a net loss of $13.4 million in the second quarter of 2025. Turning to our year-to-date results, revenue for the first half of '26 was $7.8 million, up significantly from $2 million in the first half of 2025. Cost of revenues was $7.2 million, resulting in gross profit of $576,000, compared with $143,000 in the same period of 2025.

Operating expenses for the first half were $29.1 million, down 18% compared with the first half of 2025, primarily reflecting a 23% reduction in R&D expenses. Year-to-date net loss was $25.7 million, a 16% improvement compared with the $30.7 million net loss we recorded in the first half of 2025. Turning to our cash and investment position, we spent $6.9 million during the second quarter compared with $13.5 million in the second quarter of 2025 and $13.1 million in the first quarter of this year.

The key drivers of the lower cash spend compared with last year were a lower net loss and lower capital spending. Capital spending was approximately $200,000 in the second quarter and $2.1 million year-to-date. This compares with $11.6 million in the first half of 2025 when we had significantly higher spending on additive printing machines and related investments. Cash generated from asset sales was $1.9 million year-to-date. We finished the second quarter with $132.4 million of cash and short- and long-term investments on our balance sheet.

Next, I'd like to update our outlook for the remainder of the year. Our research and development services work with the Navy has ramped up more quickly this year than we initially expected. We began the year projecting approximately $10 million of R&D services revenue for this year, but in the first half alone, we have recorded nearly $8 million of revenue. The work scope and funding under our existing Navy contracts will begin to wind down during the second half of the year with a significant milestone being the expected completion of the 800-kilowatt KARNO Power Module around the end of the year.

As that work nears completion, we will begin ramping up activity under the new $41.7 million Navy contract we signed a few weeks ago, although most of the work under that contract is expected to be performed in 2027 and 2028. Consequently, we now expect third quarter revenue to be approximately in line with the second quarter at just under $5 million and total revenue for the year to be approximately $15 million. As Thomas noted earlier, this represents a 50% increase from our previous 2026 revenue outlook and compares with total revenue of $3.5 million in 2025.

Turning to capital spending. Over the last couple of years, our investments have been directed primarily toward additive printing machines, related facility investments, and CNC machining equipment. As Thomas noted, this year, our focus has shifted towards optimizing our existing printer fleet and increasing the speed and throughput of those machines. As a result, we expect capital spending in the second half of this year to be approximately in line with the $2 million spent during the first half. This represents a significant decrease from total spending of nearly $24 million in 2025.

We've also previously discussed our expectation to enter into an equipment financing arrangement to monetize a portion of the value of our printer assets through either a sale leaseback or secured debt financing. We continue to expect this to close this year and generate between $10 million and $15 million of cash proceeds. The combination of higher revenue, lower expected capital spending, and the anticipated equipment financing has resulted in a meaningful improvement in our cash outlook for the year.

We previously expected to spend approximately $50 million during 2026, net of the equipment financing and to finish the year with approximately $100 million of cash and investments on our balance sheet. We now expect to finish the year with between $115 million and $120 million, reflecting net cash spending during the year, including equipment financing of approximately $30 million to $35 million. This improved outlook highlights the strength of our balance sheet and our continued focus on the careful deployment of capital as we begin deliveries of KARNO Power Module systems and continue to grow revenue.

Next, as Thomas discussed earlier, we are gaining confidence in the ability of advancements in additive printing technology to deliver greater speed and throughput. This is the outcome we were targeting and a reason we slowed printer acquisitions this year. Previously, we expected the next phase of the capital build-out of our additive printer fleet to begin in 2028. Based on the progress we are seeing, we now see an opportunity to restart printer purchases sometime next year, enabling a faster ramp-up in production capacity.

We expect to be able to finance these growth investments with some combination of leases, debt and cash. As we have stated in the past, we continue to believe the capital we have on hand today is sufficient to carry us through commercialization of the KARNO Power Module. At the same time, additional capital will ultimately be required to support production growth. Finally, as part of our ongoing capital planning process, we are establishing an at-the-market equity program to provide additional financial flexibility.

The program will allow us to raise capital opportunistically when market conditions are favorable and when we believe doing so supports our strategic priorities and long-term shareholder value. We intend to be disciplined in our use of the program, carefully considering both our capital requirements and the potential dilution to shareholders.

Now we'll turn the call back over to Thomas.

To wrap up, the second quarter advanced all 3 of the areas I opened with. We signed our largest military contract to date at $41.7 million and expect to close very near $50 million worth in new military awards this year. We increased our revenue guidance for 2026 by 50% from $10 million to about $15 million. And we unlocked additive manufacturing improvements that we believe support up to a 3x increase in throughput, which will enable us to begin scaling faster to meet the demand we are seeing.

For the remainder of 2026, our focus is on completing the early adopter units, deploying the first systems to customer sites, closing the remaining military contracts, and building the 800-kilowatt Navy system for the USX-1 Defiant autonomous Navy ship. We are excited about the opportunity ahead and about the position we are in to capture it.

I will now hand the call over to the moderator to open up for Q&A.

[Operator Instructions] Your first question comes from the line of Sean Milligan with Needham & Company.

I guess just real quick on the current capacity of the 30 printers. I just want to clarify, did you say 15 megawatts a year or 50 megawatts per year?

Sean, so it was 1-5, 15 megawatts per year. We see that the existing installed base and the printers that we already have on order and expected to come in, that will be able to produce up to 15 megawatts a year. Now with that, that's a rolling in all the advancements that we discussed throughout today's call.

Okay, great. And then on the data center side, I think you said like roughly half of your LOIs are data centers today, but you're engaged with a growing number of customers. Just curious, like how you see that progressing with the customers in your pipeline? Do you see the opportunity to sign additional LOIs or MOUs? Or are you going to move more to like test orders? Just trying to think through potential catalysts for the rest of the pipeline there.

Sure. So maybe just to start. So we have the -- about 750 KARNO Cores worth of LOIs that are executed that represents around $400 million of potential revenue opportunity at current pricing. What I'll say though is, that is only a fraction of the customer engagements and customer interests that we are seeing and engaging in. As we announced on today's call, we're working with multiple hyperscalers as well as numerous additional data center builders. And we're in discussions with them.

And some of the questions we've been getting from them are like, how do you get to 100 megawatts a year of production capacity just for me or one of them is even at the 400 megawatts of production capacity just for me. So they want to understand really how are we going to scale the additive manufacturing side of things.

That's one of the big reasons why we were excited today to be able to share more about what that scaling journey looks like and the capital-efficient path that we have ahead of us. But the demand we're seeing from the data center sector, the military sector, as well as the base commercial customers that we've been working with is phenomenal, frankly. And the big focus right now is how are we going to be able to scale to meet that demand.

Okay, that's really helpful. And kind of along those lines, you gave the numbers, the $1.5 million for printers and manufacturing equipment to support 1 megawatt of KARNO per year, of course, per year. Just curious about the supply chain there. So on the -- especially on the printer side, I guess, like how quickly can you order printers? Can you bring printers in?

Is there any constraints on -- like in terms of the number of printers you could bring in, in a year? Because to your point, like it seems like printers and power space is accelerating, the demand from data centers is accelerating. So just curious about the ability to accelerate your supply chain, especially on the additive manufacturing side?

Sure, so maybe let's start with the infrastructure that we have in place. So in our Austin, Texas facility, the existing square footage that we already have under lease, that has the capability of adding hundreds of additional printers to it. So we have plenty of space there. That's not going to be our issue.

The next question is, what does the supply chain look like in order to be able to source these and get them in? As you're aware, we procure the systems from Colibrium Additive, which is a GE company. The great thing is, these are the same printers that they use in aerospace. They're selling into the healthcare industry and even are being brought into other industries as well. And so these are not unique printers that are just for Hyliion. These are standard printers that they're scaling for other customers as well.

Now, in terms of their ability to deliver on those, they have been in production of additive machines for over a decade. And so it's something that they have confidence, and we've had discussions with them that as our capacity is scaling up, they'll be able to deliver towards them. We do anticipate that it's in the -- we anticipate lead times on printers getting into the low number of quarters, even potentially months of lead time as we go forward, but that's obviously something we'll work closely with GE on.

[Operator Instructions] Your next question comes from the line of Edward Jackson with Northland.

So I wanted to start out and just make sure I understood. You talked about the 10 units that you're going to -- well, roughly 10 units that you're still on track for. And you're going to have a chunk of that being the U.S. Navy, so 800 kilowatts that takes care of 4 of them. And then you said that you were going to have an initial customer site for the data center. Is that for a 200-kilowatt system? Or is that for multiple systems?

And given the fact that it's going in there and they're going to be knocking the ground, I assume there's no product revenue associated with that. So that -- I wanted to make sure that I understood that right. And then am I correct then that the remaining kind of half of the units you're going to be sort of using to do validation work with other data center opportunities. Is that the way to think about those 10 units?

Yes, so a few different parts to the question. So maybe first is the data center opportunity we're targeting for this year. So that will be with a 200-kilowatt system, as you mentioned, so the base smaller enclosure. That customer ultimately has long-term interest in the multi-megawatt product. But the great thing is that technology on the inside of the box is the same. It's more of just a scaling equation.

So that was one of the things we wanted to convey on today's call is over the early adopter units, as well as units that will be getting out into the field next year that are going into data center sites, they'll be taking the 200-kilowatt, and it's really to build confidence in the technology and get to experience it firsthand for a roadmap of multi-megawatt systems going forward. In terms of the revenue on these early systems, so until we get to actual commercialization on the unit, we would not have recognized revenue for them.

However, when we are selling these units to customers, we are charging them for them. It's more from an accounting standpoint that it doesn't go to revenue recognition, which Jon can obviously share more on. But then in terms of the additional units, so the remainder of the 10 are actually all Navy assets. And in terms of the couple of customer units that we have, or a couple of customer sites and then the remainder are going to the Navy.

Okay. With regards to -- I mean, the prior question with regards to printer availability and the quality of getting ramped-in and if it's going to take months or quarters before you can get printers from order. When do you think you'll have your -- finalize your kind of CapEx needs for '27 and be able to highlight that more? I mean, it seems that -- are you expecting to have that done in months or by the end of the year?

Just kind of trying to get a sense of when that spend is going to kick in for next year. I mean, obviously, you're doing it because you're coming up to the point where you want to be, which is selling product. But I just kind of want to get a little better sense in terms of timing because there's obviously a runway before you can get those units in and put into work.

This is Jon. I'll take that one. So while we haven't finalized 2027 capital spending needs yet, we do think that these print speed improvements are going to enable us to start spending again on printer acquisitions. So I think if you do the math on what I projected capital spending this year, it's going to be $4 million compared to like $23 million, I think, it was in '25. So next year, we would start to ramp that back up again. And what's going to dictate that -- the initial step is when you place orders, you put deposits down. There are -- there could be an opportunity to actually get the printers delivered.

Some of it depends on just when some of the technology will be available, but we certainly -- we'll certainly want to send the message that we're going to restart that effort next year and whether the deliveries happen late next year or into '28 or middle of next year is yet to be determined. But I think that the good news of the messaging is that we have an opportunity to pull ahead our ability to bring the printers in that will drive production capacity growth in the future. So I guess stay tuned, but it's certainly going to be closer than it is than we previously had expected.

And then you made a comment, Jon, that you were going to look for some equipment financing and you thought to get $10 million to $15 million in cash proceeds for doing this, this year. And then I thought you made another comment with regards to another $30 million or $35 million in equipment financing...

Yes, yes. Yes, sorry Ed. Yes, let me clarify. So initially, our projections this year were to consume and spend net $50 million. So that would have left us with $100 million of cash and investments at the end of the year. So that included $10 million of equipment financing. So now what we're changing is that our actual cash forecast of spending is improving by $15 million, and I'm also projecting that we may upsize that equipment financing from $10 million to $15 million.

So that's where you get up to a $20 million improvement in our total cash forecast. So that cash forecast was $50 million and now it's somewhere -- of net spending and now it's somewhere between $30 million and $35 million. So I didn't mean to imply that the financing would be in the $30 million to $35 million. The financing will be $10 million to $15 million. Does that make sense?

Yep, it does. Yes.

So again, I just want to reiterate that because of higher revenue, lower spending overall, and lower capital spending, our net cash burn is going to be about $15 million better than we previously expected.

Got it. And then my last question, and I will get out of line, is on the Navy contracts. You've got call it $42 million of revenue, you commented on it being rolling in and starting to contribute in 2027 and 2028. And so when we think about those 2 years, basically, we should think about some -- however the cadence is, putting at least $42 million of developmental revenue into the model across '27 and '28. And then I want to verify with the commentary of the ramp down of the original contract, there will be no revenue from that original contract carrying forward into '27?

Yes, yes. Let me try and unpack that a bit. So we had roughly $20 million of military contracts preceding the one that we just signed for $41.7 million. We've been spending on those contracts since the end of '24. I think we had $1.5 million in '24 and $3.5 million in '25. And then -- and so we've already booked almost $10 million. So that's going to start to wind down by the end of the next quarter, this quarter that we're in now. At the same time, we're going to be ramping up spending on new contracts, including the one we just signed, and then we're anticipating additional contracts to be signed this year as well.

So I think the answer to your last question is yes, old contracts, if I can call them that, will start to wind down this quarter because we'll have spent that money and the new contracts will start up in Q4 of this year. And then those will accelerate in '27 and '28. But I do want to reiterate, we do expect to sign more contracts in '27 and '28. So what we expect to see is layering of incremental government military contracts on top of each other, which will enable steady continuous revenue growth on our R&D services line over time.

And then maybe one other caveat, some of those contracts will turn into more -- it won't be R&D, they'll actually be commercial systems. For example, base power, forward operating base power, those will be R&D initially, but then that will just turn into commercial units of special types for the military. So again, we should see growing revenue over time. So yes, the $41.7 million contract we just signed, the bulk of that will be spent over '27 and '28.

[Operator Instructions] We have reached the end of our Q&A session. I will now turn the call back to Thomas for closing remarks.

Thank you all for joining today's call. As we highlighted, a lot of exciting work happening, customer demand, both with military, commercial customers, specifically the data center front. We're seeing data centers really viewing this as a viable technology to meet this power need that they have, and we're excited to get some of our early units out there into their operations to really showcase the product benefits.

And then coupling to that being able to highlight more on the road map ahead of what scaling additive manufacturing looks like. We believe we, with these breakthroughs have had a very economical way to scale the business going forward and are excited about the state-of-the-art nature of what these machines are looking like. So with that, thank you for joining this quarter, and we look forward to updating further on earnings calls ahead.

This concludes today's call. Thank you for attending. You may now disconnect.

Hello, everyone. Thank you for joining us, and welcome to Hyliion Holdings First Quarter 2026 Earnings Conference Call. [Operator Instructions]. I will now hand the conference over to Greg Standley, Chief Accounting Officer. Please go ahead.

Thank you, and good morning, everyone. Welcome to Hyliion Holdings first quarter 2026 earnings conference call. Joining us today are Thomas Healy, Chief Executive Officer, and Jon Panzer, Chief Financial Officer. A slide presentation accompanying today's call is available on Hyliion's Investor Relations website at investor.hyliion.com.

Please note that during today's call, we will be making certain forward-looking statements regarding the company's business outlook. Forward-looking statements are predictions, projections and other statements about anticipated events that are based on current expectations and assumptions, as such, are subject to risks and uncertainties. Many factors could cause actual results to differ materially from forward-looking statements made on this call. Factors that may cause such differences are discussed in our presentation and press release as well as our filings with the Securities and Exchange Commission. You are cautioned not to place undue reliance on forward-looking statements and we undertake no duty to update this information except as required by applicable law.

With that, I'll turn the call over to Thomas.

Hello, and thank you for joining us for Hyliion's first quarter 2026 earnings call. On our last call, we had said 2026 would be the year we shifted from development to deployment of the KARNO Power Module. The first quarter delivered tangible progress against that plan. We successfully completed the UL Certification non-recurring test milestone for the KARNO Power Module. We signed a new data center partnership with VFG Holdings, broadened our military engagement and further demonstrated multi-fuel flexibility. We also saw a significant increase in revenue growth, recording $2.8 million this quarter, a fourfold increase from the prior quarter. This growth reflects the accelerating pace of work with the military.

Today, I'll walk through each of these areas in more detail, then turn the call over to Jon for the financial update. Starting with UL Certification. This was a high priority milestone for us, and I am pleased to share that we successfully passed the UL Certification nonrecurring tests for the KARNO Power Module. This is the gating item we discussed last quarter and clearing it now enables us to begin delivering early adopter units to customer sites. To frame what this means going forward, each individual power module will still undergo a final operating test prior to receiving its nameplate certification, but the foundational testing we have completed does not need to be repeated. As a reminder, this testing covers the electric motor battery system and the complete power module with each subsystem undergoing separate UL Certification.

As we continue working towards our full 200-kilowatt design power rating, we expect to move to facility-level certification, after which we will no longer need to certify each unit independently. That progression is an important step towards enabling production at scale.

Turning to deployments. We continue to operate KARNO units at our Cincinnati facility while building additional systems. We remain on track to complete the approximately 10 early adopter units this year, ahead of commercialization, which we expect around year-end, depending on the timing of the early deployments. With UL Nonrecurring Testing now complete, we are beginning to work with customers to move these systems to their sites. This is important transition. Until now, customer units have been operating at our facility under controlled conditions. Over the next couple of quarters, we expect these same units to begin operating in real-world customer environments. We continue to see strong interest from the military in deploying our KARNO technology. We are now in active discussions not only with the Navy and Air Force, but with additional branches of the U.S. military as well. This interest is being driven by the platform differentiated capabilities, including true fuel-agnostic operations, low maintenance requirements and low acoustic and thermal signature, which are particularly important for applications such as autonomous operation and mobile power generation.

We continue to expect to sign $40 million to $50 million of additional military contracts this year on top of the approximately $20 million in ONR contracts we are currently executing.

We are also seeing strong and growing demand from data center customers. The need for on-site fuel flexible power generation is becoming increasingly important, and we are actively engaging with the leading players in the market. In that context, I am pleased to announce a new strategic partnership with VFG Holdings. Hyliion and VFG holdings have executed a nonbinding letter of intent, to deploy up to 250 KARNO cores or approximately 50-megawatts of power over the next 5 years. VFG is a developer of advanced next-generation data centers and offers turnkey solutions, including power, infrastructure, compute and financing. The team at VFG is comprised of industry veterans from some of the largest data center companies and is planning multiple gigawatts of power production in the years ahead. Under this partnership, the parties plan to deploy KARNO Power Modules at VFG's data center sites to demonstrate the unique benefits of the KARNO platform, including lower-than-grid electricity production costs true fuel-agnostic operation and direct 800-volt DC integration. We plan to share more details on the partnership in the periods ahead. The LOI between Hyliion and VFG is subject to the execution of a definitive purchase agreement.

Building on our partnership with ABM Industries, which we announced in early 2026, we have been engaging prospective customers alongside ABM. The opportunities span light and heavy commercial applications from single-unit 200-kilowatt deployment to multi-megawatt installations. ABM brings deep capabilities in site engineering, integration, construction and ongoing site management, which complements our focus on advancing and commercializing the KARNO Power Module.

Now shifting to updates on our product development. We continue to make progress towards our full 200-kilowatt design power rating. During the quarter, the team conducted isolated testing on new software and component improvements that yielded additional power and efficiency gains. We plan to incorporate these and other advancements into the product over the coming quarters and remain on track to reach the full design power rating by year-end.

Beyond the broader fuel flexibility we have previously demonstrated, the first quarter included a particularly meaningful product development milestone. We successfully demonstrated dynamic fuel switching within the KARNO reactor across diesel, natural gas and hydrogen without shutting the system down. Said differently, the system can operate on both gaseous and liquid fuels with very different characteristics. This achievement validates true fuel flexibility on the platform, not just dual fuel, but the ability to operate on liquid fuel pipeline gas and zero carbon fuel through a single architecture with the ability to transition seamlessly between them, during operation. The strategic significance of this capability is meaningful as many of our target applications desire this kind of flexibility.

For example, data center operators often prefer pipeline natural gas as their primary fuel source while maintaining diesel fuel on-site as a backup in the event of a disruption in natural gas supply. The KARNO Power Module enables both within a single platform, eliminating the need for separate primary and backup generation systems. This capability also expands our addressable market in defense applications, where mission requirements frequently depend on fuel availability and the ability to adapt in real time. With diesel fuel operation now demonstrated, we have commenced building an 800-kilowatt KARNO Power Module for the U.S. Navy. This system will be deployed on an unmanned Navy vessel as part of our existing ONR program. We expect to complete this build during 2026 alongside other Navy product performance and reliability milestones. Importantly, the 800-kilowatt system we are building for the Navy serves as the same architectural building block for our data center offering. By combining multiple 800-kilowatt units, we can scale to 1.6 megawatts, 2.4 megawatts, 3.2 megawatts and higher aligning with the modular power configurations required by data center customers.

On manufacturing capacity, our focus today remains on building systems, establishing the supply chain to meet our quality requirements and continuing to scale print capacity. We are continuing to make meaningful progress with our additive printers, particularly in improving part production speeds. We expect to install a few additional printers this year, and with those additions, our existing fleet is expected to support our planned production needs for 2026, '27 and into 2028.

We remain on track to take delivery of and begin testing one or more printers equipped with the latest laser technology from GE Colibrium later this year. We believe that technology has the potential to further improve print speed and throughput.

On supply chain, we noted last quarter that magnet supply was a potential risk given export constraints from China. I'm pleased to share that during the first quarter, we began to see progress with alternate sourcing options for the high-strength magnets we require, increasing our confidence in our ability to support planned production.

To recap our 2026 milestones. At the start of the year, we outlined a clear set of objectives. With 1 quarter complete, we are making solid progress. We have completed UL Nonrecurring Testing on the KARNO Power Module. We have demonstrated multi-fuel switching, including liquid fuel operation and have begun building the 800-kilowatt Navy system. Looking ahead, we remain on track to achieve our full 200-kilowatt design power rating, complete our remaining early adopter unit deployments, secure $40 million to $50 million in additional military contracts and deliver approximately $10 million of revenue for this year.

Looking beyond this year, our 3-year outlook remains unchanged. In 2027, we expect to ramp commercial deliveries and expand the range of KARNO deployments, specifically with the military and data center customers. We view 2027 as the year we transition from initial commercialization into meaningful production scale. By 2028 and beyond, we expect to accelerate commercial growth as production capacity enables us to serve a broader portion of customer demand, including expansion into multi-megawatt configurations for data center customers.

With that, I'll turn the call over to Jon to walk through the financial results for the quarter.

Thank you, Thomas, and good morning, everyone. In the first quarter, we recorded revenue of $2.8 million from Research and Development Services. This compares with revenue of about $0.5 million in the first quarter of 2025 and $700,000 last quarter. The significant growth reflects an acceleration of work under our contracts with the Office of Naval Research. Cost of revenues was $2.6 million, resulting in a gross margin gain of $210,000. As a reminder, R&D Services revenue with the Navy reflects the sale of KARNO Cores and Systems including the 800-kilowatt power module we are building and the work we perform to test and validate these units.

Operating expenses for the first quarter were $13.4 million down from $19.7 million in the first quarter of 2025. The decrease was driven primarily by lower research and development spending. R&D spending in the first quarter was $7.7 million down 37% from the $12.2 million we spent a year ago. While the absolute level of R&D spending was down compared to a year ago, most of the year-over-year decrease related to a shift to revenue-generating services for the Navy versus other research and development work. This shift included approximately $1.9 million in lower R&D expenses this quarter as we capitalized inventory and utilize less labor and materials performing R&D activities. This inventory is primarily work-in-process components that we expect to utilize in future periods for building KARNO systems for the Navy.

On the Powertrain Exit and Termination expense line, we recorded a credit of $414,000 compared to an expense of $1.4 million in the first quarter of last year. This credit reflects ongoing asset sales related to our former Powertrain business during the quarter and is not expected to be recurring. SG&A costs were relatively flat with an increase of $100,000 attributable to higher personnel expense, partly offset by lower spending in other areas. Our total net loss in the first quarter was $11.7 million, down 32% from the $17.3 million loss we recorded in the first quarter of 2025.

Turning to our cash and investment position. We spent $13 million during the first quarter. Capital spending was $1.9 million and consisted primarily of payments for additive printing machines, along with some facility investments to support printer operations. Cash generated from asset sales for the quarter was $1.6 million. Asset sales related to the monetization of equipment previously used in our Powertrain division and are largely complete. We finished the first quarter with $139.3 million of cash and short and long-term investments on our balance sheet. We are reaffirming our guidance for 2026, including approximately $10 million in revenue this year from both R&D Services and possibly some Commercial customer sales. Also, as Thomas noted, we plan to slow capital spending in 2026 compared to 2025 as we work to optimize the output of the printers that we have on hand today. We are planning to execute equipment financing for up to $10 million later this year, although that amount may shift up or down based on terms and availability of lease capital.

Overall, for 2026, higher revenue, expense control, lower capital spending and planned equipment financing are expected to result in a lower level of total spending compared to 2025. Our current forecast is for net spending of just over $50 million during the year, resulting in a year-end cash and investment balance of approximately $100 million. We continue to believe that the capital we have on hand today is sufficient to carry us through commercialization of the KARNO Power Module. Also, we anticipate additional capital will eventually be required to support production growth particularly for the purchase of additional additive manufacturing equipment to more rapidly ramp up production in future years.

Now I'll turn the call back over to Thomas.

To wrap up, the first quarter delivered against the deployment and commercialization plan we outlined at the start of the year. We completed UL Nonrecurring Testing, signed a meaningful data center partnership with VFG and generated 4x the revenue of the prior quarter. We also broadened our engagement across the U.S. military, demonstrated true multi-fuel flexibility and commenced the build of our first 800-kilowatt Navy system. The themes we identified last quarter, namely the shift toward 800-volt DC architecture in next-generation AI data centers and the demand for resilient mission-critical power across the U.S. military became more tangible in the first quarter. The VFG partnership positions us within the next-generation data center deployments where 800-volt DC-native operation is a clear architectural advantage, while expanding military engagements reflect strong demand for the platform's differentiated capabilities.

For the remainder of 2026, our focus is execution, delivering the remaining early adopter units, completing the 800-kilowatt Navy system, securing $40 million to $50 million of additional military contracts achieving full 200-kilowatt design power and commercializing the KARNO Power Module by year-end. We are excited about the opportunity ahead and the position we are in to capitalize on it.

I will now hand the call over to our moderator to open up for Q&A.

[Operator Instructions] Your first question comes from Edward Jackson with Northland.

Congratulations on all the progress towards commercialization. You really kind of move a change if you would [indiscernible]. I got a question for you. I wanted to start out on some of the R&D and contracting stuff around the Navy. So you had $12 million. I mean, you have $20 million contract you've got about $12 million of it left. You had a pretty big number for it this quarter. How do we think about how that -- it seems like it's -- from your own conversation like the activity around that contract is accelerating. So can we expect similar kind of numbers in terms of print as you go through the remainder of this year against that contract?

And then kind of tieing into the $40 million to $50 million of additional military opportunity in front of you. What's the timing for you to maybe bring some of that home in terms of contract? And would it be fair to assume that given the progress you've had with the Navy that the ability for you to kind of bring that revenue into the P&L would be a little quicker because you're building on a foundation that's already been in play? That's my first question or questions.

So Ted, this is John. I'll try and answer that. So the first part was just about the $20 million in contracts that we have today and the pace of that revenue. So last year, I think we -- last year and the year before, I think we booked somewhere around $5 million total and so just under $3 million this quarter. It's a good -- it's a really good start to the year, especially in our forecast of $10 million. We do expect that strength to continue and it's really been a matter of us getting a lot of traction with that contract and being worked done, particularly around the 800-kilowatt power module that we talked about.

And then leading into later in the year, we do expect to sign additional contracts, as you mentioned, somewhere between $40 million and $50 million. That will be a second half of the year, the expectation in terms of when those get signed. And so that will allow us to continue that momentum into the fourth quarter and into 2027. So we feel very good about our military opportunities. I think that's one thing Thomas was really trying to emphasize during his prepared remarks today is we do see the military is a huge opportunity. And we are starting to now get some big traction on the revenue side. So you should expect that to continue to grow.

Okay. And then my next question, and then I'll have more but I'll step aside and not be a hog. On the efforts to get to the certification for 200-kilowatts. I know it's kind of like a whack as you kind of go through and make different changes. But maybe you could talk a bit about what you've accomplished towards that end during the first quarter and where we stand in the current second quarter and we need to finish to get to that 200-kilowatt rate and be able to certify.

Happy to share more on that one. So maybe I'll start with where we were last quarter. We had shared that we have produced a gross of 175-kilowatts out of the dyno, and that was about 3-months ago. The focus over this quarter was around actually independently testing some advancements, and in great news, I mean, we were seeing double-digit kilowatts of improvement out of those independent tests that we're running. And so now what we'll work on over the next quarter as well as the quarters ahead by year-end is rolling in those improvements. We've got other ones that we're working on as well that we'll be rolling in.

But ultimately, it's all pointing towards we should be up that 200-kilowatts of power rating by the year-end. One nice thing is with the UL testing that we've done, the nonrecurring testing, even with improved power levels, we do not need to go back and redo any of that testing. So that's a big win. And it's just that that isolated final test that happens on each system, which will have the power rating of it. So it's not like as we increase power, we need to go back and redo all of UL. So that puts us on a good trajectory to be going into next year, starting to scale production, get more units out there and have them be at that higher power level.

Your next question comes from Martin Malloy with Johnson Rice.

Congratulations on your progress. My two questions -- I wanted to talk about the military applications. And if -- with respect to the stationary power military applications, if we could see those -- the testing and move into orders, could that be quicker than what you're seeing with the Navy? And then also on that, are you seeing any change in the interest from the U.S. military following the recent conflict in the Middle East?

Yes. Let me start with the first one on stationary power. So this Office of Navy Research, or Naval Research contract that we have, we often talk about the on-ship platform in this 800-kilowatt. One thing that we probably underemphasized is that contract also has stationary deployments planned into it as well. And so that's something that as we go through this year into next year, we'll be working on those stationary deployments as well.

We also, as Jon highlighted a little bit earlier, the $40 million to $50 million of additional military contracts, that is both for on-ship applications as well as stationary applications as well. So to your point, the stationery does have an accelerated factor to it where it really is just that commercial box that we've already been working on and developed that we'd be deploying. And then the military has plans to put it into hot weather environments, put it into very humid environments. They've got cold weather environments, they're going to be operating it in. So they're really looking to run it through its paces in order to ensure that it's suitable for any type of operation that they might need to be able to be producing power in. So then going into -- the second part of your question.

Yes, so around the conflict and is that accelerating the demand. So what I would say is, I think I would kind of look at it outside of the conflict itself, I would say that the interest from the military is just growing overall. Just over the last quarter, I've had a handful of meetings with various Generals of both -- in the military, I was up at West Point just recently. I was up at [ Fort Hood ] recently. And we're seeing that this product just touches a lot of the pain points they have, whether it's not knowing what fuel is going to be available. And so our system can run on various fuels, whether it be that they need a low heat signature, low noise profile, we deliver that. Also just even the low maintenance, one of the Generals expressed that where the military often struggles is in logistics and moving parts around and getting out to where they're needed. And so if you can move to a lower maintenance system that requires less parts, now you've improved the logistics of the military.

So I think overall, we're seeing very fast growth in interest and demand out of the military. And I suspect that next year, the military will be where we actually focus a lot of our deployments with them.

That's very helpful. And then for a follow-up question, just wanted to focus on the Commercial non-military side and the placement of the KARNO generators this year. Can you maybe help us in how to think about the timetable for testing with the customers, how much time that they'll need before those pilot tests turn into orders? Is it a 3-month testing process or 6 or 9 months with these customers? And this is -- assuming it works as expected.

The numbers you threw out were spot on. So we've got some that have tested, it's in the 3 months. We've got some that have told us it's more than 9-month time frame. But as we highlighted in today's call, we've got nearly 750 units of interest and LOI signed with various customers. And so what I envision happening is that this year we'll get some initial units out with various commercial customers. And then we'll start deploying initial units with other customers as well, and then we'll start layering in repeat orders for those first customers but we'll also be deploying new units with new customers as well. And so I don't think backlog interest, that's not our issue.

Our issue right now is how do we get these units out into the field, how do we start scaling production, and hopefully deliver more units to customers next year, whether that be through repeat orders or through just a brand-new opportunity. And I think next year, we're going to be focusing heavily on both military, as I just mentioned, but then also the data center side of things. And we've been engaging with many of the key players in the data center space, and they're excited and eager to see our product. And -- they're doing trips up to our Cincinnati facility. We've had one this week -- we've got multiple this week. We've got more next week visits happening -- and with various data center players. And so I think that's going to be a long-term growth opportunity for us.

Your next question comes from Sean Milligan with Needham & Company.

Thomas. Great update today. I was -- you hit on the printer side and just the capacity you have for this year and next year. Curious about the human element in terms of the assembly. Kind of what the plans are to scale up there? And do you see any headwinds on the assembly side? I guess personnel and then also maybe like changes you're making to the system to make it easier and quicker to assemble?

Yes. Great question. So we're working on all of the above. So maybe just walking you through the journey. Initially, all the -- very first assemblies were happening and all the very first printing was all happening out of our Cincinnati R&D facility. The first thing that we started to scale was printing, and that's where we set up -- we've probably got about 3x of print capacity in Texas that we do in Cincinnati now, and we moved that to our larger facility here and built up the team around it. Over the last handful of months, we also then moved Power Module assembly to Texas. So that was happening in Ohio. Now it's happening here in Texas. The next stage of the journey, we'll be moving the actual KARNO Core assembly to Texas. That's something that we've already started building out the team. They're actually traveling to Cincinnati, learning how to build the system properly and then we'll move that operation here.

So I think we have a very good setup where we kind of go through the learnings, the hurdles, the development in our Cincinnati facility that allows the engineers to be alongside the assembly line in kind of educating and iterating that process. And then once it's more stable, then we move it to Texas, and it's more of a focus of let's start scaling and growing it. And to that point, as I mentioned, we've already started building out the team to work on that. So do feel confident that the facility has the space and the infrastructure we need in order to scale and we'll be hiring the team as we need it appropriately and then we'll also be adding in parallel assembly lines as well, which is something we've already started exploring.

Awesome. And then on the -- you mentioned -- you're having like a lot of conversations with data center customers today. Curious kind of what they see as the most unique or like in-demand piece? Is it the 800-volt architecture? Is it the fuel agnostic? Like just curious about what do you think they view as like the differentiator for Hyliion?

Yes. So first is they just need power. So I think any data center provider you talk to, if you had even an old used turbine [indiscernible], they would buy it right now. It is unbelievable the need for power that they have. And

then outside of that, I mean that's actually one of the things that they wish we had more of. They wanted -- they would love to have more capacity out of us. Obviously, we're just scaling up, as we've talked about. But -- so what they're seeing is unique about the KARNO Power Module is, one, it truly enables behind the meter or on-site power generation. It's got very high efficiency for its form factor as well as it's got very low maintenance. So those are exciting factors. And then the other things that you touched on, let's talk about fuel agnostic at first.

So pipeline natural gas is going to be what data centers run on 99% of the time, but the reliability standards that the data centers are held to pipeline natural gas doesn't actually meet those reliability metrics. So that then forces the data center to have fuel stored on-site as well, and normally in most data centers, the on-site fuel storage is diesel. And so what data centers are being forced to do right now is not only by a set of natural gas generators. They also then have to go by a set of diesel generators to be able to handle both those requirements. With the carnal power modules, you can run both of those fuels off of the same power module and have redundancy built in. So that's a big win.

And then the last one that you highlighted, the 800-volt architecture. So just coincidentally, I had the opportunity to meet both Jensen Huang from NVIDIA this past weekend as well as Lip-Bu Tan, the CEO of Intel. And that was discussions I had with both of them. They see 800-volt architecture as where the industry is heading. NVIDIA has been probably the most kind of progressive on that end of saying that this is where data centers are going to head. And so right now, you would normally have to convert the 480-volt AC power from the grid, convert that into DC to go into the racks, versus with the KARNO Power Module, you can go straight 800-volt DC power out straight into the actual rack. So that's going to reduce your need for transformers. It's going to reduce your need for copper lines in the facility. And so it's just all around a better solution, a more efficient solution.

Your next question comes from Edward Jackson with Northland.

Thomas one way to talk about your pipeline and your growth prospects is you know that 1 year ago, your LOI covered 100 KARNOs. And now it's at 750. So that's in 12 months. A stat you might have missed yourself. My follow-up questions for you -- My follow-up question for you is as you're moving towards commercialization and the data center opportunity, and we're talking roughly 10 units, you had indicated earlier in the year that because of the data center opportunity and the evaluations of the trials that you were really going to hold back, call it, 2 or 3 units and just maybe be able to use them for a demonstration with regards to this data center opportunity. You assigned a contract with VFG or an LOI with VFG, which really expanded your pipeline in terms of our backlog or whatever you want to call it, in terms of KARNO units that are under LOI. Is there any time between those two? Or are they mutually exclusive? And I guess, where I'm going with it is, given the relationship with VFG and what they do, I mean, is that where you are looking at perhaps putting some of these units where you're wanting to have them for data center or potential data center customers to kind of kick around? That's my first question.

Yes. So I do see some overlap with those. What we're envisioning with these couple of demonstration units we're actually already in the process of one them is going to be mounted on a trailer. It can be moved around from location to location, what -- what we've seen is when people actually get a chance to see the product, I'll kind of try to put a story to it like up in our Cincinnati facility, you walk out of the building, you can see the KARNOs there. You're maybe 30 feet away, you can't tell whether the product is on or off because of how quiet it is, right? It's not until you actually get closer to the system than you can actually hear some of the fans running.

And that's not the experience that people normally have with an internal combustion engine or turbine. Data centers are actually in the problems of neighbors complaining about how loud systems are. And so when people get to experience the product, see it firsthand, see the -- how easy it is to operate and how simple it is, you truly just feed fuel in and you're going to get electrons out. The product in a lot of ways, sells itself. And so from that end, whether it's with VFG or other data centers, just yesterday, another data center provider was in and they were talking about let's take a system, put it in their like AI test facility. All these data center providers have not only the true data centers, the building, they also have facilities where they're testing new technologies, new architectures like the 800-volt. And so they want to get systems deployed and really prove out the use cases we're talking about.

So as we go into '27, whether it's with VFG and/or others, the plan will be to be showcasing some of those demonstrations.

Having been to your facilities, I stood and had a full conversation with you not like -- literally right next to a unit, and we didn't have to yell. So I mean they are definitely -- they aren't that loud.

The next question is, you're talking about '27 being the year of kind of ramp and '28 and that you have capacity to handle what you view as your needs until you kind of get through '28. Can you perhaps give us some kind of sense with regards to what you think your current capacity level is -- or there some kind of range in terms of kind of what your capacity level will be in '27? I mean, I assume we're talking units and just -- it doesn't have to be exact, I mean, maybe kind of like a lower to high just kind of frame out what we could think of in terms of your ability to produce KARNO?

Yes. So I'll start with apologizing. We're not ready to share that just yet. We do anticipate later this year, we'll start leaning in and share more on what we expect capacity for next year and the year after to be. But we obviously are expecting growth behind it. The reason that we don't want to lean in just yet and share those expected numbers is because we want to see how these initial units go when we get them out to customer sites.

I mean just being transparent, it would be naive for us to think that there won't be any learning that it will just be full steam ahead, right? I mean there definitely will be learnings. Every new technology has it. And the question is going to be what are they and how fast can we respond and adapt and fix them. So our focus right now, let's get these approximately 10 early adopter units out there. Depending on how those are performing, that's what will then set what we expect production to be next year. But obviously, our goal is to accelerate it and get a good bit more units out there next year and so on in 2028 we're just not ready to actually put in numbers to that yet.

Well ahead. My last question is when you think about your revenue, there's kind of -- right now, the revenue that you're putting out there is really coming -- it's coming really from like development work for the military. At what point -- and I know that there's actually KARNO units kind of within that. Is there -- will that always be will your revenue from the military as you move forward, always kind of be in that sort of other revenue bucket? Or as you hit commercialization will we see another revenue line that will have revenue for your commercial customers? And will your military revenue or at least some portion of it is shift into that line?

Yes, Ted, this is John. I'll take a shot at that. So I think it's more of the latter. Our initial contract with the Navy is more R&D. And if you can imagine, a new ship platform they're going to do a lot of R&D work before they put a new power source in that. But then as Thomas was talking about, there's a lot of interest for mobile power for example, in the Army and the Air Force and those types of systems, they could initially be R&D, but those will certainly turn into some kind of commercial units. So it won't all be R&D. I think from our standpoint, at least now, we're happy to have revenue in both lines. It is beneficial both ways and is a good signal of interest -- in the -- in the system and also us making progress in offering these types of R&D services to the military. So it should continue to grow in both areas.

And Ted, just to add to that one. So last week, we had a gentleman in from the military, and it's still on our whiteboard here in the conference room that we're all sitting in. He walked us through the how do you move from the R&D phase of the military into actually just being purchase orders that are being placed and how to effectively do that. So it's obviously something we're thinking about. And do expect that it's long term, this isn't just R&D. This is just we want it to be a product that they just purchase.

[Operator Instructions] Your next question comes from Martin Malloy with Johnson Rice.

I just had one quick follow-up question. Just regarding the UL testing, does that clear the way for what you've done so far, does that clear the way for placing the KARNO units at customer sites? Or is there some final certification that is needed.

Yes. So we've done all the nonrecurring tests and just a little more detail that's both on the Linear Electric Motor, the Battery pack as well as then the full Power Module. The last thing that we need to do in order to put it at a customer site is just go through a final run of the system, and then we'll then put the nameplate on it for UL certification. So the way that you UL structures this is until you get facility-level certification, every unit you produce still needs to go through a final end-of-line run. It does need to redo all those recurring tests I mentioned. It's the final end-of-line run, and then that will get the nameplate once later this year, we move into the facility level certification, then we no longer have to independently run every single system prior to nameplate, being put on it. So the way for you really look at this we've hit those UL tests that were needed in order -- or we've successfully passed those UL tests that were needed, in order to move to the next phase of customer sites. And then each one that we built will just go through a final run prior to that nameplate being put on them.

We have reached the end of the Q&A session. I will now turn the call back to Thomas for closing remarks.

Thank you all for joining today's call. Hope you can see a lot of excitement on our end, a huge milestone for us to get through that UL Certification testing. It's been something we've been talking about for a few quarters now. Obviously, excited about the collaboration with VFG and moving the signed LOIs, nonbinding LOIs to a potential of about $400 million of revenue at today's current pricing. And then we talked about a lot as well the opportunities ahead of us with the military and excited to get those $40 million to $50 million of contracts signed and executed. And and then be able to share a little bit more of what some of those opportunities are with all of you. So once again, thanks for joining for this quarter, and we look forward to sharing more on the next earnings call.

This concludes today's call. Thank you for attending. You may now disconnect.

Hello, everyone. Thank you for joining us and welcome to the Hyliion Holdings Fourth Quarter 2025 Earnings Call.

[Operator Instructions]

I will now hand the call over to Greg Standley, Chief Accounting Officer. Please go ahead.

Thank you, and good morning, everyone. Welcome to Hyliion Holdings Fourth Quarter 2025 Earnings Conference Call. On today's call are Thomas Healy, our Chief Executive Officer; and Jon Panzer, our Chief Financial Officer. A slide presentation accompanying today's call is available on Hyliion's Investor Relations website at investors.hyliion.com.

Please note that during today's call, we will be making certain forward-looking statements regarding the company's business outlook. Forward-looking statements are predictions, projections and other statements about anticipated events that are based on current expectations and assumptions. As such, are subject to risks and uncertainties. Many factors could cause actual results to differ materially from forward-looking statements made on this call.

For more information on both factors that may cause the company's results to differ materially from such forward-looking statements, please refer to our presentation and press release as well as our filings with the Securities and Exchange Commission. You are cautioned not to place undue reliance on forward-looking statements. and we undertake no duty to update this information unless required by applicable law. With that, I will now turn the call over to Thomas.

Hello, and thank you for joining us for Hyliion's Fourth Quarter and Full Year 2025.

[ Technical Difficulties ]

Ladies and gentlemen, we are currently experiencing technical difficulties. Please stand by as we resolve the issue.

Hello, everyone, and apologies for the technical difficulties there. Thank you for joining us for Hyliion's Fourth Quarter and Full Year 2025 earnings call. Heading into 2026, we are positioned strongly to deploy more early adopter units and move towards commercialization. As we shared on our last call, the Kernel power module is now performing at a level that meets our initial customer needs. On today's call, we'll provide more details on our early customer deployment plans and cover the current status of UL certification and product performance demand we're seeing across commercial and military markets and how we are preparing to scale production to support growth.

Turning first to UL certification. We made significant progress during the quarter and are now nearing completion of this important milestone. To provide additional context, UL certification for the carnal power module occurs at 3 levels: the linear electric motor the battery pack and the full power module. I'm pleased to share that we have successfully completed UL testing for both the linear electric motor and the battery pack meaning we have completed 2 of the 3 certifications we need. We have completed our initial round of testing on the full power module. Through that process, we identified several small refinements, including gasket updates to further strengthen water ingress protection and the opportunity to incorporate recent power output improvements.

With those enhancements now underway, we plan to begin our next round of UL testing shortly and expect to complete the certification in the second quarter. Overall, we are very encouraged by the progress and view UL certification for early adopter units as a near-term gating item towards delivering units to customer sites. Beyond certification progress, we now have 5 [ KARNO ] units at our facility, 2 development units and 3 early adopter customer units. These systems are being continuously exercised through a range of load profiles, extended duration testing, customer-specific operation scenarios and military representative applications.

As we shared last quarter, we achieved over 150 kilowatts of power generation, which is sufficient for initial customer deployments. We have since demonstrated 175 kilowatts of power production in testing following recent upgrades and we expect to reach the full 200-kilowatt design power rating by year-end as we transition into commercialization. These improvements include refinements to the piston design and updates to cylinder liner material to enhance heat retention within the system. And importantly, we do not believe reaching the full rating will require any fundamental architecture changes, but rather a series of incremental optimizations across the platform. The steady progress we have made in the past couple of quarters, reinforces our confidence in achieving the final 200-kilowatt design specifications this year.

One thing to note, we are still experiencing some technical difficulties, and we would ask that you please follow along on the slides that are posted on our website.

Continuing on Fuel flexibility continues to be a key differentiator of the KARNO system, and we made meaningful progress in this area during the quarter. We successfully demonstrated dynamic fuel switching with the carnal power module transitioning between natural gas and propane. The system can automatically switch between fuels without shutting down and without any user input to indicate which type of fuel is being utilized. We simply changed the incoming fuel supply and the unit continues to operate as designed, truly highlighting our unique fuel-agnostic capabilities. Another recent accomplishment was successfully running a KARNO core on diesel fuel and being able to export power to the grid while meeting Tier 4 final emissions requirements without the use of exhaust aftertreatment systems. We expect to begin incorporating diesel capability into customer-deployed systems this year. Diesel capability is particularly important for defense applications where it remains the dominant fuel source.

More broadly, the majority of installed generators worldwide operate on diesel, which significantly expands the KARNO power modules addressable market in mission-critical and prime power applications. For example, many data centers prefer pipeline natural gas for prime power but still require on-site diesel for resiliency. We believe our ability to operate on both fuels will allow customers to avoid purchasing separate natural gas and diesel generators and instead rely on a single flexible platform for both primary and backup operations. Next, I'd like to share our outlook for 2026, which can be seen on Slide 5. This includes our deployment plans, product development priorities, business development activity and manufacturing capability. We are entering 2026 with strong demand across commercial, data center and military markets. As previously shared, we have nearly 500 units under nonbinding letters of intent. In addition to a broader set of customers actively evaluating the platform. The primary focus now is transitioning from development into real-world field deployments and moving towards commercialization and scaling.

From a development perspective, our overall plan remains consistent with what we have previously communicated. We expect to deploy approximately 10 early adopter units prior to commercialization including the 2 units delivered to the Navy last year and a third unit that we recently completed and that will go to a customer site following UL certification and product validation. Throughout 2026, we plan to deploy these systems into customers' environments and commercialization to follow later this year. One area where we are seeing particularly strong long-term interest is the data center market. With recent industry announcements pointing towards a shift to 800-volt TC architectures for next-generation AI facilities. We believe our KARNO technology is uniquely well positioned. Our platform already operates at 800 volts DC, which aligns directly with this emerging standard and has the potential to reduce conversion stages lower equipment requirements, improve overall system reliability and efficiency and simplified site electrical architectures.

In 2026, we plan to demonstrate this capability in live environments to showcase this potential. Early this year, we successfully demonstrated a mission representative maybe load profile on a Navy owned KARNO asset. The system managed rapid load changes [indiscernible] in performance under high stress operating conditions, reinforcing its suitability for shipboard and defense use. This was an important validation milestone and has accelerated discussions around additional defense platforms and with NASA, who is exploring coupling our KARNO technology with nuclear power generation. We have identified several near-term opportunities across multiple branches of the military that are moving towards potential contract awards, which we expect to finalize this year. We believe these opportunities could represent $40 million to $50 million worth of new revenue opportunities on top of the approximately $20 million worth of contracts with O&R that we are currently executing on today.

As part of our current Navy program, we plan to deliver additional KARNO power modules and cores in 2026 for specialized shipboard testing. Once completed, these deliveries will represent about half of our early adopter units. One deliverable will be a multi KARNO power module for the ship demonstrating our ability to create high-power systems through coupling our 200-kilowatt cores together. Building on our work with the Navy, I'd like to provide an update on plans for a previously discussed 2-megawatt KARNO power module, which we believe aligns well with the needs for data center and other high-power applications. This can be found on Slide 6. In 2025, we spent significant time developing our multi-KARNO power module configuration for shipboard use. Through that effort, we identified substantial overlap between the Navy architecture and what is required for higher power commercial deployments. As a result, we have leveraged that design foundation to advance a modular, scalable configuration suitable for data center applications.

Our initial concept was a 2-megawatt system comprised of 10 200-kilowatt KARNO cores integrated into a compact footprint, roughly the size of a 20-foot shipping container. Recent customer discussions have led us to evolve the configuration into a more flexible architecture. The system is designed to scale in approximately 800-kilowatt increments, allowing configurations, such as 800 kilowatts, 1.6 megawatts, 2.4 megawatts, 3.2 megawatts and so on. This modular approach aligns closely with feedback from data center customers where power requirements vary depending on site electrical architectures. By enabling expansion through the addition of [ core ] sets, we can tailor output to specific customer needs while maintaining high power density and resiliency. Now turning to Slide 7. Moving on to our commercial customer deployments. We recently entered into a strategic partnership with ABM Industries to support the deployment of integrated distributed energy solutions. This collaboration, combined with KARNO technology with ABM site engineering, integration and operational capabilities, helping simplify deployment and broader customer access across commercial, industrial, data center and mission-critical applications. ABM is also equipped to offer energy as a service contracts with their customers. This partnership allows us to remain focused on advancing and commercializing the KARNO platform while leveraging an experienced partner to support end-to-end customer solutions.

Now turning to Slide 8, where I'd like to provide a brief update on our manufacturing readiness. Today, we operate more than 30 additive manufacturing machines the majority of which are located at our Austin facility. These printers span 3 different machine models with configurations that are able to produce 1, 2 or 4 parts at once. We expect to take delivery of several additional printers this year that we've had on order from last year. With these additions, we believe our additive manufacturing capability or capacity will be well positioned to meet planned production needs for 2026 and 2027, while providing a strong foundation for scaling further into 2028. Our current focus is on maximizing the speed, power and productivity of each machine.

During the first quarter of 2026, we initiated efforts to improve printer throughput and we'll be dedicating both printer time and engineering resources to these optimization initiatives throughout the year. In parallel, we plan to take delivery of and begin testing one or more printers equipped with the latest laser technology from GE Calibrium, which we believe has the potential to further improve print speed and efficiency. On our last earnings call, we discussed the potential risk related to magnet supply, particularly given the export constraints from China. We are pleased to share that we have made meaningful progress in mitigating that risk and have already begun receiving components. While this does not fully eliminate supply chain risk, it substantially reduces our exposure and improves our confidence in supporting planned production.

Turning to Slide 9. To wrap up, I'd like to share our key milestones for 2026, which are summarized on the slide and highlight the achievements we expect to deliver over the course of the year. We began the year with an early win by successfully operating the KARNO core on liquid fuel and meeting emissions requirements. Looking ahead, we expect to achieve UL certification for the early adopter carnal power modules during the second quarter of this year, which will enable broader customer site deployments. In parallel, we expect to achieve the full 200-kilowatt design power by the end of 2026. We also plan to complete the remaining early adopter units during the year. These deployments are an important step towards validating system performance across real-world applications, and will support our plan to commercialize the 200-kilowatt KARNO power module in late '26 which will then allow us to begin scaling production.

In addition, we expect to complete a multi-canal core platform featuring the systems and controls required to coordinate multiple units operating in tandem. This configuration will serve as a stepping stone towards a larger multi-megawatt KARNO system designed to support data center and other customers with higher power requirements. We expect to secure additional U.S. military contracts with a total revenue opportunity of $40 million to $50 million, further advancing the development work underway to support autonomous navy vessels and other mission-critical defense applications. Taken together, these milestones support our expectation of generating approximately $10 million worth of revenue during 2026 from a combination of commercial customer activity and R&D service contracts.

Turning to Slide 10. Looking ahead over the next 3 years, we plan to build on the progress to achieve in 2026 as the KARNO Power module transitions from early deployment into scaled commercialization. In 2027, we expect to ramp commercial deliveries and expand the range of applications where KARNO is deployed. This includes advancing the development of our multi-megawatt KARNO power module for data center applications. We view 2027 as the year will transition from initial commercialization into meaningful production scale. In parallel, we plan to expand our additive manufacturing capability or in capacity in preparation for anticipated growth in 2028. By '28 and beyond, we expect to accelerate commercial growth as increased production capacity enables us to address a broader portion of customer demand. This includes fulfilling interest in multi-megawatt systems for data centers as well as continued expansion within military applications.

Turning to Slide 11. To wrap up, '25 was a year focused on resolving product and production challenges, strengthening the core architecture of the KARNO power module and expanding its operating capabilities across fuel and mission-critical profiles while also improving performance. In 2026, our focus shifts from development to deployment, and commercialization. With UL certification nearing completion, early adopter units moving into the field, growing military engagement and strong data center interest, we believe we are well positioned to transition from validation to scaled execution over the coming years. With that, I'll turn the call over to Jon to walk through the financial update, starting on Slide 12.

Thank you, Thomas, and good morning, everyone. In the fourth quarter, we recorded revenue of $700,000 from research and development services related to our contracts with the Office of Naval Research. Cost of sales was $600,000, resulting in a small gross margin gain. In the fourth quarter of '24, we recorded $1.5 million of R&D revenue and a $100,000 gross margin gain. As a reminder, R&D services reflect -- revenue reflects the sale of KARNO Cores and related components to the U.S. Navy and the work we perform to test and validate these units.

Total operating expenses for the fourth quarter were $15 million down from $17.2 million in the fourth quarter of '24. The decrease was driven by lower R&D and SG&A costs as well as a $500,000 gain from asset sales in connection with the powertrain exit and termination. R&D work continued at a strong pace in the quarter but was lower than '24 when we were purchasing components at a faster pace. SG&A expenses were down about 6% compared to the fourth quarter of 2024, primarily due to lower facilities and insurance costs, partially offset by a small increase in labor costs. Our total net loss for the fourth quarter was $13.2 million, down from $14.4 million in the fourth quarter of '24. For the full year 2025, we reported revenue of $3.5 million, all from R&D services and gross profit of $170,000. This compares with revenue of $1.5 million and gross profit of $100,000 in 2024.

Full year operating expenses were $65.7 million compared to $64.4 million for all of '24. The small increase compared to '24 is related to higher R&D expenses this year partially offset by lower SG&A and powertrain exit and termination expenses. Our full year net loss was $57.2 million compared to $52 million in 2024. Turning to Slide 13. Our cash and investment position, we spent $12.4 million during the fourth quarter and $67.4 million for all of 2025. Full year capital spending was $23.7 million and consisted primarily of additive printing machines and related equipment, along with facility investments to support printer operations. Cash generated from asset sales for the full year was $2.2 million. As a reminder, asset sales related to the monetization of equipment previously used in our Powertrain division and will continue into 2026.

We finished the fourth quarter with $152.4 million of cash and short- and long-term investments on our balance sheet. While this outcome was a little lower than the $155 million we projected for the end of the year, we did end up deferring $10 million of plant equipment financing into 2026. Excluding that deferral, our year-end cash and investment balance would have been about $7.5 million higher than projected, mostly due to lower capital spending and lower operating expenses. Looking forward into 2026, as Thomas noted earlier, we expect to generate approximately $10 million in revenue this year from both R&D services and commercial customers. Commercialization of the KARNO power module is expected to occur late in the year. Also, as Thomas mentioned, we plan to slow capital spending in 2026 as we work to optimize the output of the printers we have on hand today.

We are planning to execute equipment financing for up to $10 million this year, although that amount may shift up or down based on actual capital expenditures and available capital. Overall, for 2026, higher revenue, thoughtful expense control, lower capital spending and planned equipment financing are expected to result in a lower level of total spending compared to 2025. Our current forecast is for net spending of just over $50 million during the year resulting in a year-end cash and investment balance of approximately $100 million. On past calls, we've consistently noted that we expect that the capital we have on hand today will be sufficient to carry us through commercialization of the KARNO Power module. Based on our current plans, that continues to be the case.

Last quarter, we noted that we anticipate that additional capital will eventually be required to support production growth, particularly for the purchase of additional additive manufacturing equipment. In anticipation of that eventuality, we have filed a standard S3 shelf registration statement with the SEC to provide us with additional capital raising flexibility in the future. An S3 is a valuable tool commonly used by established public companies to efficiently and opportunistically raise funds from time to time through the issuance of debt or equity. Now I'll turn the call back over to Thomas.

Thank you, Jon. As we look ahead, our focus for 2026 is clear. We'll be building units and getting them into the field and expanding real-world customer operating experience. We believe this year we'll be defining for Hyliion as customers begin operating the platform in live environments, and we transition from development to scaled execution. Beyond initial deployments, we see significant opportunity in both data center infrastructure and military applications. The shift towards an 800-volt DC architecture in next-generation AI data centers aligns directly with KARNO's design and we believe our modular and fuel flexible platform is well positioned to position this evolution. At the same time, resilient and mission-critical power remains a priority for the U.S. military.

We are excited by the opportunity ahead and our focus on execution as we move into the next phase of growth. I will now hand it over to the moderator to open up for Q&A.

[Operator Instructions]

Our first question comes from the line of Martin Malloy with Johnson Rice.

Congratulations on all the progress you're making. Wanted to ask about the commercialization later this year. And you mentioned, I think, half the units will be going to the U.S. Navy. Could you talk maybe about the end markets for the other units and where you're seeing customer interest from? You mentioned data centers. Is that where the other half of the units are going?

Sure. Appreciate the question. So military is a huge focus for us this year. We're at the point where we're actually starting to put the systems together, that will go into that unmanned autonomous ships. So that's a very exciting one for us this year. Other applications include just prime power. So think about like powering facilities, providing power for EV charging, those sort of opportunities. And then the last one is that data center side of things. So ultimately, data centers are looking for that larger platform that 2.4, 3.2-megawatt system. But one of the focuses for this year is we can really showcase and provide the benefits to data centers even on a 200-kilowatt system.

And so one of the things that we decided to do is take some of these early adopter units and actually use them as mobile units that we can showcase in various applications. And so one of the ones we do plan on showcasing this year is a data center opportunity.

And just for a follow-up question. Could you maybe talk about the capacity that you'll have as you exit this year in terms of manufacturing on a megawatt annual basis or give us some measures what it would cost to add additional megawatts per year in manufacturing capacity so we can get a feel for how '27 might look?

So at this stage, we haven't shared exact specifics around what you're asking. Reason being is we really want to showcase improve this printer speed improvements that we are highlighting. So that's going to be a big focus for this year. Just to give a little background there, as you buy an additive machine and you start utilizing it, there are levers you can pull like how much power in the lasers, how many parts you're producing on a build plate, things like that to increase the throughput. And so that's going to be the focus for this year. And then with that, that will give us a stronger clarity on exactly how much capacity we have with the existing installed base.

But all that being said, we are highly confident that we've the strong production capacity for '26 and '27. And then as we look at 2028, that's where we envision we'll really need to start adding additional printers to start scaling capacity. So I know that does not give you the exact color you were looking for, but hopefully, you understand why we want to focus on those printer speed improvements and proving that we can get to the throughput that we anticipate.

Our next question comes from Sean Milligan with Needham & Co.

When you talk about the incremental $40 million to $50 million potential from the military, I'm trying to understand, is that -- like do you consider that skilled deployments? Or is that like additional testing across different platforms and different agencies, things like that, that would then scale -- could scale beyond that $40 million to $50 million range?

Yes. So it is really focused on additional applications and more unique development for the military and so right now, we have our Navy work we're doing. It's really focused around being the prime power on a fully unmanned autonomous ship platform. We are also looking at stationary power deployment. So think about like powering a base doing that with the Navy. These additional contracts are both a continuation of that work plus adding actually new applications, new opportunities. So think about like different ship platforms, think about other branches of the military, how they need prime power solutions as well and starting to expand the use cases of the KARNO technology. It's not so much just placing a standard PO to just buy additional equipment. But obviously, this work we're doing with the military is in an effort to go to that stage.

Great. And then on this new 800-kilowatt module that you're talking about. I think historically, you gave some numbers around KPIs for the KARNO versus like fuel cells or traditional gas I was curious if there's any update in terms of like how this module might change that? Does it look more attractive for you as you scale up to the larger platforms here? And then what kind of risks are there scaling up from the 200 to the 800 and then the 3.2 million ultimately, what do you need to work through on your side to prove that up and get that done?

All right. So a few pieces to that question. So first is around kind of how it compares against other solutions in the market. So we see customers looking at kind of the conventional internal combustion engines, looking at fuel cell and then looking at technology like ours. In terms of how it shapes up, and I'll put some broad numbers to it, is with a normal internal combustion engine, cost per kilowatt. And maybe I'll use -- if you look at our 200-kilowatt system, what the end numbers would be, if you bought a 200-kilowatt internal combustion engine really designed for prime power and natural gas solutions. You're probably going to be approaching that $250,000, $300,000 for that actual solution.

Then as you look at our system, we're more around $0.5 million. As you look at fuel cells, you're up closer to around that $700,000. So it really still puts us in the middle of those 2, which is what we -- where we're targeting to be. And then obviously, our system has benefits over those other solutions that really drive the payback, the return on investment. So that's improved efficiency, so you're actually going to use less fuel. And then on the maintenance side, having lower maintenance as well as what we're anticipating, which will then help you with driving a faster payback and return on your investment. So that's kind of how we shape up against competition.

Now another one of the benefits of this modular platform is it allows customers to really match the power they need with the amount of capacity we're giving them. So different data centers have different architectures, and this 800-kilowatt system allows that flexibility. In terms of the risk you asked about of going from 200 to 800 to then the 3.2. So the key thing here is the actual power generation unit that 200-kilowatt KARNO core remains the same in every size iteration. And so once we validate and have confidence in that 200-kilowatt architecture, you're really just replicating it. I would equate it to think of like going and buying an electric vehicle, you can buy ones that have different ranges. At the end of the day, they're just putting more of the same batteries in the vehicle to get you longer range.

We're taking the same approach to power generation, where we're putting more of the same KARNO cores together to give you more capacity. So in '25, we actually spent a lot of time, a lot of effort on developing this 800-kilowatt system because that is what's being utilized in the -- for the Navy. And then from there, what we said is, well, as opposed to going and doing a reengineering and doing a 2-megawatt system, why don't we just take that 800-kilowatt and make that modular so that those can be stacked together. So we're already well underway with the development of that system actually.

And then on the ABM [indiscernible] partnership, can you just give us a little more background on ABM and maybe like their expertise in terms of power deployment and like what end markets they are stronger in?

Absolutely. So ABM has a lot of breadth and experience in this space in power generation. And their skills are widespread, everything from -- they've got a strong customer base. Customers are already working with that are coming to them saying, "Hey, we need additional power. We want to need to scale and expand -- to then they can actually do the site engineering. So look at how do you integrate the actual power unit, then they can actually do that integration work, the actual deployment and then they're set up to also do the long-term continuing service and upkeep of that site as well.

And in addition to all that, they also offer energy as a service solutions to their own customer, which for background there, that means that ABM would actually come in and procure and buy the technology, the solutions and then they would charge the end customer based on a cost per kilowatt type of a purchasing agreement to make it easier for customers to adopt solutions for power gen. So ABM very well versed in this space, one of the large players there. And then there focuses range everywhere from things like airports to data centers, to mission-critical applications. So we've already started discussions with ABM and end customers together, and we're excited to see where that will take us here in '26 and beyond.

Our next question comes from Ted Jackson with Northland.

So I've got a few questions for you. Let's start with just making sure I understand kind of CapEx and capacity. So right now, exiting, say, exiting '25 at 30 printers. And if I understand then, you're going to add 1 or 2 more during 2016 and rather than expand the fleet the effort is really focused on using it better. But when we exit this year, roughly speaking, you'll be at 32 units, but you'll be running them at a much higher throughput rate. Is that the message?

Yes. Ted, I think you've got to summarize accurately. We don't know exactly how many we'll have maybe a handful this year. So I wouldn't say exactly 2, it might be a little more than that. But those are units we had on order from last year. But you're thinking about it correctly. We've got a certain amount of capacity, and that capacity rate has been growing as we get more of these printers commissioned and operating to their optimum state. As Thomas has mentioned earlier, some of these are the latest generation printers, which have higher laser power than what we're used to.

So the work that we're going to do to increase print speed this year will be related to taking advantage of that higher power and also just other programming opportunities that just make the printers run faster.

So that's a lot of -- the good news is that's a great increase in output with not a whole lot of investment in terms of dollars. So that's what our focus will be on this year. So yes, we'll end in 2026 with a handful more printers and more throughput from the ones that we have.

And then the range of your printers is from there's the older ones with one laser, and then you have some with 2 and then you've had some with 4. And these -- the we'd be purchasing will be more of these of the 4 laser printers. Is that correct?

Actually, it's a mix. It's a mix. So Thomas also mentioned that GE is working on some new laser technology, and we're excited to help them out on development of that or trying to test out what that's capable of and that will actually be one of the printers with the smaller number of plates. So we've really got opportunities across the range of our printer fleet.

And then, I mean, I know you probably won't tell me this, but do you kind of get a sense in terms of just sort of understanding as you roll through this and then you start your capacity expansion and the CapEx. I mean, roughly speaking, for like the ideal printer whatever one it might be, what's the outlay for a [indiscernible] unit?

The cost -- what are you asking? For a printer or for

Yes.

I mean these are commercially available printers. So our exact cost we can't share, but that is information that if you're looking at what added the printing machines cost that's pretty readily available.

Maybe, Ted, just to add slightly more just some machines can actually come in less than $1 million and then other machines can be in the low single-digit millions.

And then going over to the kind of your revenue forecast and the development work you do, you're guiding to $10 million of revenue mix between development work and I assume some unit revenue recognition. Starting with the development work. You had $20 million of revenue with the U.S. Navy. If you kind of go through it, you rolled through about $5 million of that so far. So you've got about $15 million of that kind of left to work through. You got another, let's call it, $40 million or $50 million across other things that you think you can bring in.

When we look at -- start with the older stuff, I mean will you recognize a significant piece of that remaining $15 million in '26? How do we think about that? And then what would be the time line if you would, for the new revenue related to get these contracts and where you would start seeing it happen? And then kind of what are the milestones that you have to do to recognize that and get paid?

There's a lot of pieces there. So remind me if I forget one of them. So the -- you're correct that most of the revenue that we project to earn this year would be from R&D services, although there will be commercial revenue from following commercialization of the very initial units that are early deployment units. And part that is R&D services revenue is part of that $20 million roughly of R&D contracts that we have. Part of that we spent in 2024.

Last year, I think you're right, about $5 million spent so far there is upside opportunity that just based upon the pace of our work. Some of that could roll into 2027 as well. The new contracts that we were talking about today, we plan to get those -- we hope to get those under contract and those will -- are really more kind of next fiscal year and really will provide us that runway and pathway into 2027 and beyond. See, again, the time line for those new contracts, again, we expect it to be kind of late in the year. And then on the commercialization side, the milestones, obviously, we want to get a lot of hours on the units that we have, get these initial early adopter customer units in the customers' hands, finished testing at our location, put them on customer sites. They have to meet customer specifications in terms of their operating capabilities.

We mentioned UL certification has to be completed, and there's some steps around just our manufacturing processes and so forth that have to be optimized and repeatable and so forth. So there's a lot of steps, fairly well-defined things that we have to do, which we expect to get done over the course of this year.

Does that catch everything you were asking?

It did. And then on the 10 systems, I mean, I know Pen Systems it's a rough guestimate for what you think you'll be able to put out in the field during '26. Half of those are from the military and by I'm going to assume those are not like the boxes, if you would, that I've seen in new testing facilities, but a little more bespoke, let's call it, the remaining will you recognize revenue on any of those? I mean I'm hard-pressed to see like if you're going to do $10 million that -- and you have $10 million in revenue that you're going to recognize revenue on all 10 units that you're kind of circling in for kind of being put to customers? But would you recognize revenue on KARNO modules outside of the military in [indiscernible]

Sure. So maybe just to start off with the color of the units. So yes, about half of them going to military now. That is actually a split of some of them in at that 800-kilowatt module designed for going into the ship plus actually, just to your point, Ted, about the military will also be taking delivery of boxes as well units that are in a full 200-kilowatt enclosure, which is more focused on base deployment and prime power applications. So we're actually doing both with the military, which is pretty exciting.

And then on the units, the remaining units, which are more commercial ones, some of those will be going out to customer sites deployed. They are paying for these systems, even though we won't recognize revenue right on the front end. They are paying for the systems. And then some of those units, a couple of them, we're also anticipating having them as units that we can -- as we were talking about bringing out the data center showcase the abilities there, integrate into customer sites and really prove that application because as we look at '28, '29, 2030, there is so much growth happening in the data center space that we want to make sure that people will view us this year as a viable solution in that market. And so that's where it's important for us to showcase that.

I'll then hand it over to Jon on how it works in terms of actually recognizing that early adopter unit revenue though.

Yes. So just as an example of some of the early units that we expect to deploy initially here and some that are already built and operating. Once we have reached official commercialization, then we would expect to recognize revenue for those. There could be some that are still in the process of acceptance and so on that may slip outside of that. But yes, that's the point of of commercialization that we can recognize revenue and the earliest ones that we deliver would be the prime candidates for that recognition right away.

Yes. But just to make sure I understand. You will recognize some revenue. But when you talk about the 10 units, it's not 10 units that are going to flow through in terms of your P&L, it's 10 units that you're actually going to -- for a lack of a better term, you're going to ship? They're going to be saying they may hit have been accepted in revenue recognized

Yes, you're correct. And maybe just to decompose the revenue a little bit further, there's going to be services related to R&D services, so testing and engineering work, even stuff we contract out. And then the deliveries that Thomas just mentioned on full systems and 800-kilowatt system. And then what we would call commercial customers, the initial units that we're deploying now with our early adopter customers, those will turn into revenue once we've crossed the threshold of commercialization as well as making sure, individually on those contracts that we've met the contract -- the customer contractual requirements.

And then my very last question because I've taken up more time than I'm allowed, I think. But with -- you're up to 175 kilowatts with regards to the KARNO autonomous, you've got -- you said as you've kind of gone through testing, I mean, as part of this part of the process of why you go through beta and everything helps into the sun you got more tweaks to do to get up to the 200 kilowatts. Can you take a little time and kind of talk about where the things are that you need -- that you've discovered to -- that you need to revise and kind of where you are in the process to resolving those issues for lack of a better term, so that we can -- so that you can get to that 200-kilowatt goal? That's my last question.

Perfect. All right. So yes, obviously, great progress in the quarter get into that $175 million. We -- another core thing is we don't see it as fundamental architecture changes to get to the full 200 kilowatts. It's really about refinement. So to use an analogy, like this is a heat powered solution, right? And so it's almost like squeezing a balloon. When you go contain the heat in one area like squeezing a balloon, it wants to expand and go out other areas. And so we talked about earlier in the year of '25 the regions, that's really -- that fine mesh, the thermal battery saw a deficiency there. Once we solve that, it was like squeezing the balloon in that area to then it showcase some other areas we needed to work on.

So to give examples, those are -- over the past quarter, we've been working on a new cylinder wall sleeve that has better thermal properties, so it doesn't let heat transfer through as easily. We're working on a new piston design that reduces the amount of radiation. So heat that can actually flow through it as well as smaller things like improved thermal blankets around the solution that keeps heat and better, some unique materials that can stop heat from transferring from one part to the other. All this, as noted, it's small changes. Some of the things that I mentioned just now have already been rolled in. That's what got us to those improved power levels.

Others are at a stage where we just got the first batch of these new cylinder wall, sleeves in just this past week. Pistons, we expect to get those in, in the coming weeks. And so we are in the middle of still evolving that, but tying this all back together, it's important to note, like we're at the point where the power that these systems are producing now is sufficient to get the initial units out there. And so that is the prime focus right now. And then in parallel, we'll work on continuing to get up to that 200 kilowatts. So hopefully, that adds some helpful color.

Our next question comes from Martin Malloy with Johnson Rice.

Just wanted to ask about the control systems that you mentioned. Is that something you're developing internally? Or could we see some sort of partnership there?

We've really taken the approach of developing all the software in-house. So it is Hyliion IP, the design developed in-house by Hyliion and we see this as a key part of our solution and one that -- I didn't mention this in Ted's last question, but we even see some software improvements that we can make that will even squeak out some additional kilowatts out of the system.

So with all that in-house solution owned by us, where we will integrate with others is for site integration, right? So we will not be the primary controller on a site that is something that we will integrate with other site systems or others EV charging pedals, things like that and then let them operate it. It is key to note though, we do have the ability to integrate into other DC architectures. So think about like a battery pack, you can plug a battery pack right into the KARNO power module, and we can communicate directly with that and even control the battery.

And so we see it as a very advanced software and one that customers who have been on site have actually seen it as a little bit of like a Tesla moment where Tesla was the first to really put a large screen in display into their cars. We're -- believe we're one of the first to really put a large screen and a lot of information that a user can look at real time on the actual power module and get feedback both at the unit or through the cloud.

Yes. We've got a very strong software and controls team and they were -- many of them were some of our top people back when we had our Powertrain division. So we've been really able to leverage those skill set of those people.

There are no further questions at this time. I will now turn the call back to Thomas for closing remarks.

Thank you, everyone, for joining today's call. Apologies again for the technical difficulties at the start there, but glad we were able to get that resolved. And just setting the stage again for 2026, this is a year that we're focused on getting units out there into the field, getting those customer deployments out there and really showcasing the units working. We started the deployments of early adopter units last year, continue that this year and get units out into the field.

And then as we look at the years ahead, I mean, a lot of exciting opportunities growing, not just in prime power, but then the other 2 that we mentioned heavily on this call with the military expanding upon military contracts and then also the data center space. So we look forward to hopefully sharing further good news in those 2 areas throughout this year. Thank you again for joining the call. We look forward to chatting again on our next earnings call.

This concludes today's call. Thank you for attending. You may now disconnect.

Thank you for standing by. My name is Carly, and I will be your conference operator today. At this time, I would like to welcome everyone to the Hyliion Holdings Third Quarter 2025 Conference Call. [Operator Instructions]

Thank you. I would now like to turn the call over to Greg Standley, Chief Accounting Officer. Please go ahead.

Thank you, and good morning, everyone. Welcome to Hyliion Holdings Third Quarter 2025 Earnings Conference Call. On today's call are Thomas Healy, our Chief Executive Officer; and Jon Panzer, our Chief Financial Officer. A slide presentation accompanying this call is available on Hyliion's Investor Relations website at investors.hyliion.com.

Please note that during today's call, we will be making certain forward-looking statements regarding the company's business outlook. Forward-looking statements are predictions projections and other statements about anticipated events that are based on current expectations and assumptions as such are subject to risks and uncertainties. Many factors could cause actual results to differ materially from forward-looking statements made on this call. For more information on both factors that may cause the company's results to differ materially from such forward-looking statements, please refer to our presentation and press release as well as our filings with the Securities and Exchange Commission. You are cautioned not to place undue reliance on forward-looking statements, and we undertake no data to update this information unless required by applicable law.

With that, I now turn the call over to Thomas.

Hello, and thank you for joining us for Hyliion's Third Quarter 2025 earnings call. I'm joined today by our CFO, Jon Panzer. We're excited to share the significant progress we've made this quarter with the KARNO Power Module, along with the growing customer interest and market demand we're seeing for our technology.

I'm pleased to share that the KARNO Power Module is now performing at a level that meets the key performance needs of our early customers. On our previous calls, we outlined several system enhancements that were in development. and I'm happy to share that those improvements have delivered meaningful gains across multiple performance areas. Throughout today's call, we'll review the performance metrics that matter most to our customers, including dispatchable power, reliability, emissions and outline where we stand on each as we begin entering the next phase of KARNO deployments.

I'd like to start by directing your attention to the slides accompanying today's presentation. The first slide outlines the key performance targets that our early customers need and where we now stand against each of those benchmarks. As you'll see, we've reached a performance level that meets or exceeds these thresholds. I'll start with the power output and efficiency, which is where we saw the most significant improvement this quarter. As we shared previously, the regen component of the KARNO core had been underperforming relative to our expectations. Over the past few months, we implemented the redesigned regen and as anticipated, is delivered a material improvement in both power and efficiency. We are now achieving more than 200 kilowatts of mechanical power and north of 150 kilowatts of electrical power generation, meaning the system now meets the power needs of our initial customers.

In the months ahead, we'll continue to implement small design refinements that are expected to further improve electrical output and overall system performance. These refinements are anticipated to bring electrical power generation to 200 kilowatts as we move into commercialization next year.

Next, I'd like to highlight an important regulatory milestone that was achieved this quarter. Through our internal testing, we have confirmed that the KARNO Power Module can meet the extremely stringent air quality standards established by California's South Coast Air Quality Management District which are among the toughest in the nation. While running on natural gas, the system achieved NOx levels below 2.5 parts per million in carbon monoxide in the low single-digits parts per million, performance that far exceeds compliance thresholds and was achieved without the use of any exhaust after treatment.

In addition, we are pleased that the U.S. Environmental Protection Agency recognized the advanced nature of our technology and determine that the KARNO Power Module is not classified as an internal combustion engine under existing federal regulations. This determination means that our system will not require traditional federal engine permitting and will instead only be subject to local air district authority oversight. It's a major step forward that removes a significant regulatory hurdle and allows customers to move more quickly towards deployment as most other technologies are subject to both federal and local air permitting requirements.

Now shifting to reliability. This quarter, we reached another significant milestone with one of our KARNO Power Modules completing more than 100 days of operational testing on a customer unit without any unplanned hardware-related downtime. While the system wasn't operating continuously during that period, it underwent extensive run time across a wide range of load conditions and hundreds of start-stop cycles. These results give us a high degree of confidence in the underlying architecture and its ability to deliver the low maintenance operations that customers value.

Finally, we demonstrated the KARNO Power Modules ability to seamlessly switch between fuels while operating under load. In a recent demonstration, the power module alternated between natural gas and propane without any interruption in power delivery, automatically adjusting performance in real time. This capability highlights the true fuel flexibility of the KARNO architecture and its ability to switch fuels while in operation without impacting performance. As a reminder, the KARNO technology will be capable of operating on more than 20 different fuel types, including natural gas, propane, diesel, hydrogen, ammonia, JP-8 and many others, providing customers with unmatched adaptability across energy sources.

During the quarter, we deployed a power module to begin our UL certification process which is an important step that many of our customers expect before large-scale commercial deployments can begin. I'm pleased to report that the UL process for the KARNO Power Module is progressing well and is on track for completion in the coming months in parallel with deployment of systems to field trial locations. We've completed and passed UL testing requirements for the linear electric motor. And notably, it passed every test on the first attempt. Our UL certifier stated that this was the first time that he had ever seen a component achieve full compliance on the initial try, which speaks to the robustness of the design and the tremendous work of our engineering team.

During the build of the UL power module, our team identified several opportunities to enhance the design of both the high and low-voltage assemblies. We made the decision to implement those updates immediately rather than to deploy the unit and retrofit them later. While that choice caused some short-term adjustments to our delivery schedule, it ensures customers receive the most compliant and reliable product from the start.

Next, I'd like to share the latest with our customer deployment plans. Similar to the approach we previously shared, some customer units will first operate at our Cincinnati and Austin facilities in a controlled environment where we are doing integration and demonstrations requested by the customer. We presently have multiple customer units in the structure in Cincinnati with more planned before year-end. From there, these systems will be transitioned to customer sites for permanent operations. Other units will go directly to customer sites. We initially projected 10 early adopter units, followed by design refinements as needed based on customer feedback and system performance. We're currently building these units and aligning deliveries with customer schedules and their timing needs. These deliveries are expected to be made during the remainder of this year and in the early part of 2026. One of our early adopter customers has experienced delays in their project build-out and has shifted their project time line into 2026.

As a result, we plan to utilize a couple of these early units as demonstration systems for additional customer showcase opportunities. We'll deploy these units at various customer sites with a particular focus on data center applications and third-party demonstration centers. This strategy will help us gather further product validation and increase our visibility. We expect deployments to continue at a steady pace as we progress towards full product commercialization in 2026, demonstrating the KARNO Power Modules operational performance, reliability and scalable production capacity. Beyond that, while we anticipate continued growth in R&D-related revenue, we are not yet providing detailed guidance for 2026.

Customer interest in the KARNO Power Module remains very strong, and we continue to introduce new customers to the technology every month. These customers are not only seeking reliable power generation capacity, but also the differentiated attributes that set the KARNO system apart, including its ability to operate on multiple fuel types, improved resiliency, power density, low maintenance design and high efficiency. To date, we've executed nonbinding letters of intent with customers representing nearly 500 KARNO Cores. Based on current demand, we anticipate being supply constrained for the years ahead as interest in the system continues to grow across multiple sectors.

We're also deepening our engagement with the U.S. military as they explore broader use cases where energy security and reliability are mission-critical. As part of our existing R&D contract with the Navy, we plan to deliver additional KARNO Power Modules and Cores in 2026 for specialized testing required for shipboard use. Our team recently visited the first autonomous Navy vessel that will be powered by KARNO Cores. That ship is currently undergoing sea trials and is planned to be outfitted with its initial KARNO units in 2026. The ship is designed with numerous engine bays and is capable of housing multiple megawatts worth of KARNO Cores for that ship's power needs. This is an especially exciting program given the Navy's significant power requirements and the unique advantages the KARNO system offers in delivering efficient and low maintenance onboard power with an excellent thermal and acoustic footprint.

Next, I'd like to share an exciting update in the nuclear space. Hyliion has engaged in an exploratory agreement with one of the leading organizations in the small modular reactor or SMR sector, to evaluate how our KARNO technology can be paired with next-generation nuclear systems. SMRs generate heat that is traditionally converted to electricity through a steam turbine. Since the KARNO Power Modules fuel source is heat, we're exploring the potential of replacing the steam turbine with a KARNO Core to generate electricity more efficiently. While this collaboration represents a longer-term opportunity, we believe it is an important step in exploring now given the growing interest and investment we're seeing in the nuclear energy space.

Lastly, I'd like to touch on some important developments we're seeing in the data center space. NVIDIA recently published a technical report outlining that the next generation of data centers, particularly those supporting AI workloads, will increasingly adopt 800-volt DC architectures to improve efficiency, reduce conversion losses and enable more scalable power distribution. This shift aligns directly with one of Hyliion's key advantages. The KARNO Power Module native output is 800-volt DC. While other technologies will require additional and costly conversion equipment to connect to these high-voltage systems, the KARNO system design allows for direct integration. This positions Hyliion as a natural fit for the evolving power architecture of modern data center and AI infrastructure.

Now shifting to our capacity for the years ahead. The pace of deployments will be guided by insights gained from the initial units as we continue to fine-tune system performance and ensure consistent reliability. At the same time, our production capacity continues to expand as we transition more activities from our R&D facility in Cincinnati to our larger manufacturing facility in Austin. We now operate roughly 30 additive manufacturing machines across 3 generations of printer technology. Our current focus is on optimizing these existing systems for higher throughput while continuing to add additional machines to support growing demand.

To help lead this next phase of growth, we've recently welcomed Darrell Preble as Vice President of Operations. Darrell brings extensive experience from the energy sector with a proven record of building scalable manufacturing programs and driving operational excellence, which he'll be responsible for at Hyliion. Prior to joining, he held senior leadership roles at Cummins and Husky Technologies. I'd also like to share some updates on the supply chain front. This year, we've been focusing on expanding and qualifying new suppliers to support future production scale-up.

As part of that effort, I want to highlight one supply chain challenge we're actively managing. Like many U.S. manufacturers, we're experiencing challenges in sourcing the high-strength magnets out of China that are used in our linear electric motor. At present, we have sufficient inventories to support operations through the next couple of quarters and our team is actively pursuing multiple alternative sourcing strategies. That said, the broader supply environment for these magnets remains uncertain, and we're continuing to monitor it closely to mitigate any potential impact on our production schedule.

To wrap up, we're very encouraged by the progress we've made so far in 2025 and the strong position we're in for the months ahead. We're excited to expand our customer deliveries and transitioning the KARNO Power Module into real-world applications. As a reminder, from our last call, we shared the KARNO Power Module and its supporting infrastructure will receive a 30% tax credit. This incentive will help accelerate adoption and support our commercial ramp up over the coming years as it remains in effect for the next decade.

With that, I'll turn the call over to Jon for the financial update.

Thank you, Thomas, and good morning, everyone. In the third quarter, we recorded revenue of $800,000 from research and development services related to our contracts with the Office of Naval Research. Cost of sales was also approximately $800,000, resulting in a small gross loss. In the third quarter of 2024, we recorded no revenue or cost of sales. R&D services revenue reflects both the sale of KARNO Cores and related components to the U.S. Navy, the work we perform to test and validate these units and other development work.

Operating expenses for the third quarter were $15.3 million compared to $14.2 million in the third quarter of 2024. The increase was mostly related to higher research and development costs and lower gains from asset sales in the powertrain exit and termination line. The increase in R&D work reflects the more rapid pace of development and growth in the production of additive components. SG&A expenses were $5.2 million, down about $0.5 million compared to the third quarter of 2024 and due primarily to lower facilities and insurance costs, partly offset by a small increase in labor costs. We recorded $2 million of interest income during the third quarter, down from $3 million in the prior year quarter due to a lower level of investments and lower interest rates this year. Our total net loss in the third quarter was $13.3 million, up from $11.2 million in the third quarter of 2024, but about flat with what we reported in the second quarter of this year. Year-to-date, we reported revenue of nearly $2.8 million, all from R&D services and gross profit of $96,000. We reported no revenue or gross profit in the same period in 2024.

Year-to-date operating expenses were $50.7 million compared to $47.2 million in the first 3 quarters of 2024. The increase is related to higher R&D expenses this year, partly offset by lower SG&A and powertrain exit and termination expenses compared to the same period in 2024. Net loss year-to-date was $44 million compared to $37.7 million last year.

Turning to our cash and investment position. We spent $20.6 million during the third quarter and $55 million this year-to-date. Year-to-date capital spending was $22 million and consisted primarily of additive printing machines and related equipment along with the facility investments to support printer operations. Cash from asset sales year-to-date in 2025 was $1.2 million. We finished the third quarter with $164.7 million of cash and short- and long-term investments on our balance sheet. Throughout this year, we forecasted total 2025 cash expenditures of $65 million, including equipment financing of $10 million to offset part of this year's capital investments, leading to a year-end cash and investment balance of approximately $155 million. At this time, the timing of equipment financing is uncertain with a possibility it may shift into 2026 or end up being somewhat less than $10 million, depending on the decisions we make related to timing and available terms.

R&D services revenue has ramped up a little slower than we expected partly due to a decision to defer delivery of some of the early deployment units into 2026 based on the timing needed by the U.S. Navy. Therefore, full year 2025 revenue will likely be approximately $4 million. As Thomas noted earlier, we expect that KARNO commercialization will occur in 2026, at which time we will begin recognizing revenue from KARNO system sales.

We continue to expect that the capital we have on hand today will be sufficient to carry us through commercialization of the KARNO Power Module. Looking ahead, we anticipate that additional capital will eventually be required to support production growth, particularly with the purchase of additional additive manufacturing equipment. However, with approximately $165 million in cash and investments as of the end of the quarter, we are well positioned to be deliberate and opportunistic in determining the timing and structure of the capital raise.

Now I'll turn the call back over to Thomas.

As we wrap up, I want to leave you with 3 key takeaways. First, the KARNO Power Module is now performing at a level our initial customers require. This marks a major step forward for the product and gives us confidence as we begin broader customer deployments. Second, customer demand continues to strengthen. We have nearly 500 units under nonbinding LOIs and the newly established 30% investment tax credit further enhances the economics for our customers adopting KARNO systems. Together, these factors create a strong momentum heading into next year.

And third, we're aligning with where the world's energy needs are going. From NVIDIA's planned move towards 800-volt DC architectures in data centers to our exploration of pairing the KARNO technology with next-generation nuclear systems, our technology is uniquely positioned to serve both near-term and long-term power needs. These advancements show that we're moving from development to deployment, turning the KARNO Power Module into a real-world solution for customers who need reliable distributed power.

I'd now like to hand the call back over to the operator to begin Q&A.

[Operator Instructions] Your first question comes from Ted Jackson with Northland.

So I got, I don't know, 3 or 4 questions. The first one is the customer that has shifted its -- I guess, you would call it acceptance of the initial KARNO units from '25 into '26. I mean not that big a deal. Could you just talk maybe like what vertical is that customer?

Yes. So this is the Navy. And so as we highlighted on today's call, exciting news is that ship is now in the water, in the ocean, going through sea trials and then we're scheduled to now be installed into the ship in 2026. And so that's why what we also wanted to highlight on today's call is it's not beneficial for us to build up those assets and then just have them sit waiting to actually be deployed into the ship. And so what we're going to do is we're going to take some of those, and we're actually going to make them demonstration units that we're going to bring out to data centers to some third-party validation sites and actually utilize those assets to showcase the technology there, get more customer momentum growing, but then still be able to meet the demands of the Navy in getting those units into a ship in '26.

And then that ship, that's the unmanned ship that they're like -- I mean,, there's been a lot of press about the development of that ship, and that's what it is correct, the kind of this next-generation sort of crews.

That's correct. And Josh and I and some of our team actually got a chance to go board the ship this past quarter. And it's unbelievably neat. It's really designed for weaponry onboard and then fuel and engines to move the ship around. So very different than conventional Navy ships where you usually have to have sleeping quarters and cafeteria and medical, all those things that personnel need. The ship has none of that. It's the hull of the ship. It's fuel storage. It's a handful of engine bays that we can deploy multiple megawatts of KARNOs into.

And then the deck of the ship is designed where they can mount different types of weapons on board. And the goal of the ship is it gets deployed and then it will be out in the ocean for a long time prior to coming back to dock. And that's one of the key drivers why the Navy was so excited about our technology is because if you think about like a diesel engine that a ship is normally powered by, that requires every few hundred hours, you need to do an oil change. Well, if there's no one on board, there's no one to do that maintenance. And that's where our low maintenance design really matches what they needed out of an engine.

I've seen pictures of it. It's a pretty cool looking product or system. Then the next question for me, going into kind of the testing and such is on Slide 3, I saw that you had a couple of KARNO units sitting in the parking lot going through your testing phase. I mean, is that what I'm seeing there or is that you have -- when I came out and visited, you had one out there, now you have 2 out there, I mean that itself is an indication of progress with regards to you getting systems and going through those initial runs for your end customers?

That's correct. So we've got a couple of customer assets in operation at the facility. We also have a couple of our own assets in operation at the facility. So you're absolutely correct. And that was one of the reasons why we put that photo in there just to show that we are building these systems. We're getting them out there.

As we highlighted on today's call, one of the next big phases here we're working on is getting that UL certification. So we're going through that testing now. And then once we have that, then that really enables us to start getting these assets out to customer sites.

When do you think you'll have all the UL testing completed?

Yes. We're talking in the next couple of months here, we'll have that completed. And as we highlighted on today's call, so there's a couple of different phases of UL, right? So for instance, the linear electric motor needs to be certified, the battery needs to be certified and then the whole power module itself needs certification. So one of the big milestones we hit coming into today's call was we were able to achieve the certification around the linear electric motor. So we went through that testing. We got the approval there, which as we shared, I mean, a big accomplishment where we actually passed it on the very first attempt of doing it, which as our certifier said, he had never seen something passed on day 1 of testing, which is great to see.

Congrats on that. I have a few more. I'm going to ask one more and then if I need to, I'll come back in line. And that is on the regen rebuild. You mean you [ comment ] yourself the initial systems, and this is not a surprise, I don't have that component in it that it's not required for your customers and it's and easy swap out. Have you, at this point, taken the new components and started putting it in units and doing any testing with it? And is that like just for yourself? Or are you doing that and you will go into some of the initial beta units?

Yes. So both in our system and then also it's now going into customer assets as well. So maybe just a step back, in past quarters, we talked about we saw performance deficiency with the regen and we were working on depowdering and new architecture, new design of that part. Over the last quarter, we implemented that part into our own asset. We saw a step change in performance out of the system. So that was great to see. And that really highlighted or brought forward what we shared on today's call, which is we've now gotten the performance to a point where we believe we're going to -- we're meeting the needs of the customers.

So for instance, all of our early units that are going out to customers only require up to about 150 kilowatts of power generation. We're now exceeding that. We're now north of 150 kilowatts. So our thought is, let's move forward. Let's go ahead with the design as is, get units out into the field. And then as we near commercialization, there are some small areas that we want to continue to improve. So we're seeing some heat leakage through the piston. We're seeing some heat leakage through the cylinder wall, a sleeve that goes in there. So minor changes like change in the material of that component. So we're going through some of those changes so that as we get into commercialization or as we enter commercialization, we plan to be at the full 200 kilowatts of power output of the system.

Your next question comes from Sean Milligan with Needham.

Just a quick question on your pipeline. The 500 units that you have in LOI, you're going to be delivering some units next year to those customers. I'm curious about the timing you expect those customers to then like test those units at their own operations and sort of the conversion from LOI to purchase order, maybe what your expectations are on timing there?

Sure. So even before we get the first asset out into their operations, some of the quantity of units will move from LOI into a firm agreement. And then from there, it will be how do we continue to work through that backlog of the additional units on the LOI, which I think to your question is how long do they need to test it, which we're hearing in the 6 to 9 months is kind of the rough time frame that customers want to run the asset, experience it. And then from there, they believe they'll have the confidence and the performance to keep going forward.

It varies between customers. Some customers are saying, let's do one unit right out of the gate. Others are saying, let's do multiple units right out of the gate. But overarching the message is, there's a lot of demand in this product and the demand is continuing to grow. Even tomorrow, we have another new data center customer coming in that we're having great discussions with. And so we're continuing to build that customer interest. And as we shared today, we foresee that we will have enough demand in this product for years ahead that will actually be supply constrained as opposed to demand constraint.

That's great. And then -- I mean, the other question was around the 500 units in LOI. Can you give any context around how many customers that maybe represents? And then sort of the -- like the timing of those. I think those date back to like even 2024, first half of this year, but maybe how your discussions have expanded sort of like in terms of numbers of different opportunities you're now seeing in your pipeline? Just trying to understand how the pipeline is progressing as you're kind of moving KARNO forward here?

Sure. So just kind of rough numbers, I believe it's in the dozens of customers, not in the hundreds of customers that we signed LOIs with. We are in discussions with even more than that. Not all discussions go to an LOI. Some discussions go to, okay, great. We like the technology. We want to see it in operation at other people's sites before we move forward with making a commitment.

So with all that, how has the market landscape evolved? I would say, initially, we were seeing a lot of interest from the military, a lot of interest from EV charging. We've actually seen some sectors that have popped up that have become even larger than those. So the data center space, seeing a ton of interest there. Data centers are just power constrained, right? That's their #1 issue. When they go to build out a new data center, they talk to the utility. Usually, the answer is there's not enough power available. And so the data centers have really shifted to moving to make your own electricity on-site power generation, which that's exactly what our technology is. Even as we highlighted in today's call, NVIDIA has come out and expressed that they see the future of data centers in the years ahead here actually shifting to an 800-volt DC architecture, which that is the architecture of the KARNO. We produced 800-volt DC natively. So we think we can fit perfectly into that.

And then we've also seen just a growing demand on the commercial facility side of things. So the driver there is really can you produce your own electricity for cheaper than you could buy electricity from the grid. So I'd say data centers, commercial applications are growing. And then the work with the Navy we've been doing has been going really well to a point where they've actually pulled in other branches of the military that we're now starting to work with as well.

Okay. And then the last question is around sort of like manufacturing scale up. So if you go back to the first part of this year, you had powdering issues, which you've addressed. And then also, I think you had some outsourced -- you had outsourced product manufacturing, which you brought back in-house. It sounds like the manufacturing side today, you're kind of honing in on design and feel comfortable with where you are. But I'm more curious about your ability to scale in the future. So the confidence in coming to like a systematic manufacturing process, the ability to scale up printers and then just anything you need to do in-house there? And I think you brought up magnets, but obviously, like a lot of these award potential on the data center side specifically is very large. So just the ability to source materials. I just wanted to get your big picture thoughts there for Hyliion moving forward.

Sure. So I'll first start with the challenges we were facing. So we had outsourced production of the electric motor. We pulled that in-house, as you highlighted. And then we were having issues with the regen and the deep powering. We believe we've gotten our arms around that as well. So as we did this past quarter, we leaned in and said, yes, manufacturing is going to become a big part of the story ahead. That's why we hired Darrell Preble to come in and lead operations for us. He's built out manufacturing facilities and distribution facilities for Cummins and Husky in the past. So this is not an unknown journey to him, which is great.

And then as we look at the phase ahead here, I think one of the big areas of focus is going to be how do we continue to get more and more out of the additive manufacturing machines. So we've got about 30 machines installed so far. We've got about -- in total, about 3 dozen roughly that we've ordered. So still a few more to be delivered here. And then from there, one of the focus areas is going to be keep getting more and more out of the machine. So we know that we're not operating at their fastest parameters at the greatest throughput. That takes some just work to get them to those faster throughput. So that will be our next phase. And then from there, we'll look to continue to expand the number of printers we have and buy more assets, and that will be how we scale volume.

So the reason that, that is really key to the story is we actually expect that we're going to slow down our capital expenditures. The goal right now is not just keep buying more machines, the goal is actually going to shift to let's keep getting more and more out of the existing machines we already have. And then as Jon highlighted, then there will be another phase where, okay, let's continue to expand manufacturing and buy more assets.

Your next question is from Ted Jackson with Northland.

I wanted to circle back. You made a comment about -- and honestly, I was taking notes, I missed what you've said, but that you did -- you had a bit of delay in delivery or development because you decided to go for work with that 800 kilowatt? Is that what you just sort of -- can you go through that again with what you said in your prepared remarks?

Yes, absolutely. So as we were putting together the UL system, we saw some areas of opportunity that we can improve both high voltage, low voltage. And so we decided to do those changes prior to going through UL certification because once you certify it, it's much more difficult to go make changes after that. You have to go back through another recertification process. So what we decided was, why don't we go ahead and roll those improvements, those changes in now and get those done, which has been done. They're in the -- to go back to your reference of the photo of the parking lot, they're incorporated into that UL system that's out there.

So we made those changes. Now that did add some time to the schedule. And so that was just a strategic decision we made of let's get those rolled in and go through the UL certification process with those as opposed to have to do an alteration or change down the road.

There are no further questions at this time. I will now turn the call back to Thomas Healy for any closing remarks.

Great. And we actually had a couple of questions come in that Greg will read out for us as well.

Thomas, can you share more about why you see KARNO as a good fit with nuclear and what time line you envisioned for developing a solution in that space?

So today's call is the first time we've really talked about nuclear, but the reason we wanted to is because it really highlights the versatility of the KARNO system. So as we've shared in the past, I mean, the KARNO is a heat-powered generator. And so if you look at nuclear, all nuclear does is it makes heat. And then conventionally, nuclear SMRs will use steam turbines to take that heat and convert it into electricity.

And so we started engaging with an organization in the SMR space, in the nuclear space who were looking at actually replacing that steam turbine with a KARNO. And we're seeing that we should be able to produce electricity more efficiently than even a steam turbine can. And so why that's important is we're seeing that there's a lot of interest and a lot of momentum in the nuclear space. It's not happening tomorrow. This is still a ways out as to when nuclear and SMR will be deployed. But we think it's important to get our name into that discussion now because a lot of these companies are in the R&D development phase and could select us as their way to produce the electricity as opposed to using a steam turbine.

You mentioned on today's call that NVIDIA has highlighted a new electrical architecture for data centers. Can you elaborate on how this matches Hyliion's existing architecture?

So data centers today are a lower voltage architecture. They are DC power, but they're in the 50 -- I believe it's about 58-volt DC architecture. And NVIDIA came out with a report a white paper that said they see the future actually moving to 800-volt DC power. Reason being is that as computers become more powerful, they consume more electricity. And when you just start scaling up, you end up consuming a lot of capital and just even the cost of copper to move that electricity around.

So if you move to a higher voltage architecture like this 800-volt that we're at, you can reduce how much copper you need, you can reduce some of the other components and make the whole data center architecture more efficient and less costly. And so that's where we were excited to see this because it matches us perfectly. We make 800-volt DC power natively. So if that's where data centers head and they start adopting, other technologies like the grid, 480-volt AC architecture are going to require other components, other inverters or converters to move to that 800-volt DC versus we'd be able to couple without those components. So it just aligns well with our vision and structure of what the product outputs and where data centers are heading.

So with that, I believe that concludes the questions that we had come in. So on a closing note, I appreciate everyone joining today's call. As we highlighted, we're pleased with some of the key metrics we've hit on the performance side of things. We're seeing growing customer demand, and we're excited about not only being a solution for the near term here, but also some of these future opportunities we discussed as well. So with that, we'll talk again next quarter. But in the meantime, I do encourage listeners to follow us on our social media channels, also follow the press releases we released as we'll be sharing more throughout the quarter on those channels. Thank you for listening.

This concludes today's conference call. Thank you for participating. You may now disconnect.