Andy Lowery: Building Force Fields Against Drone Swarms

At Contrary Research, we pride ourselves on being the best starting place to understand any private tech company. So the very first question we always want to hear from the horse's mouth: give us a quick summary. What is Epirus? What's the elevator pitch?

Well, in a word, we've created force fields. In a word, and I'm not kidding. We literally create area sectors where we can put an electromagnetic field in place to deny the use of consumer electronics. In the case of drones, you can put wide-area shields, if you will, at fence lines. We've seen a lot of news lately: New Jersey drones, 17 drones at Langley. Put a system of ours at a fence line, and you stop everything like a force field. So as far as version one of homo sapien inventions, we've invented version one of a human being force field.

So, yeah, we've created an incredible platform, and the type of company that we are, a nontraditional defense contractor, sometimes they call us a neoprime, it's taken that kind of company to push this sort of invention over the line. And we've done it, and now we have something that's going to work well in a number of layers to defend different assets: bases, of course, but beyond bases, stadiums, refineries, ports, you name it.

We're going to unpack the product itself, what the market for it is, and how the trends are changing to shape the need for it. Before we dive into the details on Epirus, walk us through your background. Not just the bulleted list of what you've accomplished, but paint us a picture of your journey and how it led you to become the CEO of Epirus.

Oh, it's great. Sometimes we always tell our stories narratively, like these storylines, but it was a lot of haphazard, Mr. Toad's Wild Ride a little bit for me. I'm originally from Illinois. The first half of my childhood I grew up in Southern Illinois, and the second half I grew up in the Chicago area. I left high school and enlisted in the Navy right out. I didn't go to college. I went right into the Navy. When I signed up, they told me, "You belong in nuclear power. You should be a nuclear engineer." And I'll tell you, I didn't know anything about the Navy at the time, or the military. I didn't even know they had nuclear power plants and reactors. That was a totally foreign idea to me, but I ended up becoming a nuclear electronics technician.

I did really well in a lot of the coursework, and so they offered me a chance to be an officer. In the mid-'90s I got to go to the University of Illinois, majored in electrical engineering, went back into the Navy, and served as a nuclear officer aboard aircraft carriers. You can choose subs or aircraft carriers; I wanted aircraft carriers. I did that for a number of years.

In 2002 I went into the reserves and started a civilian career, first at MACOM, which was a big part of my journey, where I learned a lot about high-powered amplifiers. The next part of my journey, and maybe one of the most educational, was Raytheon. I spent the better part of a decade running huge organizations at Raytheon. I got very lucky to be young and still able to run large organizations. I ran a whole portfolio of space programs. After that, I ran a whole portfolio of intelligence, surveillance and reconnaissance programs. Then I culminated a career there as the overall chief engineer in charge of all of electronic warfare for space and airborne systems.

I did that throughout the time that we bid on a very important program called Next Generation Jammer, which is sort of, I'd call it, the father of Leonidas. NGJ doesn't have a cool name like Leonidas, but NGJ was very much a father to the concepts and ideas behind the approach Epirus has taken later on, to leverage things like gallium nitride and an arbitrary waveform generator.

Then for the past 10 years I've been in big data and a lot of augmented reality, a lot of computer perception, because in the world we're living in, computers are becoming aware of where they are in the world and what's around them, and then how to do a mission or navigate objects. That idea is common to augmented reality, virtual reality, drones, robots. It's this plethora of things like inertial navigation that all involve computer knowledge and perception and autonomy.

So when a counter-drone CEO is on your show and you say the word autonomous, ask him what's the first word he thinks of. I would answer swarms. You say autonomous, I say swarms, because that's really what it enables: the ability to attack with many, many different objects all at the same time, all thinking and moving and behaving completely independently, without, and I stress this, any connectivity to the outside world. So traditional ways of jamming or disrupting or taking over the signals that control them will no longer be a possible method of attack. You'll have to move into something that works differently in order to get into the electronics and have any kind of effect, other than an old kinetic shot, where you hit one missile onto one cheap drone.

My journey pretty much ended there, apart from getting a phone call from one of my old Padawans, Dr. Bo Marr, who founded Epirus back in 2018. I had called him back in the day when he worked at Georgia Tech. I hired him at Raytheon. So this time he returned the favor and said, "Hey, Andy, would you like to come join?" I joined the company a number of years ago, first as the chief product officer, and helped with some of the engineering and the design of the system. A couple of years after taking that role, I transitioned to CEO, and I've been the CEO for a little while now.

And I'm telling you, this is it. This is the number one job. I told you all that career, Navy, MACOM, big Raytheon, startups, and out of everything, I can't think of a job I would rather have than being CEO of Epirus. This time, this urgency, this mission, this team of people I have working for me: it is a dream. It is my dream come true. I'm very, very happy to be in the role I'm in. But that's my story.

Your career touched on a lot of the ways this aspect of warfare has evolved over the years. We've worked with a number of companies that work directly or adjacently with the DoD, like Hadrian and Anduril, and these companies are by necessity defined by 50, 60, 70-plus years of technological evolution, but also by the DoD's willingness to work with certain technologies. High-powered microwave traces its history back to research done in the '60s and '70s, and you touched on gallium nitride semiconductors. What do you feel are the pivotal moments that have made this the right time for Epirus to do what you're trying to do?

That's a really great question. High-powered microwave, you're right, has been around for a long time. Starting way back in the '50s and '60s, they began to work with an old technology that nevertheless has a lot of incredible performance characteristics, called transmit wave tubes. Some people call them vacuum tubes: these older vacuum-based systems that were able to take energy and boost it up.

But here's the thing. They could only boost it for short durations of time, and when I say that, very short durations of time. You may be able to get an old-fashioned high-powered microwave system up to gigawatts of effective isotropic radiated power, EIRP. Gigawatts. However, you can only maintain that power for the briefest interludes of time, maybe 20 or 30 nanoseconds. And because of that short duration, if you take the math, it takes a frequency spectrum and spreads it very wide. That's just a principle, if you're not an engineer: a short duration of time gives you a wide set of frequencies, and vice versa. So they call that old-fashioned way of doing high-powered microwave wideband.

Where we came along and did something different is we took the idea of high-powered microwave, which conceptually is just the idea that we're going to push such a powerful microwave field that we don't need to come in through any open front doors. We can come straight into the wires and the connectors and the boards. Because the energy field is so incredibly powerful, we're able to just come right in.

Now, that's where the similarity between our system and traditional high-powered microwave systems stops. OK, we both get in the back door. But once we get into the house, at whatever power level, then what does the system do? Remember, the old-fashioned high-powered microwave system can only be inside for a few nanoseconds. They don't have a lot of time to mess around. They've got to get in, smash something with a hammer, and get out of there, because that's just the limitation of that type of system.

The Epirus system is a high-powered microwave system that is also a high-energy microwave system. You have high-energy laser. You have high-powered microwave. Epirus is a high-energy microwave system. That means when we get in the back door, we stay around. We hang around a long time at the party making a lot of ruckus: making noise, disturbing things, messing up the way they think, the way they sequence. We're at the tip of the spear of even learning the profoundness of the things this energy can take over and take apart when you're inside these circuits for a long period of time.

So there's that approach, and some cleverness in how we implement it, with an arbitrary waveform, software-defined system where almost anything we can think up in MATLAB, we can get out the front end of our system. The combination of those two elements led to this moment, along with a third element, which you alluded to: the gallium nitride progress we've seen over the last two decades.

You have a substance called semiconductors. It was originally all silicon. These semiconductors have very unique properties that parallel some of those transmit wave tube properties we were talking about: they can amplify things. For the longest time we were stuck with just silicon. Then in recent decades they came up with new substances. Sometimes they call them III-V devices, because they sit on either side of the periodic chart. Here's a fun fact for your engineers out there. The periodic table is labeled by the number of valence electrons in the outer shell. Silicon's at four. Well, you have some that are at three and some at five, like gallium and nitride. You put those two together and you have a III-V material. Same with gallium arsenide, same idea. III-V materials have a much lower depletion region energy, much easier to get over, and they have inherently better performance.

Gallium nitride came along and became able to withstand enormous energies, enormous powers, enormous heat fluxes. Do you realize that inside a gallium nitride junction, there's a higher heat flux than there is on the surface of the sun? This is the type of magical rock this is. And gallium nitride has only performed better and better over the decades. For a little while we joked internally and called it Marr's Law, for Bo Marr, because a similar idea is happening as with processors, more and more processors over time. Gallium nitride is getting higher and higher voltage, more and more power density. If you look at the progress over the last 20 years, you see a trend that doubles every so many years. Dr. Marr picked up on that trend and said, "Look at this. Notice this, guys."

Right here was the line where we thought gallium nitride became significant enough that it could participate alongside the older technology in a fair fight, meaning we can get in the back door. But that's when the fight stops being fair. Because, like I said earlier, a solid-state gallium nitride device can hold that signal for a long time, one millisecond, however long. All of that energy dancing inside the house after you get in the back door creates profound effects.

And you push that energy into a large section of the sky, because you put it all in a phased array, which is another thing solid state allows you to do: stack these things up into bricks and create scanned arrays. When people don't know what I'm talking about, I say, "Have you ever been to a laser light show?" They say yes. "You ever see a laser that draws an elephant in the air? Just one laser, and you see the elephant?" That's how our system works. It can draw microwaves. You can't see them, but it draws microwaves in a square, in a circle, in whatever shape we want, and takes up a whole section of the sky in an instant. In less than a second, we can scan every single space of the sky within about 60 degrees by 60 degrees. That's what puts up our field. It's a temporal field, very quick, that scans again and again and again, and that scanning ability of the front end is what makes it persist.

You mentioned software defined. What have you been working on on the software end, and what capabilities does that unlock that didn't exist in previous generations? For example, I know you have features where you can carve out zones in the air that you don't want to target.

Yeah. Of course, we're leveraging many decades of gallium nitride, gallium arsenide and phased arrays. Look at Raytheon's systems; lots of them are phased arrays. SPY-6 is a phased array. Next Generation Jammer is a phased array. APG-79 is a phased array. The phased array capabilities are profound and mature. We can notch out frequencies to make sure the frequencies that blue drones are affected by aren't affected by our system. We can steer the beam very accurately and know precisely where it's steered, so we can steer away from friendly forces.

But in general, what we've done is build a system, again similar to Next Generation Jammer for those in the audience with some familiarity with it, for a game that is always cat and mouse. Let me start with that. People will naturally ask, "If you do this, why don't they do that? If you figure out a frequency to attack them on, why don't they change some shielding and move things around so that frequency doesn't work anymore?" Well, yeah, that's what they do. A hundred percent. That's exactly how the game is played. They'll see that I have this powerful system that hits them here and hits them there, and they'll change by shielding or something like that.

But that does two things for me. One, it drives them back into the hardware, because they don't have a jammer where they can use encryption or signal processing to get around me. They're going to have to get into the hardware. They're going to have to put on heavyweight shields, different things that weigh down the drone, that the drone might not be designed to support, and very, very manual processes, like putting globs of RTV, which is this jelly that resists radiation. All of these things drive them to lesser and lesser capability to build, lesser and lesser volume, and put them more and more into the military lane. It brings them back to our playground, where we do all the real damage, right? It brings us back into our lane, where the DoD is by far the strongest. The consumer electronics lane, we've given a lot of the keys of the kingdom to China. They own a lot of strength in that category. But in the DoD lane, we still reign supreme. So what Epirus has done is build a machine that drives them back into the military lane of weapon systems, because we deny anything that's simply and cheaply produced with consumer electronics.

And when they change their electronics, let me give you the cherry on top. We change our system, but we don't have to change the hardware. We just reprogram it. That could take us minutes to days, but not more than days. We've not had an experience where it's taken us more than days to pick the lock. We've been able to do it in minutes, or even seconds, with the system in real time, running through a whole litany of different magazines and techniques that do different things. The system has many tools, hardware tools, so we can play the guitar. We can do polarization. We can do different frequencies. We can do different pulses, different pulse repetition rates. All of these things are part of a very complex, very complicated equation that we can leverage so that our cat-and-mouse motion is just software. It's just a software upgrade.

That frames the picture nicely. As conflicts change and people make specific changes to their hardware, especially with the onslaught of consumer drones being used in military use cases, you're zeroing out a lot of that high-volume, low-cost capability and forcing it back to being a military creation. Before we jump into the changing face of those conflicts, we've touched on a lot of the core enabling innovations around the Leonidas system. How has the product evolved? From the core stationary system, you now have several other form factors, like mobile and the pod. Walk us through the surface area your products cover.

Today we're in generation one, which we're sending to the Army now. That generation one was preceded by several prototypes. Believe it or not, even though they're very expensive machines, we build these almost like a cell phone company builds a cell phone, with a process that's known in the industry as EVT, DVT, PVT. We spent a year or so building our first engineering verification and test article. That was our first prototype. Then we spent a couple of years designing several iterations of our design verification and test article, which involved a lot of what we call soldier touchpoints.

We ultimately landed on our generation two system, which we're building right now. The generation one systems are deployed. Some are in Central Command and some are going to Indo-Pacific Command somewhere. I can't say where, but they're getting deployed now, the early prototype version. The next prototype version, which will be our final prototype prior to a program of record, we're calling generation two.

Those prototypes will be much more capable systems. They're going to run on batteries, for one. They'll have a huge amount of lithium batteries that can burst a lot of energy in a very peaky type of motion, because we want a lot of peak energy over time. They'll leverage diverse polarizations, where we can select the polarization depending on the target we're going after, and we've found the right polarization against the right system creates quite a profound pairing. We've been able to increase the total pulse width of each of our pulses. We've been able to increase the duty cycle under those burst periods. All of those additional features are going to be part of the new system, and they're going to really elevate our range.

We have a low-end range and a high-end range because we're a field. As things enter the field, they have different susceptibility levels, and as they get deeper and deeper into the field, that susceptibility gets ferreted out. So if it's a super non-susceptible system, it might get in pretty close. Not too close; we're still talking hundreds of meters, not a few meters away. But it gets in pretty close. We're trying to continually push that bottom-end range up and up. By doing the things I just mentioned, in the lab as we run experiments inside anechoic chambers, we're able to see an extremely measurable increase in our ability to have effects at a much longer range with generation two.

Generation three, which is just being discussed, has a lot of new features even beyond that, including new gallium nitride we're working on with MACOM that may even eclipse the highest gallium nitride in the world today. They've built a device that's even higher. We're talking about potentially three times the power density of the current system per individual brick. We have about 150 of these bricks, and each of those bricks will have three times the power in the generation three system.

So we constantly try to get EIRP better and better. That's delivered through more and more power-dense gallium nitride, but also just by increasing the size of the system. If we elect to do a 20-by-20 system, it's going to be much more powerful than a 10-by-10, for example, as we scale the system up and down. And since we've got that core foundational IP done and established and working and reliable, we've been able to scale it to everything from a six-element system that we're talking to Army Futures Command about, for rear protection for Abrams tanks against first-person-view drones coming in and hitting them from the rear in Ukraine, all the way up to big base protection, and getting larger and larger to protect ports and other vectors for attack. It just depends. If you're defending the north and south of Taiwan, you probably want to go out a couple of miles at least, if not further. You need a bigger system.

The system is very parametric that way. You can make it big, as long as you can contend with the size, weight and power requirements. You ever hear of something called the PAVE PAWS system, Kyle? That's an old early warning radar designed in the '80s. The reason I know about it is that MACOM provided the silicon bipolar L-band devices that produced the transmit power for that radar. If you Google it, the PAVE PAWS radar is a building. It's a round radar the size of a building. Well, if you put a building-sized Epirus Leonidas out there, you might have 15 or 20 miles of effective range that you could push that beam out to. Obviously that's a lot of work and engineering to get that big of a system put together. But if you're thinking about the north and south of Taiwan, you could put systems of PAVE PAWS size and capability there and have 20 miles of protective coverage of the north and south of that island.

So lots of different opportunities. The system is very parametric. We've run through a number of iterations where we've learned more and more about energy, and just how important energy is over straight power. Straight power is good, but with great energy comes a great system, with a great number of effects that are much more powerful and much more lethal than traditional high-powered microwave.

Is there anything you can share about the timeline, whether that's when Gen 2 will likely be deployed or the longer-term path to Gen 3?

Gen 2, we're building. If you were here at the factory today, you'd see some of the initial steps to build four Gen 2 systems. So we're in process to build at least four systems, unless we get the word to order more, which we may do with this new administration. They've been talking pretty seriously about that.

The Gen 3 system is the program of record system. That's the system that's going to be the catch-all for all the soldier touchpoints and all the learning we're going to do over the next couple of years as we deploy the Gen 2 prototypes. The Gen 3 system will likely be a program of record in 2027, if not sooner, if an acceleration button gets hit again. Right now we're angling for 2027, and we're on track, knock on wood, to continue to deliver, so the customer gets enough confidence in us to push us into the regular PEO lane, which is the office that will run that program of record.

You touched on expanding both the power and range of the product as it evolves. If I were to map that onto areas you're not addressing today that you could address as the product evolves: when people visualize drones, they visualize everything from several-thousand-dollar consumer drones strapped haphazardly with a payload to more sophisticated drones. Where on that spectrum do you tap out today, and what complexity could you expand to over time?

What I'm going to go with here is more consumer electronics applications, and then I might talk about the application of defending against a wide-bodied aircraft, and people can assume what they want to assume there. But I'm not going to mention anything about higher-order targets beyond consumer drones, even though that is a possibility, and there have been some profound discoveries. Those discoveries are getting into dangerous waters, with wanting to say too much.

So here's what I'll say. The tail end of your question was what other applications and targets. Beyond drones, anything consumer electronics related will have a computer on board. Let's look at an outboard boat motor, for example. We took a smaller system, albeit a powerful one, out to a Navy exercise this summer and worked it against a bunch of boat motors, because USVs are becoming just as problematic in the Ukraine-Russian war as the flying drones. In fact, there's an interesting fact that was told to me not too long ago, that 50% of the Russian Navy has been damaged or destroyed by the Ukrainians, who have no navy. So it's no wonder that both Putin and Zelensky have decided to create drone branches of their armed forces.

USVs are just as threatening to the Navy as drones are to the ground and short-range air protection fight. And the good news is, Epirus is here once again with our force field. You push a force field out to an outboard motor and it knocks it offline, and the only thing that brings it back online is a hard reset, where you take the key all the way off and all the way back on. And again, we're talking hundreds of meters, not tens of meters or less. So you can imagine Coast Guard cutters or other ships defending themselves and other squadrons from these types of USV attacks, while also having it against close-in drones and other flying systems.

Now, one of the things I want to say when we talk about higher-order targets is a lot about the class of effect we have on the systems. The typical effect of a laser system is called a level five. That's where they burn a hole in it. There's no getting the drone back and rerunning it. It's broken and it's dead. Then there's the level four effect. Most of our effects, because of the high-energy effects, end up being level four, which means it requires a hard operator reset for the system to come back online. For a drone, that's a death sentence, because it falls from the sky and crashes and breaks on the ground. But for a boat motor, you can take the beam off, the operator can turn the key all the way off and all the way back on, and the boat will begin to run again. For a ton of applications, that's a better level of kill.

Level three, again, it's getting less and less. Level five is kill it dead. Level four is hard reset. Level three is a more subtle effect, and those can be almost at electronic attack ranges, where we're looking at kilometers and kilometers out, making subtle attacks on electronics that do subtle things. A good example would be an ISR camera, a commercial camera on a drone that's there to do some spying. We might have the ability to make the camera look like it's got white fuzz on it from many kilometers out, but the drone isn't going to take effect until it comes within maybe a kilometer or something in that range. So there are level three effects that can be pretty profound and begin to incorporate other targets beyond the motors and the wires and the ailerons of a drone.

So, yeah, we're after drones right now, because that's the threat that's most material, most threatening and most prevalent. But as we talked about with the software definability of the system, it has a complete ubiquity against any target with consumer electronics or any computer-like, sensitive electronics in it. Anything, even if it's militarized, is susceptible. There's nothing magic about the IMUs inside a seeker. They're not special IMUs. They're susceptible like a consumer IMU. All of these things could be potential target vectors for future systems, on board different types of platforms. You can think aircraft. You can think airborne. You can think space, and what's going on in space.

The breadth of the platform is akin to force fields. I keep saying that as a joke. I once had a person, I won't say who, to spare them any embarrassment, ask, "Is what you're doing at Epirus a platform business?" And I said, "Sir, with all due respect, we've invented force fields. Do you understand what a platform is? I mean, yes, it's a platform." It could be worked into space applications, airborne applications, surface applications. Subsurface, I've got some thinking to do, because the waves don't go so well underwater, except at super low frequencies, and we're not nearly low frequency enough to work underwater. But above water and up, we're good to go. Land, air, sea, space, all of that, we're good to go. So it's got a ton of applications, both in the military lane proper and outside it, like we talked about earlier.

You touched on the program of record in 2027. For listeners who aren't familiar with how the Defense Department works, how do you work with the DoD? You're putting your own money into R&D, and companies like Epirus, Anduril and Shield AI are pioneering a move away from cost-plus and taking on more risk. What has driven your success? And do you see any potential for partnerships with some of these other defense companies?

You may or may not have heard of this, but back in, I think it was 2017, the Secretary of Defense at the time, Ash Carter, came up with an idea to implement something called the Adaptive Acquisition Framework. It basically allowed the DoD, in certain circumstances, to flex away from the FAR, the Federal Acquisition Regulations, and do things with a bit more agility and nimbleness. Obviously they needed law and policy to do that, and they call that law and policy the Adaptive Acquisition Framework.

Within the Adaptive Acquisition Framework, there are a number of ways material can be bought by the DoD, in a whole bunch of different categories. The category we're in is a little different from the category Anduril is in, or the category Shield AI is in. The category we're in is what they call a mid-tier acquisition. What an MTA does is allow for a hybrid with a traditional defense program, which needs all of the rigor, all of the support: operations, sustainment, training, a whole ecosystem, laboratories. Think of a Patriot missile system. You need an entire Army ecosystem to support building, fielding and repairing a system like that. Our system, being a force field generator, is akin to that. It's going to need to find its way to the FAR eventually.

Systems like a Roadrunner, or other very sophisticated, incredibly impressive technology, are still considered dual use. And because they're considered dual use, a lot of their sales can be commercial off the shelf. They can't sell it with a weapon payload, maybe, but with a kinetic payload, those roads have been plowed. They know how to get Coyote payloads into country. They know how to get them exported. Those roads have been plowed. The directed energy roads have not been plowed as hard. So we're doing the plowing, through a process called a mid-tier acquisition.

Now, you're only allowed to do a mid-tier acquisition if you have a very urgent material request. By luck, the drone problem is so big that it issued a trigger that allows them to flex and do a mid-tier acquisition. We competed for the mid-tier acquisition in 2022. There were five other competitors, and of those, we won. We won the prototype shoot-off, and we went to the JCO to do this, by the way. I think it was General Gainey or something at the JCO. So when we were doing the JCO event, General Thurgood at RCCTO said, "I looked at all of the different systems out there, and we want to elect to go with Epirus."

That triggered the RCCTO, the Rapid Capabilities and Critical Technologies Office, to do a buy, try, decide process. We went through 15 months of try. I didn't know try would be so long, but we've run through the gamut. We've done everything: people safety, fratricide safety, DRAP processes, performance processes up the wazoo. Red team, blue team scenarios where they're separated, no one knows what everyone's going to do, and they're operating the system like they're in the Middle East. Now, as a final step, the cherry on top of the sundae, we bring it to CENTCOM and INDOPACOM and do final operational exercises so everybody gets the confidence that these things are as good as they say they are. Once that happens, the program of record is kind of a short thing, as long as all that happens right. We should know in 2025. We should really know in 2025.

You're also not just a defense company. Your smart power technology has a lot of applications. What are you mainly doing in commercial, and what are the future use cases for your smart power technology?

Sure thing. As I lead into that, I'll talk a little bit, real quick, about the second part of your question earlier. Palantir and Anduril very much are brother companies to us. We're in the family. We're the little smart brother, I guess, that knows a lot about microwaves. You asked about the linkage we have to Palantir and Anduril. Well, it's tight and it's strong. We're brothers fighting the same fight, not against the primes, but as a different category. We have to make room for ourselves. We're a different category, giving a different option to the DoD, and numbers provide strength.

Recently there was a leak of this sort of activity going on with Palantir, Anduril and some others. Well, we're very much in that brotherhood. That's our team there: Shyam and Chris Brose and Palmer and all that. We're very close personally. We work together. We talk about things. The lawyers at the various companies all talk with one another to figure out the best ways of going about doing this. And for the most part, that one voice boxed together becomes very, very powerful, on par with some of the big primes. I think that's a neat thing that began to develop in the last part of 2024, and I think you're going to see a lot more of it in 2025.

OK, now I'll pause and answer your second question. When you look at Epirus, in a lot of ways Epirus is a hybrid between being a neoprime and a traditional prime. If you look at the construction of the team, we look very much like a program office out of Raytheon or Northrop Grumman. We have chip and wire. We have system level. We have software level. We have the whole gamut, like a program office of engineers. And therefore, as we're assigned to these high-tech projects, we can do a lot of different things.

One of the things we've done while creating these systems is learn a lot about power storage and power conversion, especially as it applies to gallium nitride and high-powered RF. By learning a lot about that, we've made a lot of innovation and discovery in the transmit chains. Our first prime partnership has been with L3Harris. We're looking at the transmit side of their radios, because a lot of those radios get hot when you try to use them, or can't sustain a lot of battery life, or don't support certain signals because they're too complicated and the radios don't have enough efficiency to carry them. In all of those categories and situations, our technology solves for that. It solves for the coolness of the radio, the size of the radio, the weight of the radio. It makes more signals possible. It makes the battery life a lot longer. Those are things we are really global experts at.

From that perspective, we can sell modules with this technology embedded in them. As a commercial motion, that makes a higher gross margin business, to help support the net income needs required to maintain a position as a neoprime, where we have healthy gross margins and healthy net income, and we reinvest that net income into innovation. Because to your earlier question: yes, we've invested hundreds of millions of dollars into commercial IP. Yes, we have a Category XVIII ITAR weapon on the weapons list, a weapon system, but it has been commercially derived. We have commercial patents protecting the way we've done this and the way we're coming to market. It's not a military-rights system at this point in time. Now, some of the rights we gave to IFPC-HPM flex into the Army in a certain way. But the commercial system, the Leonidas system, is 100% derived from our own money.

So there's some uncharted territory we're in a little bit. I don't know what that's going to mean for the future. I'm not opposed to Raytheon and others helping out and building force fields alongside me. I think Raytheon, L3, a few other companies like Northrop, maybe even Thales, could do an excellent job with the right know-how on how to build these types of systems. I'm not opposed to more competition being out there. But the future is a new one. It's a new path we're on these days, and we're doing things a little differently. Now, with this administration coming in, I think we're going to do things a lot differently. So we may be in for some new ideas, some changing times. Change is always scary, but at the same time, change can bring about a lot of good. I'm looking for a little bit of both going into this administration, but leaning into the good stuff, not into all the scary stuff people talk about too much in the news.

That's a perfect segue into the last question. We always like to end by dreaming the dream, and potentially the nightmare scenario, of the future of the business. Maybe start with the key risks. Whether it's the countermeasures folks could be developing in response to your technology or whatever it might be, what keeps you up at night as you try to make sure your business is future-proof?

Well, one is IP protection. We need to protect our IP strongly as an IP-based company. All of our IP is Epirus-owned, and we need to own that IP. That's part of the differentiation between a neoprime and a standard prime: we are an IP machine. When you look at the finances of a prime, you'll see a top-line number, a gross margin number and a net income number. When gross margin is very close to net income, you know there's not a lot of self-funded OPEX in there. We need a good differentiation between gross margin and net income, because that means we have a lot of R&D OPEX, and we need to continue to fuel that engine.

My biggest worry is that we have such a supreme platform of a business that we're going to get sucked into a cost-plus type of world, and we might as well get bought by Raytheon at that point, or Northrop or L3 or GD, because we're going to look a lot like a normal prime. I think that's really the worst-case scenario: we look like a normal prime. We're at a point where this tech is so proven and so viable that I think we're going to win at some level. I'd consider it a bit of a loss if we end up having to bow and say, "Oh, yes, sir." And of course, if I'm talking to you two years from now and I'm getting bought, I'll be like, "Oh, no, this is a better idea." But for now, I'm telling you, tried and true, the best idea for me will be to stay independent, not to be similar to the primes, and to work with them very, very well, like I used to do. I used to be a prime. I know how to work with them.

But at the same time, be that 21st-century skunkworks. Not a Skunk Works or a Phantom Works under a Boeing or a Lockheed, shielded, but pure, straight into the customers: straight into the Department of Homeland Security, straight into the Army, straight into the Air Force. Have the ability to access a highly technical team that can do whatever needs to be done on an urgent issue. I dream of that. I dream of being, in a way, a Stark Enterprises or the Batcave from Batman, take your pick, where the government comes to us and says, "Andy, I'm in an urgent pickle. We've got this new thing the Chinese came out with, and we don't know how to solve it." "We're here for you, sir. We'll get it solved." That's the Epirus I look to build. That's the one I want to see in the future.