School of War - Inside America’s Fight for Drone Dominance

Episode Date: August 25, 2026

The U.S. military needs more drones, and it needs to build them without Chinese parts. Today, Defense Innovation Unit Deputy Director Travis Metz joins School of War to discuss how America can build drones quickly, cheaply, and on U.S. soil; what America can learn from Ukraine’s drone war; and the Pentagon’s new billion-dollar effort to upgrade its aging military technology. 2:23 - Why America Is Behind on Drones 4:32 - The Drone Dominance Program 6:51 - Breaking Dependence on China 10:23 - The Different Types of Drones 13:34 - Ukraine’s Mid-Range Strike Revolution 16:00 - Cheap Drones at Long Range 18:47 - AI, Autonomy, and Electronic Warfare 21:47 - How Drone Dominance Works 24:33 - Building an American Drone Industry 33:55 - Lessons From Fort Benning 39:19 - Cheap and Expendable Drones 42:01 - Testing Drones for Different Missions 50:14 - The Next Phase of Drone Dominance 55:09 - How Many Drones Does America Need? 1:02:49 - Recreating Ukraine’s Innovation Cycle This episode of School of War is brought to you by Maven Smart System, the builder platform for the American warfighter, fielded across the force and live on the frontline today. Learn more at palantir.com. Learn more about your ad choices. Visit megaphone.fm/adchoices

This episode of School of War is brought to you by Maven Smart System, the builder platform for the American warfighter, fielded across the force and live on the front line today. Maven Smart System isn't just software you use. It's the platform you build on. Every day, thousands of uniformed service members use AI coding agents to build their own agentic applications across all warfighting functions on top of the DOW's integrated ontology. What once took a contract in a year, now takes a warfighter and an afternoon. Administered by OSWCDAO, powered by the Palantir platform. One platform deployed everywhere at all classifications.

Build at the speed of war. When it comes to drones in the U.S. military, there are a few things basically everyone agrees on. Using unmanned systems of various kinds will be essential to battlefield success in the future, And America is behind, possibly very behind on this issue. Today, we're going to talk about a billion-dollar initiative run out of the Pentagon's Defense Innovation Unit to accelerate how we buy and how we use drones while building a supply chain for these things that doesn't rely on Chinese parts. By the usual Pentagon standards, this program is running at lightning speed.

We'll go into the weeds of the kinds of drones operated by the U.S., how many we should be buying, and how we should be using them on the battlefield. Let's get into it. It is for a war with a rocky invasion of Hawaii. December 7, 1941, a date which will live in him. A bloody experience of Vietnam is to end in a state. We continue to face the great situation in France. Hi, I'm Aaron McLean.

Thanks for joining School of War. I am delighted to welcome to the show today, Travis Metz. He is the Deputy Director and Chief Operating Officer of the Defense Innovation Unit, long career in investing in business. Before that, sir, thank you so much for coming on School of War. Thanks for having me. Long time, first time, as they say, in Boston Sports Radio. Wonderful. It's great to have you here. We are going to talk about drones today. We're going to talk about the drone dominance program, which you superintend, some exercises that are going on right now. But before we get into the specifics, Yep. I want to start bigger picture and ask you, this program that you're running, these efforts about essentially developing and buying drones for the Department of War, what is the problem that you are trying to solve as you see it?

So there are, so I came in the department about a year ago, and it has been the case for some time that there's lots of efforts on the part of the Pentagon, on the part of the services to address the need for, the obvious need for various types of autonomous systems. What the secretary determined last summer when he stared at the problem and when he asked Me and my boss Owen West to stare at the problem was that it wasn't going fast enough that the efforts in primarily of the Army and the Marines on quadcopters on the small versions of these on-man aerial systems They had they had lots of Cycles invested in trying to procure drones and figure out how to use them and stuff. It just hadn't gone fast enough and so the reality was if you see stared at both the use cases for drones inside of the U.S. military, and then you looked at the industry, and you said, based on what we see happening at other parts of the world, in particular

in Ukraine, which is this terrible, tragic, but live battlefield, with huge amounts of data flowing out of it, you would posit that even though the United States has a different force structure, is going to fight differently than Ukraine would and so forth, even conceding that point, we're going to need lots of drones, and we're going to need them in a way that is fundamentally supply chain resilient, and we're going to need them in a way where we have done the hard work of the dot MLPF and the doctrinal work and the TTBs and all the stuff the military does to take a system or capability and figure out how to integrate it into their plane of operations. And that had just, from the Secretary's point of view, just hadn't gone fast enough or far enough,

and so he decided to take a personal view about starting up this program called the Droner's drone dominance program. It was funded with reconciliation money. And he said to my colleague and I, go forth and figure out how to scale the U.S. industrial base for small drones and go out and figure out how to make it go faster in terms of actually giving these drones to the services so they can do the very hard work of inserting them into their sort of doctrinal stuff. We started with a couple of very basic premises. The most important premise is we don't know if we were at war based in the type of war how many FPV quadcopters we would actually need. The Ukrainians this year are going to manufacture and use somewhere in the range of 8 million of these quadcopters.

We don't know how many we need. We know that we need a lot more than we can manufacture today in the US. And the key assumption that we're making is that if we are at war, if we need to buy 100,000, a million, 10 million, or 20 million drones, we should not assume that our adversaries will sell us one piece part that makes up those drones. And the reality is the Chinese have a dominant sort of market share or did in all things related to UAS. So if you stared at the key components that go into a drone of any size, the Chinese generally had the low-cost position, They generally had the leading market share. They had the MAVIC, the DJI MAVIC, which is a, if you take off your sort of geopolitical

hat and you just look at it as a consumer product, it's an astonishingly good consumer product. Low price, incredibly capable. And obviously, driven by industrial policies that China has pursued for decades that put us at a disadvantage in a number of ways, but it had dominated the marketplace. So we fundamentally in the US did not have a scale industrial-based. around drones. And if we need drones at scale, my view is the Chinese won't sell us any of those parts. And so we have to figure out how to get from approximately zero. So yes, there are drone companies in the U.S. Yes, we have had good entrepreneurs trying to think that this problem through

and so forth. But when you look at the scale of those companies compared to what we're seeing on the battlefield, very small compared to what we actually need. So on this question of China in the supply chain, It's really interesting that this is this core focus of the program. I've heard it said about the war in Ukraine. You know, everybody knows that Chinese support for Russia, for the Russian economy, for the Russian war effort is essential to Russia's campaigns in Ukraine. But I've heard it said that actually if you think about it in terms of the Chinese role in the supply chain for drones,

as you just laid it out, in a weird way, they're supporting both sides because the Ukrainian drones, these millions of drones produced every year, I'm assuming a vast number of them rely on Chinese parts. So this leads to the obvious question. What is it going to take to achieve what you're looking for? And how much of a margin do you think the Department of War should be willing to spend on boxing China out? Because, of course, the advantage of the Chinese parts is that they're affordable.

Yeah, so two thoughts on that. The first is, so I'm just back from Ukraine. I was just back 10 days ago. The Ukrainians are actually far ahead of us. in terms of isolating themselves from the Chinese supply chains. So it is true. It was the case that when the war started, the sort of first examples of product used in drone warfare in any scaled way

was the Ukrainians basically taking Mavix and very mad maxi, strapping grenades and mortars on them, figuring out the fusing and so forth. But they decided a few years ago, sort of call it mid-24-ish, maybe late-24. that they have an existential risk in the same way that I worry about it, being reliant on Chinese supply chains. So they've done really hard work since then.

So most of the drone companies that you would go see in Ukraine have, they brought their supply chains onshore. Well, even for magnets. One of the things I want to talk to about is actually what goes into a drone, but I'm assuming there's some fairly high-end stuff that you're not building in your garage that you need at the component level. That's right. So have they done it fully?

they have not. So actually magnets is maybe the hardest problem to solve, these rare earth magnets where the Chinese sort of domination of both the primary mining and then more important the refining of those is hard to surpass in terms of the cost structure and so forth. But virtually everything else, maybe there's a couple of sensors where it's hard to figure out in a cost reasonable way to get away from the Chinese supply chains. But the Ukrainians, they have programs going across every single one of these components because they're at war and it's existential for them and they're thoughtful. And so they're staring at the problem the same way that we're trying to stare at now and saying, how do you do this? How do you go from zero to, you know, in their case,

if you had asked them four years ago, I don't imagine anyone would have thought that you would get to eight million drones in a year. It's an insane amount of capacity for these things. And yet that's what war drove, right? This sort of iteration of cause and a effect with the with the Russians drove them to use this you know the Ukrainians are working with all kinds of asymmetries right the dominant ones are lack of manpower and they don't have the high end exquisite stuff that's that's you know high up in the air and so drones were the answer for them and and they've scaled it mercilessly and primarily away from China so we're in many ways I think the the Ukrainian example is one that we're trying to do a good job of of leveraging all the

learnings from, recognizing that it's not this, it won't be exactly the same for us. Right. So we'll come back to these questions of costs and supply chains because they're, they're obviously essential, but for the moment, maybe you could help the audience. I think we'll probably have some people watching who are very familiar with this subject. Others know that it's important, but don't really know the details. You're focused on these, for the purposes of the program that you're leading right now, on these attack missions and smaller drones. How does the Pentagon think about drones? Like how does the kind of system that you're working on fit into the broader Pentagon scheme of what is a drone? So you're focused on the air and you're focused on

smaller stuff. What's the overall scheme into which this fits? Yeah, it's a good question. So the Pentagon, Department of War and the Pentagon is broadly focused on what we call UXS. So unman something systems. Those cover all the modalities, right? So we say UAS, unmanned aerial systems. UGVs, unmanned ground vehicles, U.S.Vs, unmanned surface vessels, and then UUVs, underwater vessels. And then there's a whole, you can think about it as being a whole tech stack that sort of is like cross-cutting across the autonomy. The drone dominance program that I've been helping run since last fall is only focused on a very narrow slice of that. There are lots of other efforts across all the modalities, and there's a lot of efforts starting to spool up across these different domains. So we're doing FPV quadcopters.

So those are, using my hands, sort of small. Right. Right. They carry about two kilograms of an explosive plus or minus. Today they go out 20 to 25 kilometers. In Ukraine, depending on the sophistication of the technology, they cost anywhere from $500 to a couple thousand dollars. They are attritable.

They generally are, we call them one-way attack. They generally don't come back at that sort of cost range. there are lots of different versions of that sort of, by the way, they're VTOL. They take their four motors, they take off and off they go. Generally in Ukraine, they've been flown, right? So there's a person with some sort of control stick, some sort of remote control, maybe goggles, that's generally piloting those, at least they were. But increasingly, the technology stack on those drones is taking over some of the cognitive

loads, some of the piloting requirements for the actual thing. So that's the FPV quad. There is an unmanned aerial system for up every echelon, right? So we're also, we announced and are running a competition for what the Ukrainians call a bomber dropper. So this is a group two-ish drone. It's maybe as big as your sort of wingspan. It'll drop 15 kilograms of explosive. It'll go out 35 to 50 kilometers.

It's loaded up with thermal sensors because it generally hunts at night. it is not necessarily attritable. It comes back after dropping its load. It's not $500 to $2,000. Maybe in Ukraine, it's $7,000 or $8,000. The dominant model over there, they call the vampire, the sort of Baba Yaga. The military units over there have literally hundreds of these hunting out at night.

And it generally is the thing that allows them to control the sort of 20-kilometer to 40-kilometer range. It's been big in the news. just this year, Ukrainian, if not dominance, certainly progress on these mid-range strikes on supply routes leading up to the forward line of troops. It's this kind of system. That and the next one up. Okay. So the next one up, what they, would generally when people say mid-range strike, they're talking about, so the Baba Yaga is a octacopter. Okay. So it takes off with eight propellers, and it's quite terrifying looking. Next up is what they call the mid-range strike. So that's a fixed wing. It's generally generally being launched from

a rail. It's typically propeller driven. It's not a rocket of any sort. It is, it can go out 100, 150 kilometers. It's got say five to 10 kilograms on the nose. It's one way typically. It has more technology on it. So it tends to have a combination of technologies that allow it to deal with all of the contested environment stuff over there. So the GPS denial, all the RF jamming, all the other sort of tools that they've developed in this sort of cat and mouse game. And that is the thing that is absolutely tearing apart to logistics in what they call the Kramian land bridge and sort of allowing these battlefield commanders to control, I think what they would characterize as their sort of deep battle space thing.

The key with those, and this is relatively recent on the part of Ukrainians, is they are inexpensive enough that they can mask them. So they're attributable. They don't come back, right? So the cost matters a lot. So they're building those in Ukraine, as I think I said, in the $5,000 to $7,000 range. So the ability to go out 100 kilometers and hit basically anything that moves.

So tanks, trucks, you know, people, and do that at that cost is a new capability that didn't exist anywhere as a year ago. That's literally scaling in the last few months. So for those of us who watch, watch, you know, sort of Twitter religiously to sort of try to understand what's happening in the battle space. A lot of the videos you'd see of Russian trucks being destroyed is happening with those fixed wings. Got it. And then the levels above that, we would be into the space of, like, predators and MQ9s and things like that, the things that are borderline airplanes,

not attritable. Is that right? So, so that would have been right. Okay. This is good. This is continuing education for me. Yeah. Well, I mean, part of what, you know, Ukraine's at war and against a very determined capable adversary. And one of the dilemmas of thinking about like the Department of War ecosystem is how do you create innovation and iteration if you're not at war? Okay. And the war is terrible. It's blood soaked. It's horrible. It's an eye-opening to be there. But it has created this, the perfect lab experiment for development. And so one of the things the Ukrainians have developed is what they would call these long-range one-way attack capabilities and long-duration ISR things. So you can think about this as sort of MQ9-ish on the one hand,

or you can think about it as maybe it substitutes for other long-range fires that we have, maybe the T-LAMs, maybe what the Air Force does. The key for the Ukrainians is they don't have the, they don't have a Reaper. Right. Right. They don't have F-22s and F-35s that can deal with contested air spaces. But what they do have now is these things that go out, we think the longest range ones they've operated have been out at like 2,500 kilometers. So that's like 1,500 nautical miles.

They put 50 to 100 kilograms of explosive on the nose, has a lot of technology in it to allow it to try to figure out where it is and so forth. And it's inexpensive. Right. It's the inexpensive part. And by the this is low and slow. It looks like an airframe,

It's got the sort of weird inverted tail if you see videos of these things. And they're, I think they're at like 200 meters maybe. They're going 100, 150 knots. They're actually not that survivable in a setting where you can mask the right amount of sort of counter. But they mask them for these attacks. And they don't cost $15 million a pop. No, they cost, you know, they cost tens of thousands of dollars. Yeah.

And so the Ukrainians hold very close to the vest. the survival rate is for those. It's not 100%. They don't, they don't, they don't, they don't, they don't, no one believes it was intended to be 100%. But the key is that tens of thousands dollars, you can mask them in hundreds for these strikes and you do overwhelm the capability of the, the low anti-aircraft systems to figure out how to defeat them all. And so if you get through, so when you see Russian oil refineries burning in like St. Petersburg area, east of the Urals, when you see these, they've gone after these warehouses, the wild bears.

Yeah, spectacular videos that are out there. That's what this is. Yeah. And then there's another version of that that they've modified, that it's actually taking out ships in the Sea of Azav, I think. So it's a slight less range because you don't need the range. And you've got to put a lot more explosive in it to matter for a vessel. Just sticking with this for a second, to your knowledge,

the guidance systems are just how command and control works in the terminal phase of these missions. presumably the electronic warfare environment when you're close to a Russian warship are one of these high-value targets, industrial targets that the Russians know are someone's going to take a swing at at some point. It has to be a hard place to control something from as far away as Ukraine. So is this all programmed into the system itself for a set target? Is there some sort of AI system that's making choices? What is the state of the art right now in that war? Yeah, all the above. Okay. So a lot of times, and you see this, with, I don't want to suggest this is only happening in Ukraine. You see many of the American

companies that are either in Ukraine or even just not in Ukraine developing these types of technologies. So you can think about them as one end of the spectrum, it's what we call pixel lock. The other end of the spectrum is sort of more of like ATR automated target recognition. They have the same effects, which is you're using technology or using sort of machine learning AI capabilities to guide the thing in the terminal phase. because the assumption is that for most high-value targets, the Russians will have put up some type of EW bubble, where you will lose the ability to fly the vehicle down to the target.

We saw this, so we ran our first testing event for the drone dominance program for the quads at Fort Benning in Georgia in February. And we went out as far as, I think, 10 kilometers and into the Georgia Pines. So turns out, that was, We did not run any explicit counter UAS there. We did not bring any RF jamming. We didn't bring any other types of countermeasures. We are doing that in phase two,

which just started this week out in Fort Carson. But the Georgia Pines, 10 kilometers away, turned out to be a really good RF shield. Oh, interesting. You can't really maintain, unless you actually put a repeat, some sort of repeating device up. Yeah.

It's actually really hard to maintain RF at 10 kilometers. So beyond visual line of site with the trees. Yeah, of course. So if you put the drone high enough, you can keep talking to it. As soon as it would enter, but the test was to hit a target.

Right. And as soon as it got somewhere below the tree line, you lose communications with it. And so the companies that did the performance the most highly in Fort Penning were the ones that had a version of this technology on the edge. So it's not back at the home station. The drone itself has enough compute stack

that the operator said, This is the thing I'm trying to kill. If you lose comms, you fly into that sort of thing. Yeah. So that's happening writ large. I mean, that's happening across, and it's trying to solve all these problems that you described, where the counterparty of the Russians in this case

are not allowing you to, or very much trying to disturb your ability to command and control the thing at the moment of truth. Well, let's talk about this program that you're running, the drone dominance program, and then the exercises that you're referring to are the gauntlet. That's right.

exercises. I've seen you quoted in another venue saying that the pace of this thing probably seems slow to a certain kind of business, but is lightning speed for the Department of War. What is the basic idea of, we'll stick with the first exercise back in the woods of Georgia. What were you trying to accomplish with that phase of the program? How did you get from the starting point to there? And then we'll talk about the phase right now. Yeah. So if you go back to what the mandate from the Secretary of War was last summer, last fall. It was scale the U.S. supply chain and the way that we determined for small UAS. UAS.

And so the way that we determined to do that was to use orders for quadcopters in order to try to shape the industrial behavior we wanted to see. And the idea was to use basically advanced market commitments and a sort of rolling set of gauntlets, as we decided to call them, to send a signal to industry, a long-term demand signaled industry so that industry would do the thing that we would prefer them to do, which is help us build a U.S. supply chain for drones so that if we get to war and we need to buy $100,000, $1,000, $10 million, we have something to work with on that front. So we set up to the particular mechanical framework was four gauntlets, each one were increasing

the volume. So at the end of every gauntlet, we buy drones. Okay. And people have to deliver those drones in the following six months. Every gauntlet, we increase the volumes that we buy. We reduce the prices that we're willing to pay, and we increase the difficulty of the testing that we're doing. So we're looking for more capabilities. And the idea was to see if we can get industry to work with us on this treadmill of volumes going up,

prices coming down, capabilities going up, and the restrictiveness of our supply chain requirements going up. So if you said to, you know, if you had said to us last fall, we want to buy 100,000 drones, U.S. drones. no foreign parts of them at all. It was not available to, it was not available to us. It's not available to us today either. Doesn't matter about the price. So we are always, I hope we are always going to pay more for our drones than the Chinese do, because if we're paying the same price as the Chinese,

it means something's gone really henky with the labor rates between China and the U.S. Right. It is a damning observation, though, that available at no price, that number in 2026, when the issue that we are discussing, even though I'm learning a lot from this conversation, this is not news in 2026 that this kind of technology is going to be central to war fighting for some time to come. Yeah, here we are.

Completely agree. And by the way, there are hundreds of really talented, smart folks in the services that have been on this problem for far longer than I have, and they've been working through the system to make this happen. The Secretary's observation, which I think is profoundly right, is too slow. So I, the secretary, I'm going to turbocharge it with this dose of reconciliation money and we're going to pull it out of the services so that it serves as a bridge until they

are able to have it in their normal, their sort of normal procurement processes. So our program is effectively a bridge until the services have the budget and the supply chain exists for them to buy these things at scale. The key is at scale. Like it matters a lot whether or not the drone is $5,000 or $25,000 or $50,000. to point out the obvious, right? You can buy twice as many when the price is half. Pentagon's resources, we're so wealthy as a nation, but the Pentagon's resources are finite.

And so the sort of price matters on these things. The supply chain, so we imposed basically, so we have all these regulatory regimes or statutory regimes. The primary one is the NDAA, the National Defense Authorization Act. Yeah, it's funny. I don't even know if I ever knew of that, It prescribes where various munitions or systems that the Department of War buys. It prescribes where they can come from.

So the NDA has a whole theory about drones. It has had for some time. But if you're trying to build a supply chain that can give you millions of drones with no parts from China, the NDA is actually not sufficiently restrictive. And so our first phase, so Benning, we said all the drones we buy have to be NDA compliant. using what's called the Blue List. But we then imposed coming out of Benning, I'll talk more about what we bought coming out of Benning,

we then published our own policy document, we call it the supply chain framework that basically puts all these companies on a journey to a fully U.S. drone. So the policy of the administration, the policy of the president, is that we are building an American drone industry, one that is resilient,

it's American and it's resilient so that we can buy sort of millions of them. And so that document, that framework, It's not statutory. It's just a policy of the Department of War, basically has sort of increasingly difficult gates or increasingly difficult requirements in terms of where can the brushless motors come from. So every quadcopter has four brushless motors in it. The NDAA allows you to buy them from China. In the long term, we need to build brushless motors first outside of China and then eventually in the U.S. Okay. So our phase two, the one we're running now, does not allow China. brushless motors. So that'll be the, that's the first time in the Department of Works for purchasing history that we're saying to them, I think it's the first time, we're saying the marketplace,

you've got to sell quadcopters, but they cannot have Chinese brushless motors in it. Chinese brushless motors are the dominant brushless motors in the world. I don't know what the market share is, but it's somewhere in the 90 percentage range. So it is in that supply chain framework that we're trying to drive the behavior we want sort of in the industry. This episode of School of War is brought to you by Maven Smart System, administered by OSWCDAO, powered by the Palantir platform. Maven isn't an app. It's a builder ecosystem. Right now across the joint force, thousands of uniform service members are pointing AI coding agents at their hardest problems and building the tools they need themselves, targeting workflows, Intel products, battle management.

No ticket, no vendor, no wait. The warfighter closest to, the fight is now the one building for it. That's how you fight at the speed of war. So this all raises sort of an obvious leap of faith issue for industry. This is a really interesting program. We're going to talk more about what you're doing. You have about a billion dollars. This is going to run in its current iteration for some time. But as resource right now, It could be finite. So if I'm industry and I come in, I'm a small company, and big company, it doesn't really matter.

And I start to build the supply chain I need to meet these increasingly raised bars for ultimately onshoreing the building of these different components. Either I'm doing it myself or I'm buying it from somebody. The leap of faith, right, is that at some point, years down the road, when this program, in its current iteration, meets its end, and it's succeeded by something new, that whatever it's succeeded by won't suddenly lower the bargain

or won't suddenly have a different set of standards, and all of a sudden, I've built something that actually costs a fair amount of money because I'm contorting myself for good national security reasons to meet these standards, but some future administration, some future policy pulls the rug, essentially.

How do you address that? Yeah, it's exactly the right question. So when I talk to industry about this, One of the things I say, and I think it's right, but the industry has to make their own, effectively their own underwriting determinations as to where they want to flow capital to, is that the Department of War is likely the biggest buyer of autonomous aerial systems in the world in the near-medium term. Certainly once the Ukrainian war ends, hopefully, blessedly ends, I think it's fair to assume. that notwithstanding the observation, which I think is fair, that we've been slow on this front,

we're now after it. And the United States has the biggest defense budget in the world still. And we're likely to be the biggest buyers. And so it's the total, you know, I believe it's reasonable for industry, notwithstanding the fact that I have a, you know, we have a set of finite funds around this particular drone dominance program to believe there's a very large total addressable market in the context of not just the quads, but all of the systems. that you would use in this sort of aerial space.

And I think it would be, I don't think it's a big leap of faith for them to assume they're gonna maintain the same restrictiveness around origin of components. It would be very odd for us to say we're now willing to buy these from potential sort of peer adversaries. I don't, I think that's easy to,

I don't know that I've ever had to say that, but I think it's pretty easy for people to actually underwrite against that particular problem. And the reality is, and this is the great strength, right? So we can criticize and complain and so forth about the fact that we have been slow on this. But we have unbelievably deep capital markets, private capital markets. We have venture capital investors who in the last few years have moved huge amounts of capital

into defense technology businesses across the domains, not just in quadcopters. And we have this entrepreneurial class, which is deeper than any place else in the world. And so that combination of capital and this sort of entrepreneur, spirit and what we're trying to do, which is give people a long-term demand signal and a framework for what we're after, I'm an optimist about that. I think it actually causes the effects we want. I would suggest we've only been out this. So we ran our first, so you asked the question before about the history, right? So we ran our first. We announced it in November. We issued an RFS, a request for submissions in December. We got 200 applications in January. We brought 25

companies depending in February. We competed them over three, two and a half, three weeks. We ran like a thousand missions there, I think. We placed orders in March, March, April, and we've taken deliveries since then. So we bought $150 million worth in drones. We're going to buy 60,000 drones on September 7th coming out of Fort Carson, which is Gauntlet 2. The price has come down. The capabilities have gone up. The supply chain framework has made it more restrictive. But we're going to buy, and we're going to buy all of those weaponized, or virtually weaponized. We're going to spend something like $400 million of that. I think it's anecdotal in its early days, but I think you can see the effects of people

reacting to that, properly reacting to that demand signal. You've seen huge amounts of capital flows into how do you build brushless motors in the United States? How do you do battery packs in the United States? How do you make all these, you know, one of the more expensive components on the bomb, and the FPVs is a thermal sensor, something that allows you to see better at night. You see all these startups,

you see all this sort of entrepreneurial class, you see a lot of the scientists coming out of labs, thinking through how to solve these problems in a US manufacturing context. It's messy, I suspect a bunch of them will not succeed, and yet this is what we do. This is what the sort of American capitalism is actually good at.

Yeah, so I've been asked before, like, how much will cost, and how long we'll take and so forth. And I beg off of all those things because I think it would be folly to predict any of that, except to say, we're going to place orders and then we're going to let the machine,

the American sort of economic machine, try to do the hard work of actually, like, building this up from, like I said, what I think was approximately zero before. So shifting for a moment from the supply chain and business side to the actual gauntlet exercises themselves, and these large numbers of missions that you're running with the competitors,

put some color on it for us. What worked, what didn't work at Benning? You've already given some indication in terms of your sense of terminal guidance, being able to survive once you go into the trees. What came out of that first gauntlet exercise that was characteristic of a successful competitor? And then for this second exercise that is running, as we speak, in Colorado,

what do you think is, what are you looking for out of that second exercise? Missions that were running, so like at Fort Benning, the missions were running were designed by the armed services. So in this case, primarily the Army and the Marines. The Army and the Marines of the services, those are the two components that are likely to consume quadcopter FPVs at scale. So we had the Army and the Marines and Sok on the special operators there to actually operate the drones. So they designed the missions, they design the testing.

We are lucky enough to have the Ranger Regiment there at Benning, and so we were sort of on their home turf to be able to sort of use their facilities and use their expertise around drones. And we gave each of the companies basically a morning to train the uniformed operators. So these are drone pilots from Army, Marines, and Socom. We gave them basically a morning to train on these things. So you're building into the program a kind of,

I don't know how to say this, but like a sort of semi-competent operator. Or are you already training people who are experienced drone operators on the system? I guess that's the question. Yeah, it's good question because it was actually one of what I think is one of the mistakes or the failure modes that we made on it. So we're using experienced drone operators.

We probably used at Fort Benning drone operators were too skilled because we do have very skilled drone operators, in particular in some of the elite units in the Army and the Marines and Socom. But if you think about the future state where we're gonna buy 100,000, a million, 10 million drones, the reality is those drones will be flown by young soldiers, young Marines,

who have, it's probably, not an MOS, it's probably not an occupational specialty. It's probably a schoolhouse they've been to. Maybe they've had three, four weeks of training. Maybe they've had a bunch of evolutions, but it's not someone who is a top gun pilot. Whereas actually some of our socom operators with drones are absolute world-class wizards at this stuff. Right. Like it's really impressive, but they're not the ones we're actually trying to build the drone for. I can imagine if I were a vendor showing up to compete, there must have been some sort of market, some sort of competition for the pilots, right? Because you're going to perform better than more...

Not a good question. We sampled this to make sure we had different skill levels. Okay. Interesting. Yeah. And so in that we did it, we realized it soon enough at Benning, but we explicitly planned for this at the event that's going now in terms of having the right mix of pilots for this stuff. But the other key point is if the system, a, the way to, one way to think about these FBs is, if you think about how the Ukrainians have caused effects with these, it's generally not just the material. It's the entire ecosystem. It's the entire con-ops around the material.

It's all the dot-ml-pf stuff, right? So you've got to have the drone. You've got to have the pilot. You've got the pilot had to be trained. You have to have the logistics tail that got them there. You have to have intelligence and censor. All the stuff that makes things actually have effects at the sort of tip of the spear. One of the things that matters is how hard is the drone to use?

It gets back to your question around the autonomy and the sort of pixel lock and so forth. And one of our, you know, base beliefs is the longer it takes you to train the sort of average Marine or soldier on the thing, the less good your product is. It's an, the nature of your product is defined by how hard it is to use in a way. And so we specifically limited the amount of training that we allowed them to do. And then our soldiers and our Marines ran the missions. So they actually flew the drones in these particular missions. What did we learn?

Well, we learned that the Georgia Pines are good at blocking radio. We learned that they're for the like the CQB, the close quarter missions that we were testing on. And this is probably obvious to most, but we proved it. It's very hard to do fixed wing. So we had some fixed wing. So one of our keys was we actually don't want to be prescriptive about what the solution looks like. We want to describe a problem set and have industry iterate to solve it and take advantage of that. we don't want to do classic requirements here.

So the problem sets were go out 10 kilometers and hit a dinner plate, hit a small target, or go into a building, go two rooms into a building, and kill someone. Those are the two broad problem sets we set. So fixed wings actually at low cost, because we set fixed prices for these drones. So when people submitted, they were agreeing to sell us the drones at those prices. So it turns out 10 kilometers on a target set, fixed wings, even though they're slightly higher velocity can do a good job. You can't put a fixed wing in two rooms into a building. So as much as we want to be sort of pure around problem sets, it turns out that form

factor actually matters here on this thing. I think the other thing we realized was these are, you know, in Ukraine there are a couple, you know, $500 to $2,000 here. Our first round was $5,000. Our second round is $4,000. These are attributable systems. The, they do not have a success rate of 99.9% like we expect out of almost all the munitions that we actually have. People disagree about this and Ukrainians have the most data on this. What is the base failure rate of an FPV? That's a great question. I have no idea. I think it used to be like 10 to 15%. I know, this is pure swag on my part. I think it's in the 5 to 10% range. So just the battery clip comes off, the solder didn't work. And you say, well, we should buy higher quality ones. And I would say,

how do you know? It's all about the trade-off on price and quality. So it turns out, just to state the obvious, right, $5,000 drone, 10% failure rate, and you say, well, I want one that fails only 1% of time. And the industry says to you, okay, you can have that drone, but it's going to cost you $50,000. Okay, I don't want that trade. And this is where the Ukrainian experience is instructive. We should draw conclusions from this. They've evolved to a certain place on the price quality curve. So why do I mention that? Well, if they have a high base failure rate, you actually need to run, a bunch of repetitions to sample the, to properly get to the ground truth of how effective is this drone. We need to run more missions, basically. So Fort Carson, each company, so

19 companies are showing up at Fort Carson's, I think, starting yesterday. Takes us about three weeks. We run a few companies for every day. Each company brought 120 drones. Okay. So, and these are tradable drones. The intention, we'll see whether or not how many we destroy, but the intention is to destroy all of those drones along the way, meaning run missions with them so that we have enough repetitions to have to take a view as the statistic that we've properly sampled the ground truth right yeah that's the that's the actual goal and that that imposes if you have to show up with 120 drones um i guess they got some capital in by being successful uh in for benning nevertheless you you're not building 120 drones out of your garage you you have to

have some actual resources you have to have a real at least the ability to scale a bit that's exactly the intention right is that we we as much as we admire the hobbyists who are tinkering away, we want to scale the U.S. industrial base to millions, and so we are after the companies we can partner with who can do that. And one other point. Turns out Benning wasn't a downselect. We didn't, you didn't have to participate a win at Benning to get invited to Carson. Oh, interesting. Okay. Yeah, because part of it was we laid out, when we originally announced a program, we laid out each of these phases and said, if you don't believe that you have a quad copy that compete at Benning, but you do believe you have the industrial capabilities to be a long-term

player in this space. We would like you to build a quadcopter and come to Fort Carson. Okay. Interesting. On the operational side of things, you know, you describe these two things that these two kinds of exercises at Benny. I don't know what's going on to Carson, but you have the Urban Challenge and then you have that, it's not long range, but, you know, long range for a squad, you know, five to ten kilometers out. Aren't those two such different problems that you might just, in the end, conclude that actually it's fine to have two different. kinds of systems for it. That is to say, if I'm a squad leader, I'm very concerned about what's in the building next to me. I'm only margin, well, maybe this shows how dated I am. I'm only marginally

concerned about what's 10 kilometers away. If I'm a company commander or battalion commander, I might be very concerned about what's 10 kilometers away. And so it's okay if I have the quadcopter that's maneuverable, but I can't really do much with it 5,000 meters away as a squad leader. but my company organic assets or my battalion organic assets, I have this somewhat fancier but still cheap system that can operate out there. In other words, did you think about maybe shoehorning all these items into these two missions might be imposing an unnecessary layer? That's exactly right. So at Carson, we actually bifurcated it into two separate purchases. So we have a CQB purchase and a long-range purchase. So we will pay slightly less for the CQB. It's a

slightly less capable platform. It doesn't need to have the same battery duration. You can't take two kilograms into a building and comply with like law of armed conflict stuff. So it's just a smaller looking drone that can actually, and by the way, it's more, it needs to be more maneuverable to go into multiple rooms in a building. And then the longer range one has bigger battery packs, bigger explosives and so forth. The question, the interesting question that intrigues me is, So we're going to go out at Fort Carson. We'll run multiple missions in an urban terrain setting, hit someone on a roof, hit someone in the alley,

multiple rooms into a building through a small environment. On the long-range side, we're going to go out 15 to 20 kilometers. We're going to do it with contested, so with RF jamming and other contested measures using the group in the Pentagon that does fundamentally responsible for counter UAS. this group called the JADF 401, the joint interagency task force 401, and try to knock them all down.

And so we just ran a qualifier. So we did a physical qualifier for the second phase at Camp Grayling in Michigan in June, and we brought Jada 411 out there with a bunch of vendor-operated counter UAS, primarily RF jamming. Do they get contracts if they do better? You could run two exercises

and collision with one other. Yeah, not from the DRON dominance program, but from General Ross's program at Jadda, yeah. That's part of the point, which is that what you observe in Ukraine is that it is the measure, countermeasure, which has caused them to develop their capabilities so fast. In a perfect world, we would be doing that here with Jadif, with our, you know, so counterposing the defense of and the offensive capability.

So we're working clever for it. So they'll have a whole team, they have a whole team in Carson. We brought a bunch of systems over. We were very careful not to tell any of the vendors, on the offensive side, what systems were bringing on the defensive side, you don't get to know that in a contested in a true battlefield environment. I'm not sure if this is a dumb question or not. I mean, I was in Ukraine, but it was about six months ago. So my knowledge of the state of the art is dated at this point, because as you and most people watching know, I mean, the cycle there is

extremely fast. Wouldn't the way to circumvent a lot of the counter-UAS stuff just to be for every entrant in your program to be a fiber optic cable guided system? And if not, why not? So yes, and we are testing for fiber optic capabilities, but there are plenty of missions where the fiber optic cable creates disadvantages as well. So the assumption is that you need, there's not one drone for the sort of full range of missions. You need to have these various capabilities. Most of the FPV manufacturers in Ukraine and most of the strong contenders in the U.S. can bolt that capability onto their particular drone.

Okay. Yeah, so it's just a different module, it's like a modular package you can actually use. You know, one of you, the company that performed the most strongly at Benning was a fiber optic drone. Going out 10 kilometers and maintaining comms is trivial. Right.

As long as you don't break the link on that fiber optic thing. And that's, that does impose some complexity on the, on the operators and on the drone package itself. but they solved it. The particular company had solved it at Benning. So we'll test for that at Fort Carson as well. Got it. And then none of these vendors are making the payloads. The explosives are coming from other vendors. What are the advantages and disadvantages of that? Yeah, well, I'm a free market capitalist at heart, and so I sort of expect to see people specializing in what they are best at. And so it's not, I don't find it particularly surprising that there's a different set of people in making the ESSADs, the electronic safe and arming devices, the triggers, and the sort of payloads.

We don't have a scaled manufacturing base for that in the United States as well. We have a couple, we have in particular one very good strong company that has done a good job of getting through the various safety release boards and all the stuff that we properly impose on companies. if they're going to sell munitions to the US government and put them in the hands of our soldiers and Marines. But we also can't sustain being single-threaded through one vendor for this. So one of the things we did as part of the drone dominance program is run a parallel, it was a prize challenge,

sort of term of art in terms of procurement, but to find four other vendors that have a combination of an ESAD, the trigger, and the pay and then married them up to, made them available basically to the other drone vendors so that we can start to put that demand signal out more broadly into the ESAD and payload space with the same assumption that if we place orders

for ESADs and payloads, it will drive the price down, it will cause innovation. People will iterate on this in a way that we don't really know how they'll make the price lower. We don't know what capabilities they'll build into it, but they'll do it because they had the incentive. to actually do it. You said something a couple minutes ago regarding explosives that I don't understand about the law of arm conflict and the amount of explosive you can send into a building.

Speaking of somebody who had a prior life participated in the dropping of 500 pound bombs on buildings. What do you mean? Yeah, and I'll asterisk this by saying I'm not an expert in law of arm conflict or the other weapons safety issues, but there is at least a convention, I think, that says you should use proportionate force at any moment, I think. Sure. And so if you're at the squad, if you're in a building and you don't have full understanding of who else is in the building other than the bad guy you're trying to kill, you should not bring down the whole building. I see. And two kilograms of explosives often brings down a whole building, and it is not required to kill someone in a room. I see. And so you-

So tied to the specific mission that you were, okay, I think, I think I had a picture. Yeah. And so So that's point one. Point two, a groan that can go out 15 kilometers and kill a tank or disable a tank or kill a vehicle or kill anything really that's, you know, within 30 miles of the 30 kilometers of the plot is a reasonably big quadcopter. It's reasonably unwieldy. When you're in a building, you have a different set of requirements around space and transit and so forth. So you just, by the nature of the mission, you need a smaller vehicle, a smaller airframe to get through. That smaller airframe, by definition, will have less payload capacity. So you just end up with a different sort of payload.

Okay, so walk us forward from the Fort Carson gauntlet, too, that's running now to the other major mile points, milestones in the program as currently conceived. And then what I'm really interested in kind of goes back to what we were discussing earlier. In your perfect world, how does this pot of money? that you have from reconciliation, it will be expended at some point. How does all of this land, everything that you're working on, and get either preserved or reintegrated or some combination of the two, back into the ecosystem of the Department of War, in a way that's maximally helpful?

Yeah, so we'll buy 60,000 drones weaponized, so all with ESADs and all with some type of payload, either a training payload or an actual energetic payload. coming out of Fort Carson. So those will be, we'll place those orders first, second week in September. We don't know how many companies we're going to place orders with yet.

We'll know that after we see the results of the testing at Fort Carson. It'll be less than the number of companies we purchased from in the first round. We purchased from 11 companies in the first round. The goal here is to sort of narrow down to the core set of companies for whom we want to,

with whom we want to partner with over the long term. Those companies, when they take those orders are obligating themselves to deliver over the following six months. So in phase one, and we'll repeat this more or less in phase two, they had to deliver 50% within three months and 50% within the remaining 50% within six months. There are financial incentives to delivering faster, penalties for delivering more slowly. The folks that delivered more quickly in phase one, we reserved a pot of orders that they could then go back to, so call them bonus orders, actually issue.

issued the first bonus order last week to a US company. And so we're trying to create all the dynamics to sort of incent people to basically reward people for the behavior we want, which is scaling. So people are scaling. Sort of manufacturing is hard. And then we will rinse and repeat again. So right now, February of 27, sort of end of February 27 will run Gauntlet 3. It will, again, the goal will be to narrow.

the aperture further in terms of the number of companies for which we're placing orders, with whom we're placing orders. We will go up in volumes. We aspire to go down in price on both the system itself and the ESADs and the payloads, and we will continue to increase the degree of difficulty on the mission sets. So we would expect to impose meaningfully more contested things so that the drone needs to be better, technology-wise. The pilot needs to be less loaded cognitively with the challenges of the contested environment. We may test things like, can one operator operate many drones?

Can you do the drones have any ability to communicate with each other, these sort of notions of swarming and so forth? We're still working out all the details, but it'll be, suffice it to say, we'll be, our goal would be to evaluate based on to evaluate such that a drone that is on the journey to what you can imagine these drones must look like as technology progresses. We're basically incenting those particular companies.

So that's about compute. It's about sensors. It's about getting those things at low cost because we're going to drive the price down again. We'll order, right now the plan is to buy 100,000 drones in February. We'll order those March deliveries over the following six months. and then we aim to do the same thing again in August of 2027. This is all a bridge, right?

So the Army and the Marines in particular are, have now begun putting scale budget allocations against FPVs and quads along with other unmanned aerial systems. We're still, we're just still talking about this one narrow thing. And of course, there's all this activity on these other things.

And we would expect to transition this to their programs of record as we sort of deplete the funding. from the sort of reconciliation money. Part of the benefit of actually buying the drones and sending them to these units and having the units use them, actually attrit them,

is that it accelerates the ability for these units to do the training, the doctrinal development, the TTPs. We're giving the material to actually work with in the training setting. And part of the deal we have with every unit we're sending to is if we send them to you, you must use them.

You must have trit them. And so we're sending them all across the military. So I think we're out to like maybe 25 different actual military units between the Army and the Marines and Socom. This leads me to my next question, which is one of a couple questions that are, they're live in my mind, and you'll hear even as I ask it that I don't fully understand how to ask it,

but you're the perfect person to sort of help us think these things through. Ukraine is buying millions of drones a year now. And then it's using a great number of them in combat. How many drones should the Army and Marine Corps be buying in, you know, 2027, 2008, assuming that we are not in the kind of conflict, but we're using a lot of these sort of quadcopter style FPV, one-way attack drones at scale. It's, for me, it's a hard question to think through because the Ukraine case,

well, yeah, you're buying millions of drones because you're using millions of drones. We're not going to use millions of drones in combat, probably, we think, we hope, in 2027. But you have this issue, as you know, and as you're working on, where whatever the manufacturing capacity is, we'll need it to be scalable very quickly for such a moment. Is the comparison here fundamentally the purchase of munitions? I can keep going, I will keep going, but I'll leave it there for now and I'll come back to it. I think it's a good analogy. So the Army and the Marine.

believe and I believe I won't say the exact number because I think it is classified but they need hundreds of thousands of FPVs to train on properly so when you do the math at the squad platoon you sort of work your weapon you think about what it would look like to train on these because you do need to a trit them to properly train on them that's the key thing right they are like one-five-five's they're like flying claymores like if you're not actually training with the thing because it's precious and you have to fill out paperwork because you lost one, then we're failing.

Been there. Yeah, right. With a much more expensive situation. Yeah, so one of our deals with these units is you can't put them in the inventory. You can't serialize them. You can't have to report if it ran into a tree because the goal is for it to run into a tree. So hundreds of thousands of these things just for properly training purposes. There's disagreement about where in the hundreds of thousands that lies.

but our goal is to by the end of the drone dominance program is to have made purchases in that sort of scale and that will have created if we're successful in the supply chain framework that restricts where the drone's components come from come from it will have created a U.S. industry that's capable at least of doing that one of the one of my beliefs is that it is much harder to get from zero to a few hundred thousand than it is to get from a few hundred thousand to to a few million to twenty million and so forth once you have the muscle memory and the rhythm and you have a supply chain which is building brushless motors in Orlando or New Hampshire wherever or San Francisco Bay where someone's just started up it is much more tractable to just expand those

I know what it looks like to add another line and add another line, add another line, as opposed to building that first line. And I think that's true across all the sort of component levels. So we, I'm artfully dodging most of your question here, by the way. We don't, it is very hard to figure out how to have an industrial base that makes good profits at building you, because we want them to be profitable. We want this to work for them to make us, say, 200,000 of them and have the capacity to build $2 million or $20 million.

Is the government paying for optionality on that capacity? It's a very difficult question to sort of answer. We're relying on the fact that if you've gotten people to $200,000 and you continue to have like a long-term buy in place which will have to, that

industry will be able to get us to scale sort of very fast on these things. What you cannot do is you cannot warehouse these things. That's the problem. It's not like a 155 in that sense, right? The iteration, I think even the Ukrainians, Ukrainians send everything east, right? Everything that comes off of their manufacturing lines basically gets flown. But it is the case in Ukraine that the drones

that they would have come off a line, I don't know, six months ago would be less competitive, less good than the drones that they're making today. I'm going to push you on that, but first, a brief anecdote to show what life was like. Maybe life is still like this in some places. But the year was 2009. I was a platoon commander. The place is 29 palms. My platoon had done its work and the exercise we were engaged and was finished. So I and my squad leaders were having a beer at the local camp. Marines drink beer. This was a place that sold beer to Marines for the battalion executive officer, Major Henderson, maybe he's out there listening somewhere, spotted us. Apparently the battalion was not done with its exercises until the following day.

So in this gray area, I had seized the initiative. I knew with one look from him that he did not approve of my seizure of that initiative. My reward was to conduct the investigation for another company that had lost its drone out in the desert. I can't even remember what the name of the system was. It did cost, I do remember it cost about $30,000. Feeling that my career was on the line here, it was like I was the NTSB investigating the loss of an airliner. Like this document that I produced at the end of the day was like 50 pages. I investigated wind patterns.

It was enough. I mean, they had flown it too far. It got dark. They couldn't figure it out. It ran out of battery and a crash. That was the whole story. We never found the thing. But it was a big deal within the battalion.

And I think that company commander felt uncomfortable with the whole circumstance, as you would. I felt uncomfortable doing the investigation. But that was the ethic. That was the culture of the day in 2009, perhaps reasonably when you lose something that cost $30,000. dollars, that obviously can't be the ethic now for the kinds of systems you're talking about. Yeah, I think one of the reasons why there's an imperative to simply bring down the cost of these. So, like, by, we believe the fundamental way you bring down costs is through the cost

curve, order more. Right. Like order more properly. Like set the right conditions to drive and distributing the cost down is so that you can entreat them. Right. It's so that the device that was 100,000, so you never flew it. So you didn't train on it. And if you don't train on it, it's not going to have effects when when you have all the stresses and so forth of war that that thing is now $20,000 and now you're willing to invest training resources and the possibility of loss so that you know how to use it when the ball drops yeah yeah i think that's so that the there was a memo the secretary published a memo um last july it's called the drone memo um he published it with a very good he did it with a video release which had drones flying around his head out on the pentagon lawn

which is quite something. A lot of the prescriptions in that memo are of this nature. They were basically deleting policy or imposing policy on people, up and down the commands, saying an 06, or an 05, I don't want to be specific because I don't remember it, but an 05 has the right to basically force the range to say yes to a drone or some such thing. It was intended to get after this issue of,

we really do want these things expended. It's not that we're trying to lose them, but it's in the nature of the training of the military that you have to use them appropriately. Here's the other layer, as you know, and I'm genuinely fascinated to hear how you field this, to the question of warehousing. So I get it, you can't warehouse these things.

You have to expend them in part because the technology is evolving so quickly that treating them exactly like 155 shell. and just having them in stocks and ready to go, just it won't be nearly as combat effective as if you're working through it and exercising and the stuff is evolving constantly. But of course, in Ukraine, you can do that

because you're getting live feedback from the battlefield at all times about what's working, what the enemy's doing, what you need to iterate. My impression is that there are lots of vendors with their own independent communications with different units. I mean, the system in some ways is a mess

in other ways is very flexible. In the United States, we have, even with the, you know, to express an opinion here, the very impressive stuff that you're doing, we're still very centrally controlled. You're a project of the Department of War and of the Pentagon. When it lands and devolves,

it will devolve to the services. Like these are still very centralized, high-level organizations that are going to buy in great quantity. How, as the years go on in terms of peacetime purchase of these systems, is there any way to measure? How do you create the feedback

loop of what the next iteration should be as you're expending the stuff so that you are as ready on day one of a conflict as you can be. Yeah. So ultimately, the drone dominance program is not trying to buy like an end product drone because that end product drone would be outdated very quickly in any iterative cycle caused by war or just technology innovation generally. We're trying to find the, we're trying to partner or find the business systems, the companies that have the capabilities to actually iterate on the drones at the pace of relevancy. The way to do that is with tight feedback loops. It's with iteration that implies the mesh between the companies on the one hand and the warfighters on the other hand. As you well point out, that has been

taken to 11 in Ukraine. So you actually see it's incredibly organic there. And it doesn't quite map to how we think about how we do things here, where you see vendors in the field, with the military units, you see vendors very tightly linked and then feeding back to their product managers daily with iterations, fix this clip. the, you know, they filed Jira tickets, I think, like after a mission. That's actually really, that's, it's exactly what we would like to be able to do to recreate that. But our system, our procurement systems and are just our general way we think about material,

I would say, doesn't easily lend itself to that. So we're spending, so we're spending a bunch of time thinking about as this transitions to the army and the Marines and some to the Air Force and the Navy, how do you link up units? How do you link up the exercises are going at 29 palms or the training centers or the army? How do you link those back to the vendors so that you, maybe you're not quite recreating the pressures and the iterations, innovation that comes from the Ukrainian battlefield, but you're getting some connective tissue sort of going on that. I actually don't know, we know that that's a goal. How you do it, I think, in our procurement system is hard.

No, these are real questions for me as well. I'm thinking through this stuff in real time because you have the added, complexity, that what may work in, you know, Central or South America for a particular adversary or set of missions is useless in the first island chain for very bespoke reasons that are not the obvious ones of terrain may just be the nature of the enemy situation. So unlike 1-55 shells, where there's a pretty strong case for standardization, costs a lot of money, takes a lot of time to build the tube that fires them. With UAS systems in particular, moving, the procurement system to where the services ultimately, presumably, are less, it's less that

they're buying large numbers of product X and then turning around and distributing it. It's more that they're almost, they're wholesalers, they're Amazon, they set the standard, they maintain quality control, they make sure corruption isn't seeping into the system, all this kind of stuff. I'll bet if we go peer under the hood in Ukraine, we'd find all kinds of stuff that's just not acceptable to us. So they're making the system sound, but then allowing for a fair amount, much more choice, than we traditionally have been comfortable with and regional variation than we've traditionally been comfortable with.

Yeah, that's exactly right, right? So you can imagine, we don't presuppose that we're buying like one ubiquitous drone, right? So what is the 25th ID out of Hawaii or out of Paycom have that the Ukrainians don't have? They have humidity and they have salt in the air, right? And they have a different canopy, different coverage when they think about where their mission,

where their operations would be. So is a, this is where we, you know, we had to be very careful not to over-index on what we're seeing in Ukraine on this front, which is would a drone that's optimized for the Ukrainian sort of geography or territory work for the 25th ID? Probably not. So one of the ways that we aim to solve this is a little bit modeled after something Ukrainians have done where they've created this sort of open marketplace and this sort of feedback loop where the more successful you are with various types of kills, the more access you have to this marketplace in terms of funding,

and it's up to the unit commanders to try to tease out which particular hardware system or system solve their mission needs the best. So the Army, the U.S. Army, has what's called the Army Marketplace. It looks like Amazon. I believe it was developed by AWS. And it's very new, but it's working. It actually has this. It has, we've worked through all the issues,

as you said, we've set up the contracting or set up the contracting for it. We're giving people feedback about the capabilities of the particular system. And then the goal is to let units, so try to decentralize the demand signal for those particular drones. That's not usual. Right, right, right. This is a new thing. And I think I applaud the Army for doing it because I think it's the proof will be in the eating and the putting on it, whether or not you actually see the volumes flow from it.

But the ability to say to the commander of the 25th ID, we're not, we don't know what, we don't know what particular configuration of what FPV is that you should want. You need to figure it out. We presume you will need some sort of FPV, but in terms of what the right hardware is that solves your problem set, I think that's for them to actually decide. Travis Metz of the Defense Innovation Unit.

The project is the drone dominance program. You know, we have a lot of conversations here on School of War that end frankly in kind of a dark place. This one actually sounded some optimistic notes. So I commend you on your efforts and thank you so much for coming on the show. Thanks. Enjoyed it. This episode of School of War is also presented by Strato Launch, a leading pure play Hypersonics company. Hypersonic weapons systems travel five times faster than the speed of sound, making them difficult to predict, track, and defeat.

America's advantage in hypersonics depends on delivering. testing, targets, data, and systems at pace. Stratolunch is built for that mission. Stratolunch's reusable and autonomous Tallinn A testbed exceeds Mach 5 speeds, enabling customers to evaluate advanced hypersonic technologies under true hypersonic flight conditions while collecting critical flight data. Together with its expendable hypersonic targets and weapons subsystems, Stratow launch helps the U.S. and its allies move

new hypersonic technologies from concept to capability faster and cheaper. Learn more about how strato launch is accelerating the future of hypersonics at stradolonch.com.