The Vertical Space

#66 Dan Magy: From Citadel to Firestorm, a journey through drone and counter-drone technology

• Luka T • Episode 66

Share episode

Welcome back to The Vertical Space! Today we're joined by Dan Magy, founder and CEO of Firestorm, a California-based company developing low-cost and modular military UAS systems. Dan's entrepreneurial history includes the founding of Citadel in 2016,  a counter drone company employing various technologies, including radio-frequency jamming to detect and mitigate unauthorized drone threats. 

As you can imagine, Dan's rich background sets the stage for a really insightful conversation. We kicked things off by framing the counter UAS challenge, highlighting the inadequacy of most technologies developed in the past decade in mitigating modern drone threats, particularly on the battlefield. Dan talks about the founding and early days of Citadel and what it was like to navigate the complexities of selling to the government while responding to the urgent need of countering small drones in various hotspots in the world. 

We then explore the evolving nature of the drone threat and how this is leading to a sophisticated cat-and-mouse dynamic. We talk about the new realities that warfighters and drone operators in regions like Ukraine are adjusting to and how this is shaping the counter UAS landscape. 

Throughout the conversation, Dan draws parallels to Firestorm's strategy of overwhelming adversaries with mass and modularity. We also touch upon the significance of product and user interface simplicity, how to ensure long-term defensibility of drone companies, the reasons why U.S. drone companies are struggling in Ukraine, and some emerging technologies on the horizon that Dan is excited about.

Dan:

I try and keep this really simple. I love sports and I always think that speed kills. What I mean by that is if I put so many things in front of you or move so fast that it causes you to have to make a really quick decision, the human brain and even computers can only make so many right decisions before they make a wrong decision. So if you can create that chaos and use that to your advantage or the threat of that chaos, I think that's a really good deterrent and that's what led us to found Firestorm with the idea of democratizing the sky. Build so many things at such a low cost that are modular so that you can use this platform for a variety of missions that you could literally paint the sky black.

Luka:

Welcome back to The Vertical Space. Today we're joined by Dan Magy the founder and CEO of Firestorm, a California based company, developing low cost and modular military UAS systems. Dan's entrepreneurial history includes the founding of Citadel in 2016, a counter drone company employing various technologies, including radio-frequency jamming to detect and mitigate unauthorized drone threats. As you can imagine this rich background sets the stage for a really insightful conversation. We kicked things off by framing the counter UAS challenge, highlighting the inadequacy of most technologies developed in the past decade in mitigating modern drone threats, particularly on the battlefield. Dan talks about the founding and early days of Citadel and what it was like to navigate the complexities of selling to the government while responding to the urgent need of countering small drones in various hotspots in the world. We then explore the evolving nature of the drone threat and how this is leading to a sophisticated cat and mouse dynamic. We talk about the new realities that warfighters and drone operators in regions like Ukraine are adjusting to and how this is shaping the counter UAS landscape. Throughout the conversation, Dan draws parallels to Firestorm's strategy of overwhelming adversaries with mass and modularity. We also touch upon the significance of product and user interface simplicity, how to ensure long-term defensibility of drone companies, the reasons why U.S. drone companies are struggling in Ukraine, and some emerging technologies on the horizon that Dan is excited about. Enjoy the conversation with Dan after a brief sponsor message. Dan, welcome to The Vertical Space. Great to have you on.

Dan:

Thank you for having me. I'm really excited to be here and talk all things drone and counter Drone over the next hour.

Luka:

We usually start by asking if there's anything that very few in the industry agree with you on.

Dan:

Very few. I think, I have a very strong take on, where, counter drone is going, and I don't think that a lot of the solutions that have been funded or fielded today are going to work, I think, in, like, literally months or a year. So that's probably where the biggest disagreement in this field comes from. I think if you're asking me domestically versus globally, I think domestically people still think that, now if we're going to talk about drones, that what we've built for the last 20 years is going to work just fine. And when you actually talk to people overseas, they need a totally different type of drone because we're now fighting in environments that have massive jamming capabilities. So that's a two part answer to your question.

Luka:

Cool, and we'll explore both of those, but just as a bit of a teaser, what counter UAS technologies specifically do you think are, not overhyped, but given probably too much credit and likewise, what are the methods that you think really are promising?

Dan:

I'm going to answer the second part first and say, I don't know a lot of stuff that is promising right now, because I think the world has changed so fast in the last year that we've not caught up to the reality that, most FPV drones, right? We'll call them$1,000,$2000, whatever you want to say,a lot of them are now have basically a terminal guidance chip that's put on them for 50 bucks that once you figure out what you want to hit, a yellow box comes around and you hit go and you let go of the drone, right? And so even if you're jamming it, it's going to go run into what it sees. And that's just a fundamental change that has happened. And, you know, these things move fast, so if you're talking audio or, optical, maybe you have at max six to seven hundred meters at max, and that's probably in a quiet environment. or a non busy environment. there's not a lot of those out there in the world, right? It's one thing if you're on a FOB in the middle of northern Iraq, it's another thing if you're fighting in kind of a ruined village in Ukraine. So, I think the future is going to be kinetic or it's going to be EMP based. Those open massive sets of, other problems in the counter UAS space, and we can dive more into this, but like, shooting bullets at a drone or shooting projectiles at a drone when you miss, you have to now think about collateral damage in a way you don't have to with like, RF or spoofing or hacking. I think those three areas are going to have a lot of trouble. RF, spoofing and hacking. One from an encryption standpoint, two because the Ukrainians and Russians are literally hopping between channels and signals depending on what the electromagnetic environment looks like. And then three, because one, people are actively building some really interesting non spoofing technology and I think that you're going to see a ton of SATCOMs come into this or drone over LTE, which opens an either larger can of worms, as far as like, what are we okay, either one violating from a privacy standpoint or two disrupting from just a fundamental usability of all the electronic systems we use in our life standpoint.

Luka:

And in a peacetime environment, certainly huge Fourth Amendment consequences Okay well how about we dive right in this Counter UAS theme For those who don't know, you founded Citadel back in the day. That was the first wave of Counter UAS technologies, and maybe to frame the conversation or actually, you know, what, how about you frame that state of the market at the time?

Dan:

Yeah.

Luka:

what the technology was like and what enabled this, initial crop of counter UAS companies to emerge and just, tell us a story of what that journey looked like for you, of founding Citadel.

Dan:

So the founding of Citadel came out of like a need that was shared with me from my company before that from a variety of heads of securities at sports stadiums saying we have a problem with drones. This is 2015, and that we don't have anything that can stop drones from flying in. Mind you guys, this is nine years ago, right? So this is an old problem, right? And I literally Google searched what was out there and then wrote a patent on an airplane flying back from Boston after the Patriots head of security told me that, like, they're not just at the practice field now, so they think it's other teams spying, but they're, like, coming into games, and I was like, okay, now I've heard this from enough people where this is something we should go after. I am not an RF technologist. I really didn't know what I was getting myself into. I put together a really good team and, we basically used a very cool algorithm we've built, to detect and then mitigate drones and that's what we got. When we started, it was simply just like, find the drone, find the controller and attack it. And what we didn't know at the time was that for how much of a security threat drones bring to public infrastructure, the laws were not going to change anytime soon to interfere with the drone or to jam a signal. And that, there's kind of two groups and one of them has to ask permission from the other group to use that type of technology domestically. So we by accident became a defense contractor when Naval Special Warfare found us end of 2015, early 2016, when ISIS was using DJI drones and modifying them much like the Ukrainians and the Russians are today, to legitimately own the airspace from 500 feet and below in the conflict in Northern Iraq and Syria. Like, they owned it. and they had drone factories where they were modifying them to have servo motors to drop bombs. They were using them to coordinate the vehicle based IED attacks, artillery, RPGs, like everything. They were their ISR and little bomber drones, right? They're a true multi mission,$500 system. What Naval Special Warfare told us is they tried to repurpose a bunch of the jammers that were on the trucks for the roadside based IED stuff, and it wasn't working. So they were desperate to try and find new stuff. And, I think they were, we were one of the first handful of companies that were out there. Another prominent one here in San Diego is SkySafe. They're still around. But we got that first contract. We got, VC, and they invested in us, and, we then had the time and the money to really build out the technology into what it is today. I'll say, we didn't know how to sell to the federal government, right? And so, Luka, I think I've said this to you multiple times. It was like getting punched in the face every day until you learn how to maneuver in this area. And eventually you hire the right salespeople who actually understand how to sell in this area, to one, you know, build the hype train around the product. And we had and have. Probably the best RF counter UAS system in the world. it's a SOCOM program of record. But we also didn't understand how the game works. We didn't understand you got to be going to DC. We didn't understand you got to go win the Labs, right? Like, there's just a formula for how to do this. We didn't understand that. And so I think you saw from 2015 to 2017, maybe four or 500 counter UAS companies started in that first wave. We had a customer where we provided excellent customer support. They liked what we were doing, they liked our vision, and we made some really crucial technical hires that dramatically brought down our false positive rate and then increased the ability for us to interdict a drone without bringing down other Wi Fi systems. Which allowed us to gobble up parts of that market.

Jim:

What was novel about your technology? Why did yours work where others didn't work in the past?

Dan:

Yeah, I'd say, you know, and you could ask any of the founders or a lot of the team that's still there today. The company was acquired by, Blue Halo at the end of 21 right before we became Program Of Record, it's small, okay? A lot of counter US systems are really big. It has what I would say is the best user interface in the entire industry. It is so simple to use and set up that an 18 year old kid could figure it out. I mean, and maybe they're better with technology, but you could say maybe a 50 year old person who's never really touched the system. How's that? Maybe that's a better analogy, right? And then the two other things are we invested very heavily in our AI team and we had a really good data pipeline and training approach that allowed us to remove false positives, right? Build a more robust model to match a drone signal in a very noisy RF environment. And then we built some proprietary kind of amplification technology and integrated in there. One that let us rapidly move between channels to step on, the signals, and two kind of increased power from what you could just buy, if you need to make noise on a drone signal or channel, for without, like, without what's available, on market today. So does that kind of give you a good idea for why we're, it's easy use. It's correctness and identifying things. we talked about this idea at the time called the Hunter algorithm, where we built this algorithm that would be able to see a drone when it comes out of a box before anyone else did that. And to this day, it's pretty amazing. you talk to guys who are still there and they're like, yeah, new drone comes on the market. The machine already knows it's a drone, right? it's pretty cool. So I think that, that really helped. And, we kept our BD team pretty small. And that's something where we'll, when we talk about Firestorm in a minute, like it's really easy to go spend a lot of money on marketing and BD. And we kind of had to go all in on defense because we didn't think commercial was going to happen. And there's a couple of companies out there like Dedrone that have raised unimaginable amounts of money and they're going after those markets. And I don't think as we see wave two right now, unless there's like a massive attack that you're going to see a change.

Luka:

If we think back to those early days of the market that you're describing right now, 2015 ish timeframe, commercially available drones were rather, the fleet variety was not obviously at the level of today. And so it was arguably easier to build RF libraries for these drones to then go and counter. And for the most part, the radio links that they relied on were based on the highly proliferated electronics coming from China, that operated in ISM bands, not really encrypted, no frequency agility. It was a, good old dumb radio quote unquote. And so those early RF detection mitigation efforts worked fairly well. But, and by the way, from what we hear in Ukraine, there's still, those kinds of early,

Dan:

They're there. Yeah.

Luka:

e lectronic warfare solutions that work because most of DIY FPV drones that people put together are relying on those same, non intelligent radios. But, but, but the threat is changing, the attacker is getting more sophisticated in terms of frequency agility, deception, spectral awareness. This, move counter move, dynamic is evolving to the point Electronic Warfare of the past is no longer good, especially if you're facing, as we mentioned before, low, fast, erratic moving, break line of sight type of drones. And so when you project that arc of evolution on the drone side even further, what do you think EW 2. 0 looks like?

Dan:

Good question. I want to add a little bit of context to that story you just told, and I think it's actually really important for like, why we're in this situation today. When we started Citadel, there were actually a lot of different kind of, we'll call them smaller providers of the DIY solutions. And I think DJI made such a good product. And then when they introduced Lightbridge, Lightbridge 2, all of the iterations that you can make when you have 200 PhDs working on this stuff, right? they really consolidated a lot of the providers in the market, both from a product standpoint and then from a like, who the suppliers are going to work at. And so, why I'm saying that is the DJI tech stack, even if you have an intelligent jammer, it moves around when it sees interference, okay? And that's what should have been warning us five years ago that what you just described, right, let's call it intelligent RF, was going to come. And now that there's a war and you have Ukrainians and Russians and Iranians building tens of thousands of drones a month, and they're disposable, the unit costs to go build, those competitive technologies are going to drop down. So what's gonna happen what we've seen and what we've been told, by both Ukrainians and, friends who've spent a lot of time on the ground there, and from our trips to Ukraine is this that. Basically, everyone's monitoring the spectrum at all times, okay? And a lot of the, wideband jammers you're starting to see are being built out there where you have to find the drone in the noise if you're looking for that control link because you can now move between, channels and frequencies with ease, right? And I think we're in the first inning of that. I think by the end of this year, we're going to see some new products that do it even better than they're doing today. But what the Ukrainians and the Russians have done is they've basically adjusted their tactics around a couple key assumptions, right? One is that there's going to be no GPS for 40 to 50 kilometers along the FLOT the front line, right? Maybe 10 kilometers on each side, 40 kilometers in the middle. You are not operating with GPS. That's your first assumption. Okay? So if you start with that, and then you start with, oh, and by the way, everyone has jammers or RF signal, detectors that are able to triangulate sitting in trees so that when you launch a drone, if you don't launch it super far back or put it in the air really quickly and move, we will be able to lock on to where you are in your region and we're going to go put a lot of artillery shells right there. So those are just the key assumptions that everyone's operating under right now.

Luka:

table stakes basically

Dan:

That's it. Correct. And so what do you do and how have they adapted? One, you look at alternative frequencies. So maybe things like Starlink or flying over LTE. And remember, there's also not a lot of LTE here because they know this happens. So they go and blow up the towers, right? But the second thing you do is you build signal repeaters and you either put them in trees, right? Or you put them on other drones. And that is how today, a lot of the stuff, if you like go on my LinkedIn, you're watching coordinated attacks from ISR drones to FPV drones. That's happening a lot of times with, the Ukrainian drone that uses, they like to call it the Baba Yaga, Which is like the witch that, steals children. You put it in the air, it's a lot harder to triangulate it because you can one, move it around, or if it's tethered to a car, you can drive the car around and the drone will follow it, right? You also, when you're up at 1, 000 feet or 800 feet or 500 feet or whatever, you also get a better line of sight. So now, these FPV drones that usually, at the start of the war were going 1 to 5 kilometers, they're flying them 30 kilometers. So I think this whole thing starts to change and we are building a fleet of products out of the same manufacturing cell and now we're building the software to coordinate all of them where people are going to have coordinated drone swarms that do that. You're going to have ISR drones that tell smaller kamikaze drones to go attack stuff and today they're in five to ten person kill teams but in the future that'll all be automated and it's coming in the next 12 to 18 months from Firestorm. We're not the only people who are seeing how this works, and they're basically building, one or two person drone teams that can go achieve this type of success. And I think that's going to make fighting RF really difficult. You're going to need bigger, more expensive solutions. You're going to need layered defenses, right? and I would not be surprised. a lot of this is starting to be open sourced, like the Russians and the Ukrainians are using drones that they basically preload an image of what they want to go run into the drone, and then they go send it in the air and they have it go hunt. And I think that's where one, humans get really scared because robots are making decisions. Two, we as the West, there's tons of material out there saying is this the world we want to live in? Right? And so there's some philosophical discussions we need to have as a society. I think whoever starts getting to that technology first is going to have a massive upper hand in the next coming decades. It's, and it's going to happen because there's a war where it's already happening and they're getting better with every rep.

Luka:

Of all of the counter UAS response options that exist today from guns and drone interceptors and direct energy weapons and electronic warfare, nets, falcons, eagles, streamers, aerosols, you name it. Which have the potential for the most rapid adoption and scaling? If you were to, start another counter UAS company today, what would you focus on?

Dan:

Well, I can't start an RF one cause I still have a non compete for a couple of years. But, no, I would just invest super heavily into something that does drone on drone. I would attack a very certain part of the chain. I think the game is going to be go after the motherships, whether that's the Baba Yagas that are the signal repeaters or the ISR birds that are at, a couple thousand feet doing laps for five or six hours or 10 hours trying to find targets. I would try and basically either cut off the repeaters or the eyes. And I think you can do that in a way because a human is still in the loop today, and they will be for, I think, a while, running that part of the operation. If it's a guy with an FPV drone flying this one to five kilometers, one to 10 kilometers, those people are always going to be there. But I think as you see the coordination of those things, you got to try and either attack that or even go left to launch, And that's when stuff gets super gnarly and that's where our intelligence agencies need to be playing. If they're not already.

Luka:

And left of launch is easy to say, and it makes sense for, larger aircraft that have a more meaningful launch and recovery footprint. But when you're talking about drones that you can launch, from a backpack or just on the go, it's a lot more difficult, to implement in practice.

Dan:

Let me run the scenario in my brain and I'm shooting from the hip here. Okay. So guys poke holes in this, but like we did a phenomenal job in the war in Afghanistan and then even in the ISIS fight in Syria of figuring out where these, roadside bomb factories were coming from by tracing where the money was going. And I remember distinctly when we were working with Naval Special Warfare, when they told us yeah, we found the 2 million dollar order of DJI drones that went through Turkey that were obviously then just walked across the battlefield into Syria and were dispersed into the factories. That's the type of stuff if you look at like supply chain for these things, right. You could have a hundred people swiping a credit card, but if you're going to do this on an industrial scale, you got to figure out where are, and again, this is maybe more terrorist than like state actor, but who's buying, 15, 000 batteries,

Luka:

right? This is far left of launch.

Dan:

I'm talking way far left to the launch. If you want to go closer to left to launch, you're going to need persistent ISR and satellites. That's what you're going to have to try and figure out, right? It's like, where are their launch fields? Are there patterns? But the problem with a lot of these things too, is you can take them out of a car, put them in the air or use a catapult, right? Pack it all up in two minutes and leave. So I'm not smart enough to say, that's how you should go after that. But I would just tell you like the RF triangulation and then some of the interesting things like the Ukrainians, like they've talked about how they have these 50, 000 Samsung phones all listening for the sound of a Shahed across their border. And then once they match that sound, like that's the type of creative thinking you need, which is a lot of low cost sensors doing fusion, and crunching data, in my opinion, if we're going to start figuring out, how do you actually maybe stop some of these things from happening? We're just not there yet today. And don't, I haven't personally given it too much thought yet. Cause it's so daunting of a challenge

Peter:

Dan, I just keep coming back to the question of mass, and how if you look at the countermeasures in the RF spectrum, you look at the idea of countering things left of launch, but if your opponent is throwing a lot of mass into the air, a lot of air vehicles, they're transmitting on a lot of different channels, how does that change the assumptions in this approach? Because when we're talking about drones that are at this type of price point and that can be mass produced this way, I think mass takes on a new role all its own that we haven't seen in recent decades.

Dan:

Completely. And it's because we didn't have to fight people who had these type of counter measures for the last 20 years. Right? So the global war on terror basically leveraged a lot of kind of legacy Cold War systems and repurposed them for a fight against an insurgency that was not a high tech insurgency. And I don't mean that like the Taliban and ISIS and Al Qaeda, they figured out ways to do a lot of really nasty stuff. But it's not like you have, 80 years of state sponsored universities like you have in Russia or, Tehran or China. Where you pump out really good engineers who are now being co opted to solve some of these problems. For me, Peter, I try and keep this really simple. I love sports and I always think that speed kills. What I mean by that is if I put so many things in front of you or move so fast that it causes you to have to make a really quick decision, the human brain and even computers can only make so many right decisions before they make a wrong decision. So if you can create that chaos and use that to your advantage or the threat of that chaos, I think that's a really good deterrent and that's what led us to found Firestorm with the idea of democratizing the sky, Build so many things at such a low cost that are modular so that you can use this platform for a variety of missions that you could literally, paint the sky black. You know what I mean? And the only way we're going to get there is with kind of three key things. One is new approaches to manufacturing. The day of the exquisitely made drone, in my opinion, are numbered. General Atomics and the Reaper is going to have a very, they own that part of the market. They're gonna keep selling birds. We're seeing the Houthis shoot them down now with 50 to 100 thousand dollar missiles. So they're gonna need to stay back and they're going to need to do just like the Ukrainians and Russians are showing us. They need to be the signal relay or the pipe back to the command center. And you're going to need to have a ton, and I mean, like hundreds of smaller, cheaper platforms extending out in front of them, right? That's where the world is going to go. And that's where it needs to go. And because the conflict in Ukraine is showing us that it's changing every couple of weeks. I think, LePlant, yesterday or two days ago, he said, we're seeing EW change every three weeks in Ukraine. And we are not built for that. You need to build systems just like the Ukrainians are building the small drones. We need to do this for the big drones. They need to be modifiable. And for us, that means at Firestorm, it means from propulsion, so a pusher prop that can stay in the air for hours, right? Or it means a jet engine where we can go scream at a hundred to 200 miles an hour and go somewhere really fast and be really torquey to an EDF where we can go like pretty silent, right? That's one part of the modularity, but also from a payload standpoint. I think it's whoever builds a lot of trucks with a really simple integration plan to it and the operating system to make everything work together is going to win because we're already seeing them do that in these disparate little manufacturing cells in Ukraine. We're not seeing that on fixed wing and that's why we've, raised a bunch of money and started to win contracts for that, but it's all about overwhelming. If you overwhelm you control. That is just, I think what's going to happen. I mean, look at the picture behind me, right? How do you shoot all those airplanes down? It's the same exact idea.

Peter:

And the thing we keep reminding ourselves of is that the new status quo that we're talking about here is only 12 months old, and it's not going to remain a constant, just wait two years and see what this, landscape looks like. If people identify that repeaters are a single point of failure, and those are the things to target, then guess what? They're going to be 10x repeaters in the air few months from now. It's constantly adapting. And so, we need to build not for what we see in the battlefield right now, but we need to anticipate another leap in scale coming in, months, not decades.

Dan:

Exactly. And what you just said requires us as a country to not be arrogant or the West in general, right? We fought insurgents who had no resources. We're watching the world bifurcate into two camps, and they both have a lot of resources and a lot of intelligence. So, for us, it is payload bay, electronic hookups, a operating system and a universal flight controller for a fixed wing drone, and no arrogance on our side. We just need to make a lot of those and allow you to modify it. We have our preloads, but we need to basically not have that arrogance. And I tell our team almost on a weekly basis, we are in the first inning, if that's the analogy you want to have, or for Luka, maybe we're in the first five minutes of the soccer match, right?

Luka:

Thank you. I appreciate that, Dan.

Dan:

yeah. I love Modric, he's my, one of my favorite players. so, it's going to be a marathon,

Jim:

I ask you a quick question, Dan? there's a side of me here that says you're using a sledgehammer. In other words, in some ways, are you giving up on the technology solution with an overwhelming dumb solution? And if that's, so one, and then number two, if there is going to be a technology change, that obviates the need for overwhelming volume, what would it be?

Dan:

So some of the more interesting things we've seen out of Ukraine, let me frame it in that debate, cause I think that'll help explain why I think that could, you know, could go either way. Ukrainians are supply chain constrained. They can't put a$2000 NVIDIA chip on every drone they make. So what they've gone and done is they've found a bunch of Texas instrument chips that are$50 that give them a high level of compute that are built in Europe that they can build terminal guidance on. All right. And so you're not needing to do this on an NVIDIA. You can do this on a$50 chip, maybe have a couple of them, okay, so you have some more compute. That's the first step in one of those exquisite technologies, basically being dumbed down and put onto something that is truly disposable. And I think that what we really got to see here is, I think it's going to be a mix of both, if I'm going to be honest. Because even the Ukrainians are telling us why they like our technology is we can make a ton of stuff quickly. So you're going to have to marry the sledgehammer with the sniper rifle, or the shotgun with the sniper and whoever does that in the right way, and that's why we're saying like, Hey, we think that we have some baseline technology that works, but if there's other stuff that's out there, like we will integrate it in days, right? Because we want to be able to provide that overwhelming force with the intelligent capability. Look, you could very well have something that comes out of left field here that totally overwhelms everything we've ever seen and it's like, true neural networked swarms of intelligent drones doing things. For quadcopters, there's some really promising stuff, for fixed wing stuff it's still in development. Right. And we're partnering with a bunch of people to start making these swarms make decisions at a hundred miles an hour, as opposed to 30 or 40, but I don't think we have enough information yet, Jim, to be like, this is what's going to win one way or the other. Is it the sledgehammer or the tech? We're betting that the manufacturing is always going to be a player and so we'd prefer to basically partner with the tech. And so if, and when the killer app is built, we have a way to go field 50, 000 of them.

Peter:

Can you paint a picture for how the human operator's role is going to change as the scale of these swarms increases? For instance, in Ukraine today what does a mission look like for a human operator? Does that person manage the drone from launch all the way through to the completion of its mission? That one air vehicle, or is there a lot of handoff going on, and as we head towards bigger and bigger swarms and more complex attacks, how does a team of operators manage a much larger number of drones? What is that going to look like in the future?

Dan:

There's a couple of different pieces to this question. Today, a lot of the teams, you have one person reporting to the people flying the ISR bird with like, hey, this is what I see, and we think it's over in this area. And then you have somebody with a map standing right next to the pilot of the FPV, saying like, oh look, yeah, that tree looks like it should be right there, keep moving to the right, right? Remember,'cause there's no GPS here. I think that with more exquisite kind of systems in the air, like with laser target designators or a way to hand off, terrain mapping features of what this thing saw to another drone. If you could synthesize that, then the drone could figure out where it should be going, will be coming in the next, in a while. Today, the teams really consist of people behind the front lines, controlling and flying the drones, the repeater drones in the air, the ISR drones doing the circling, and then teams that have to sneak up towards the front line to launch the FPV drones, then get out of there before they get shelled, right? So we're talking like eight to 10 person teams. There are expensive versions of that now by Western primes. Too expensive for where the world is going. I think as people develop ultra low cost, I'm talking like, let's call it, a couple million dollars for dozens and dozens of big and small drones that can work together. That's where you need to be going if you want to be competitive with what the Russians and the Chinese and the Iranians will be doing in the future. So the team of the future, we want it to be one or two people, and we want to have kind of systems that can automatically launch and recover, placed in a variety of areas. And a lot of that kind of, without giving too much away, a lot of that feedback comes directly from talking to tons of Ukrainian drone operators and tons of kill teams as they explained to us the tactics that work and how they've had to build them. Fighting with those two baselines we talked about earlier, which is no GPS, and assuming that you only have like 90 seconds, maybe, before they triangulate where you are, and they just whack you with an artillery shell. So it's how do you get the stuff from behind the jamming area, or behind the targeting area, into the area you need to find the targets, either people or vehicles, and then how do you launch the supporting vehicles in a way that doesn't harm, people who have to sneak up to launch it. And we're, we've already won a contract with the Air Force to do a bunch of that, where we're carrying other drones into combat. So now we have ISR and then smaller drones. And so that's one way in which you're starting to see that happen in Ukraine as well. And I'm glad the U S government's really leaning forward with that project here domestically too, because we think that's really important. And, you know, that's one way in which we get to that overwhelming intelligence plus sledgehammer, right? You put a lot of stuff at the edge and maybe some of the smaller things have some really good capability and you have so many of them that who cares if you lose a couple thousand, we've at least identified all the targets we need to go run into and or follow, right?

Luka:

How mature are swarming, true swarming operations in Ukraine? And I don't necessarily mean you send more than one drone per operator, but true swarming behavior that's being implemented and how do you overall see, the trajectory of that evolution?

Dan:

I don't see them optimizing to it, to be honest, right now. I think they have a system that works and I don't think they feel comfortable with a lot of the technologies that have been developed yet to deploy that out there. I'm just giving you guys like no bullshit when they say to us, do not give us your exquisite stuff, because if it doesn't work on the first or second try, we will literally put it in a garage and start tearing the pieces out of it we want, and we will not fly it. We will cannibalize it.

Luka:

That's what we hear. We hear from drone operators that, people throw all kinds of, feature rich, complex, very, shiny products at them. And the feedback that we're getting is, look, I don't think that those people understand the kind of user that is going to put this in their hands. This is a sleep deprived, afraid, stressed, a hundred other things on their mind type of user who just needs very simple buttons.

Dan:

8 to 12 hour shifts, right? Every single day, for months, that's who you're dealing with. it, I, Luka, you nailed it. So, like, a couple things need to happen here at the same time across the Western version of our drone industry because I'm going to just collectively put, heh, the bifurcating world is happening, right? We need to make these things super easy to use. And so for us on our team, it's taking the lessons of Citadel for how easy it was to run a counter UAS system, right? Making it look magical on a UI/UX screen, and doing all of the complicated stuff in the background. And we're about to debut our new mission planner where it ingests real world data into a 3d environment with real physics engines and allows you to modify. What do you want to do? Do you want to do an ISR capability? Great. Set up your mission, right? You want to do a loitering strike. Great. Set up your mission. But again, this is V1 of this thing. It's probably already way too complicated than it needs to be. But for us, we see a world where you need to use existing technologies that people have been trained on and make the magic happen in the background so that they can run it on a cell phone or a tablet, something that they're already really familiar with. Right. But, Coming back to your original point, we're hearing the same thing. And I think the problem we've run into, and we have to be very delicate here, the Ukrainians need materials. They are in a life or death struggle. Where they have a problem is that they go, not all the stuff you send us works because it's over complicated and it's not built for a jamming environment, full stop. But it's hard to tell somebody who's giving you a lot of money that you don't like what they have. Now they've gotten better in the last year of saying, this is the type of stuff we want. And if you guys look at the Wall Street Journal, I'm not going to name anybody, we've been hearing this for a year of companies that have sent tens, hundreds, maybe a billion dollars of product over there that when we go over there, we see them in warehouses, full stop, right? And that's material. And all they're asking for and what they like about Firestorm, and I don't mean to just keep plugging it, but we are listening to the customer, they're like, we need a system that has propulsion, that's modular and that we can easily get into and modify. That's what they need. Cause that's what they're having to use constantly on the battlefield. Right. You modify stuff.

Jim:

What's the greatest limiting factor of this let's say high technology that's being sent to them? What are the biggest reasons it's not working for them? So one, you said it's a jamming environment. It doesn't seem like these people are anticipating the jamming environment. Number two, you mentioned these people are sleep deprived. They don't have the mental awareness that's needed to be able to operate these. What else is the fault of the billion dollars of technology that's being sent over, that's not working for them.

Dan:

How it's built. It's not built to be modified. Like literally changing a nose cone is not modular. Need to be able to get in there. You need to be able to put bigger payloads in. You need to put smaller payloads in. That stuff's really important. Change comm systems. These are hard problems. We understand this is a marathon, right? And we understand that we can't do everything. We're very, egoless over here about where we are as a business. We're a young drone manufacturer, but we listened to the end customer that says like, look, have your build outs, but have it be able to be modified. And there has been some disconnect between what the U. S. is funding and what the Ukrainians are asking for. And it's a very delicate situation because you don't want to like, do you know what I mean? Like it's hard when someone gives you a lot of money to tell them like, Hey, we don't want that. Right? But you are having people in the U. S. government, like LaPlante, come out and openly say, we gave them all these GPS guided artillery shells, none of them work. You know what I mean? And then there's a couple Wall Street Journals UAVs that are not working.

Luka:

Dan, what are you hearing from end users in the battlefield, as it relates to existing, call it workflow, call it kill chain, but what parts of that workflow are still manual, difficult, time consuming? On the one hand, what we hear is that for instance, the target detection or the ISR bird that's looking out for targets of opportunity that on the one hand, it feels like that is something that could be somewhat easily automated, but when you actually talk to the people who are on the battlefield, you realize that this is still a very manual, far from being automated, part of that workflow. What other things are you hearing that people are having most problems with?

Dan:

Yeah, look, What I think you're going to see in the next two years, and I'm just going to say like where I think it's going and it will show where I think a lot of these pieces will be automated. You're seeing like ground robots in cars, right? A lot of money. A lot of talent has gone and worked in that space over the last, what, decade, probably, as we've started to use computer vision to solve real problems. And guys, I have a four year old Tesla. There's two parts of my drive where it still can't determine something and it freaks out a little bit. So we're not there on that stuff, but there's hundreds of millions or billions of hours now of video where these models can go and learn, right? We don't have that yet for UAVs. The Ukrainians are building that. They have tons and tons of footage, but I think that's going to help with automatic target recognition and then terminal guidance, honestly. And you're starting to see some of that stuff pop up on GitHub. Which means I think a lot of that stuff will be commoditized over the next couple of years as you build the right kind of data training pipelines on that stuff. And I think that there's opportunities for Western governments to work together to collaboratively build some of these models and then to figure out how to run them at the edge with very low power. Do you know what I mean? so that's one area. So let's call that the kind of target, recognition. I think the second area, which has been just fascinating to me is you talk to some of the guys who run deep strike okay. So we're talking like hundreds and hundreds of miles or, a thousand kilometers. Right. And, They will literally have someone flying with a map next to them, right? A map on a phone and they are doing visual identification,

Luka:

dead reckoning,

Dan:

reckoning. it's. That, that, that'll change. There's a bunch of companies working on that. It's very compute intensive right now, but at some point, someone's going to basically be able to upload real time satellite, there's already real time satellite imagery, right?

Luka:

that, that seems like a problem that's already solved or being solved.

Dan:

being solved, right? It's, I'll be honest, like when you ask the guys, what are you guys using for non GPS denied navigation? And they're like, well, not a lot. And you ask them why? And they're like, well, cause it's, again, maybe we're in the third inning of the ball game, not in the first inning, but it's coming along, but then there's some also really interesting non visual, INS stuff that's out there. We've seen, we're integrating one of them right now that we're pretty excited about, that we are being told by end users is very successful out there, and it does not use GPS, which is interesting. But I think you're going to see multiple little bites to be taken out of like where you need a human to solve a problem, right? And I think it's going to be a lot of little steps every week, will result in like a year from now, we'll be like, May 1st of 25, we're going to be like, Oh my goodness, can you believe we had 10 person drone kill teams? And we had, a guy looking at a map on his phone or tablet trying to figure out where he was. And we had, a drone not being able to coordinate with another drone, what to go run into. Do you know what I mean? Something like that. I think some of that stuff's coming. They don't actively talk about the need for a swarm, which is very interesting because the swarm dominates our conversation here. We talk about drone swarms all the time, both from a national security perspective and what we need to build. They're just like, well, there's a lot of targets and there's a lot of data we have to crunch. Let's just be smart and get stuff out there and then solve the problem. Because the humans still make better decisions because they've been trained at this point, right? So, I, and I don't think that goes away anytime soon, but in the next couple of years, maybe you have something where you're simplifying the decision making process for, is this a friend or a foe or a threat?

Luka:

You mentioned two assumptions that people are taking as table stakes: no GPS, and then, jammers and direction finders everywhere. What other assumptions both the users and people who are providing technologies to those users should anticipate in the future of drone warfare, and especially from the perspective of a founder, right, or an investor where you want to, build products that have relevance in the years to follow and in an environment that changes so rapidly, what additional assumptions do you think that are safe to make? Arguably you're talking mass and modularity and production. what else? And what are the trade offs?

Dan:

I think you should be trying to build everything to be future proof, right? And the problem is that when you go, what I mean by that, of course, is leading into modularity and like the ability to rapidly change

Luka:

When you say everything? Sorry. Is that like vertically integrated end to end, or do you have something else in mind?

Dan:

Yeah. Yeah. More, more along the lines of that type of vertical integration, right? We have six or seven propulsion systems we use, and that's because I just have concerns about supply chain for everything. I do think some things like sensors and cameras and radios, I think over the next couple of years, I have less concern about that kind of being decoupled. I think there's enough demand now, from China. Right. But what I would say is this is where the problem is because the U.S. DoD is not built to buy modular stuff. It just is not, it's used to buying a SKU. And so if you're going to look at the war in Ukraine, we need to really think about how do we align the incentives for the procurer and then the entrepreneur to solve that problem. And it's really tough. And so for us it means buying a lot of time and raising a lot of money and selling the stuff to a variety of folks to buy that time, because I do believe it will catch up because there's no other way we're going to fight the conflicts of the future. Right, because of the jamming and the GPS, that's just like the start of the table stakes, but the issues I think founders are going to face is if you're building really cool software or one piece of the puzzle, if you do not have a platform to put it on, it is very hard to find customers. And I think that, again, everything, including drones, is going to be commoditized over the next four to five years. So you have to have a business model where you can build lots of stuff, put it on things to basically get it into the fight, and, and still be able to make money. Right.

Luka:

Right. and that's critical and sorry to interrupt, but, you're raising a really important point because traditionally, people associate big moats, with some single piece of really interesting proprietary technology. But if you're moving towards a world where a lot of the products that work are actually very simple, very affordable things that you need to be able to produce at volume. If you're trending towards this model, which, broadly is commoditized hardware, then innovating on the business model, on some of the other dimensions really becomes the key source of, defensibility for founders, which is a different way of thinking about ensuring this defensibility of the business.

Dan:

Totally. And I think that means then as founders, you have to look at markets outside of the U.S. as well. We're addicted to buying big, expensive things. The Marine Corps just bought 450 million of product that you can go read the Wall Street Journal to see how it's performing in Ukraine. That was a week ago. I mean, we're, again, we're addicted to doing this. In the world I see it, and why we brought Lockheed Martin on as an investor early, the big guys know how the game works. And defense is exclusionary by nature. The world, if we're going to be successful, is small companies working with big companies to effect change within the Department of Defense. And the same thing needs to happen with our allies around the world, These big defense companies exist because they've been a very secure way to build big cold war types of systems. And so the systems have been built to protect them and keep them in business so that if conflict breaks out, you have the ability to scale exquisite systems. The world is just changing, right? Again, you can go watch my LinkedIn. I do this a couple times a week where I show a thousand dollar system blowing up a multi million dollar system. We're talking millions of dollars of ROI on small drones knocking out armored vehicles.

Jim:

You've made comments on Replicator recently. So what should change as it relates to Replicator based on what you're talking about right now?

Dan:

We, look, I commend Doug Beck and the DIU team for getting in front of this. I think the key thing is going to be, and I think what all of industry and Congress is asking for is to take risks so that we keep doing things the way we have done them. And what I mean by that is how do we build more of stuff faster? That's the key. The delivery delays are just too long for a lot of the big more exquisite systems. And you're looking at Ukraine with 10, 000 people making 10, 000 plus drones a month. And I'm not saying that everything needs to be FPV, but that should be helping you dictate where you go with this side with this and the problem we're going to run into, it's like, so say DIU funds, the start, who's going to fund it after that? Right? That's the big problem we're going to see, I think, on some of this stuff. I'm not worried about where does the first tranche of cash come from? I'm worried about where does it go longer term? Because again, to become a program of record you need money. You need palm cycle there's all of these things that kind of prevent that from happening. And, I want to make sure we're taking risks now, cause in the scheme of things, a billion dollars that they have to spend is nothing for the DoD, right? This is where you should be really risky in my opinion. But it can be really hard for people to, not push the easy button and just buy more of all the people they know

Luka:

We talked about some of the technologies that are on the horizon, namely, automating certain aspects of the kill chain and the swarming behaviors. But is there something that you are particularly excited about, that you're seeing on the horizon or something that you would point our entrepreneurial audience to go and innovate in and solve.

Dan:

I think the more we can bring in human and machine teaming to the decision making in the build up, to running a mission, the better. You know what I mean? I'll just leave it broad like that, but a lot of data goes into like, where do we spend our resources and where do we put things in the air? There's a lot of interesting technologies now you can fuse from satellite images, to weather data, to terrain, to news that's happening, like all of that stuff is out there. And if you can synthesize it to help you make better decisions for where you should fly your systems, I think that's going to be, play a huge role over the next 10 years, right? Like go run 50,000, 100,000 it's like the General Mattis, right? I want my soldiers to run, 50 bloodless battles before they go into combat. Well, 50 sounds light. Let's go run 50, 000 scenarios before we go in and do something. I think that's where this is going to get really interesting because that data can then feed what type of platform you put in the air and what type of payload you put on the platform. Some of the more interesting startups I've seen that have thought about this in a variety of ways have gone after kind of the data problems that those things will require. That's a really unique way to tackle something that both the primes, small businesses, and the government would probably all pay a license for

Luka:

Dan, what about use cases on the commercial side of things? You've been in this industry for a while. what are your thoughts on the major demand and supply forces at play here?

Dan:

I'm bullish on components and basically the decoupling of the supply chain. It's a little bit of chicken and egg. Everyone's like, well, do you ban DJIs? Does that then kill the whole drone industry? Does that jumpstart the domestic stuff? We're seeing NDAA compliant pieces and products go to Mexico. There's a robust supply chain being built in Vietnam. Okay. Cause don't forget guys, before the Vietnam war, the Chinese are their enemy, their generational enemy. Right. And they did fight a war after the Vietnam war with them. And they have really good manufacturing technologies there. I think that, as more of the beyond visual line of sight stuff kind of gets ironed out here, you're starting to see people over the last couple of years, starting to get waivers. I think the future is autonomous. I think you're going to see more and more human and machine teaming. And a good thesis has always been like, what can you do, or what can you put on a drone so that you don't have to fly a And we see a lot of helicopters flying over power lines where I live in Southern California, looking at stuff. I think that's a good use case for drones. You're seeing a bunch of these drone first responders software companies come out where they're like, hey, we'll build a whole management solution. But again, I think for a lot of the different stuff, it's going to be, you have to commoditize the prices of it. And that's where this thing gets really tricky because to force a change you got to have something either one, be very expensive to force someone to change their behavior. Or two, you have to offer an overwhelming product. And I think that some of the stuff's really good. Like you're seeing a lot of drones being used in ag because, there's less people working in ag. So that's a natural kind of replacement, if you will, of humans. I still think we're probably five or 10 years away from like mass adoption for these systems.

Luka:

I want to ask you about the supply chain. We have talked in the past about that and you mentioned that you're less concerned today than you were, say, a year or two years ago in terms of the ability to source some of these critical components. Any additional light that you can shine on the state of the supply chain and how disadvantaged is the Western supply chain still to the consumer electronics supply chain coming out of China?

Dan:

Yeah. It's very disadvantaged. If you guys have ever been to Southern China. If you want to build a product, they have their almost internal Amazon, you can say, hey, I need 10 of these motors and they'll show up that afternoon. It's incredible. And you can start testing them. And a lot of that is just Southern China was the factory of the world for the last 20 years. So it's going to take some time, but look, NDAA compliance, a desire to move away from just the reliance on China for everything is happening not just here, but around the world and I will just say, like, at Exponential last week, I was really impressed by how many products are saying we're either moving out or we're already out of China. It's fascinating. So, I think it's just a trend that continues and maybe that makes some of the subsystems more expensive in the short term, but in the long term, I think, look, as things move out of China, everything's going to get more expensive. That's, let's just be honest, but I think you're seeing a lot of the kind of manufacturing supply chain be rewritten. Living in San Diego, you guys would not believe how much construction is happening right at the border or right over the border. It's millions of square feet of manufacturing is going on there as companies near shore. So, I think that's a trend that just continues.

Luka:

Any parting thoughts for the audience or advice for entrepreneurs or, otherwise.

Dan:

I think it's an interesting time in the world. And so my thoughts for the audience are, we just cannot be arrogant that we know how this is going to go because we're not actively fighting over there and we don't have the right feedback loops coming back to the decision makers in the Pentagon for how this stuff works or doesn't work. That's just, a general thought. And then for the entrepreneurs, in my opinion, it's always been like find a hole and then jump in with two feet to try and fill it. So, I'll just keep it high level at that.

Luka:

Dan, this was a really interesting conversation. Thank you very much for, carving out a bit of time to talk to us today.

Dan:

I really enjoyed it guys. It's always good to talk about, this type of stuff. And I hope you guys learned something and thanks guys.