In this episode, we’re excited to host Srdjan Kovacevic, CEO and co-founder of Orqa. Despite a low public profile, this Croatian team has made significant waves, initially through ultra-low latency FPV goggles for drone racing and now by building their own drones through a fully vertically integrated model. This strategy not only optimizes cost and performance but also positions Orqa as a strong Western alternative to DJI.
Srdjan discusses his view that hardware in the small drone industry is far from commoditized, emphasizing its critical role in achieving optimal performance. He traces this misconception to the early consumer drone market and examines DJI’s innovation-driven dominance. Our conversation explores the benefits and challenges of vertical integration versus system integration, with insights into its impact on ultra-low latency video links. We also delve into how market dynamics and geopolitical events like the war in Ukraine are shaping the drone industry, and Srdjan shares his thoughts on FPV systems, their modern warfare applications, and future drone technology developments.
06:02 Hardware will not be commoditized
10:18 State of the drone industry and how we ended up in a polarized market
13:10 DJI's vertical integration playbook
18:34 Why vertical integration matters for small UAS
21:02 Low-latency video as an examaple where vertical integration matters
29:32 Why are there so few vertically-integrated drone companies?
35:20 Demand drivers for ultra low latency video
37:49 Impact of ultra-low video latency on managing autonomous systems
39:42 Opportunity to displace DJI in the West
45:09 War in Ukraine as a perfect storm for the small UAS industry
48:23 Competitive dynamics across segments of the drone market
50:53 Small UAS cost drivers
52:55 The future of FPV drones
55:05 What are FPV drones?
58:35 Are FPV drones here to stay?
01:06:26 Technology gaps in FPV drones
Srdjan: 0:00
This vertical integration playbook really benefited DJI at first on the logistics level and on a pricing level where they could simply afford to go down with price because they didn't have a"guy" that was doing all of their manufacturing that had to make their money. They were "the guy". But later, this system integration really enabled them to produce systems that are simply much more performant. If you are not just doing your own manufacturing, but also rather than system integrating components that are produced by other parties, you are actually building all those technology blocks yourself from the chip out and then from within the chip itself, you can start to unlock all these little pockets of efficiency that sit between the black boxes that you would otherwise connect being a system integrator. And it boils down to, in order for these subsystems to be compatible they need to talk over certain protocols, which are designed for interoperability, not necessarily for efficiency, and they're always a compromise. When you have the luxury of owning the black boxes, the two black boxes you need to integrate, then you can open them and you can tweak them and you can optimize for whatever you're optimizing for. And then unlock a tiny bit of efficiency here, tiny bit of efficiency there. And when you look across the whole system, these little efficiencies compound.
Luka: 2:40
For those in the small UAS space, Orqa is a fairly well known name. Despite maintaining a low public profile, this innovative team from Croatia has been making significant waves in the industry. Starting with ultra low latency, First Person View goggles for the niche drone racing market, Orqa has expanded to build essentially every drone component, including their own drones, through a fully vertically integrated business model. This approach optimizes both cost and performance and is positioning Orqa is a strong Western alternative to DJI. Welcome to another episode of The Vertical Space, where we have the pleasure of speaking with Orqa's CEO and co founder, Srdjan Kovacevic, about the past, present, and future of small UAS in both civilian and defense applications. Srdjan begins by expressing his disagreement with the belief that hardware in the small drone industry will become commoditized. He underscores the crucial role of hardware in achieving peak performance for small aerial robots. And he traces this misconception back to the early days of the consumer drone industry and highlights how DJI's relentless innovation enabled it to dominate the market. The conversation addresses the challenges and opportunities of vertical integration as opposed to system integration, drawing parallels with the mobile phone industry. We delve into specific examples demonstrating the significance of vertical integration in small drones, particularly in the case of ultra low latency video data links. The discussion also explores how market dynamics and geopolitical events like the war in Ukraine are shaping the drone industry. And Srdjan also talks about the intricacies of FPV systems, their application in modern warfare and the potential future developments in drone technology. As mentioned, Srdjan is the CEO and one of the co-founders of Orqa, the world's leading FPV company. Orqa is dedicated to advancing the state of the art in unmanned technologies and remote vision. Before founding Orqa, Srdjan had a distinguished career in the financial industry with over a decade of experience in asset management. He served as the chairman of the management board of Croatia's 150 million Euro sovereign equity fund and was the Head of Risk Management and Compliance at Croatia's largest asset management company. Srdjan earned academic degrees from University of Oxford with a Master's of Science in Mathematical Finance and University of Zagreb with a Bachelor's Degree in Mechanical Engineering He is a certified pension fund manager and certified investment advisor. Enjoy the conversation with Srdjan after a quick sponsor message. Srdjan, welcome to The Vertical Space.
Srdjan: 5:57
Thanks for having me.
Luka: 5:58
What are some of the things that very few in the industry agree with you on?
Srdjan: 6:02
My pet peeve is this old saying that hardware is going to be commoditized. Especially for unmanned, I strongly disagree with that. I don't think hardware is going to be commoditized. I don't think you can ignore hardware. Or, ignore hardware at your peril if you're doing small things that fly.
Luka: 6:22
Expand more, please.
Srdjan: 6:23
Okay.
Luka: 6:24
Give us some more details. And then how long do you think this will persist?
Srdjan: 6:28
I think this is going to persist until, critical mass of people are proven otherwise. And, if you ask me that fallacy goes back to, 2014 15 so speaking specifically in a small UAS space, in the drone space, right. And I believe you could trace it back to, point where DJI, really, defeated, 3D robotics, and 3D Robotics decided to pivot in software And, despite the, overwhelming proof otherwise, we keep hearing that mantra that, oh, hardware is going to be commodity, so we should focus on software solutions in the drone space, where, it's glaringly obvious that, the way DJI won was by relentlessly focusing on improving hardware from a generation to a generation of the product and, again, I'll say it, neglect hardware at your own peril when it comes to small things that fly. Software a part of the solution, we shouldn't be neglecting software, but I don't think you can achieve the level of performance if you think you can ignore hardware, or run your software somebody else's presumably commoditized hardware.
Jim: 7:49
Why do we hear hardware can be a commodity? And is there any particular part of the hardware that you think is potentially becoming commoditized?
Srdjan: 7:58
so I, I think that goes back to the experience that we had with, a PC world or computer world, like hardware is arguably commoditized there, but, then again, it only takes a look at, what Apple is doing to realize that, yeah, hardware can be commoditized, but if you want peak performance, then you need to vertically integrate and you inevitably end up doing your own silicon, which I believe is going to be the case with the small unmanned and by the way, which is what DJI is doing at the moment. In saying that, people are, willfully or unwillingly ignoring, cherry picking the examples that fit to preconceptions. And I think it's just like, one of those things that are so easy to say, so we don't think about when we say them. But I hate hearing that.
Luka: 8:45
Do you think that, this argument is true for component level hardware, or are you referring to the art of integrating a hardware product together with all of the, details associated with it, which we will get into more detail later.
Srdjan: 9:02
It's an interesting observation that you ask it like that, and I think, by saying that hardware is going to be commoditized, we gave up on, the possibility that we can compete with DJI, like a collective, rest of the industry and we're settling for a world of commoditized hardware where, we'll just buy components, that should be interoperable and system integrate them in our own solutions. But it's a good question. And I do think, in a world of other drone manufacturers or other drone systems, arguably the base ecosystems, meaning the Pixhawk universe, or Ardupilot universe, or Betaflight universe. All of these subsets have been driving the development of the software stack, autopilot stack, or flight controller firmware stack towards hardware agnostic solutions to a point where, okay, everybody's running on STM, but like on, on a module level. So I think it's also in specifically in the drone world this approach is actually, giving up on the possibility of competing with DJI which was a fair conclusion up until couple of years ago with decoupling and everything.
Luka: 10:18
From your perspective, how did we end up in the drone industry where we are today, where DJI dominates in a polarized market, dominating both in terms of growth and product quality, while drone companies in the West are having a hard time competing both in performance and price.
Srdjan: 10:36
I think first of all, we need to agree that DJI is just an amazing engineering company, being started by an amazing founder who relentlessly pursued certain goal and would not accept compromises and infuse the company culture with that sort of attitude. And then, going back, I think again, I'm going to go back to what for me was the pivotal moment in the development of the drone market is 2014, 15 3D robotics coming out with the Solo and then DJI countering with the Phantom. I think that was the turning point where all of us users were finally, able to get a really cheap and highly performant vehicles, drones. And DJI has proven that they're going to be super tough to compete with. And, at the time 3D robotics was one of the most funded Silicon Valley companies. And it was being run just like, pretty much every other Silicon Valley company was run, you do your design and your maybe prototyping on shore. And then, you build wherever it's cheapest. And I think it was the pace of innovation that, that was happening in that space, really just shown how having your manufacturing super close to where your R& D is, that just sets you up, to innovate much more rapidly, to crunch those cycles that you can just out compete anybody else who's trying to do that onshore, offshore kind of approach. DJI started, in Hong Kong, having a vast ecosystem of service providers and manufacturing companies in that local ecosystem certainly helped, but I think this was the, deciding factor that then at the later stages, just, help them compound this advantage and really just, build massive distance between them and everybody else. And, they were just better at building amazing technology to a point where the difference in performance was such that nobody could compete with them. And if like we as work as a, as an FPV company, we felt that, when DJI entered the FPV space, which was incidentally just months after we launched our first product. And I like to say that this was a super scary moment. Imagine you're building an app and your app is, running on an Apple phone. And then Apple decides to, look into that space and all of a sudden that app is part of the standard Apple package. You're pretty much dead, right? So, it felt like that, big 800 pound gorilla, the DJI was entering in a tiny market, which was FPV at the time. And, they were basically able to leverage their influence and access to certain, certain parts of the hardware ecosystem in China in order to provide solutions that were simply out of reach for us. And specifically, this was like, custom built chips that did low latency video, over, over RF, and that was for any other company or almost any other company in the FPV space, that was just completely out of reach from the standpoint of existing know how, access to capital, et cetera, et cetera. So, this is basically an advanced version of what was happening back in 2014. It's just access to different aspects of, technology.
Luka: 14:44
But DJI was also very aggressive in their pricing. If I recall back in 2015, with the Solo, the one that had the GoPro camera and the gimbal, that was priced around 1, 700, while the Phantom 3 Professional package was sold at the same time about 1, 300, but then, a year later, DJI brought the price down to 900 and then slashed previous year's models to as low as 500. And so that pretty much put every other company in the West out of business. And, perhaps because they saw software as an advantage or because it was out of necessity, but Western companies pivoted to software. It was very difficult to compete with that. So, yeah, I agree. It was. It was a very, bleeding edge, innovative product and culture, but it could afford to have such price cuts. And so from that perspective, how did the strategy of their vertical integration play into their ability to outprice the competition. Perhaps this is a good transition to talk about their business model of vertical integration, as opposed to what we see currently in the West and compare and contrast.
Srdjan: 15:58
It's an interesting observation and, for me, when I was, looking at that, my first thought was, okay, what the hell they were thinking 3d Robotics pricing their, product at a price point that was higher than DJI, you know, fiercest competitor to begin with. And then, DJI it even went lower than the original pricing. So that, that, that was the first thing that they struck me. But I think that, this vertical integration playbook really benefited DJI at first on the logistics level and on a pricing level where they could simply afford to go down with price because they didn't have a guy that was doing all of their manufacturing that had to make their money. They were the guy. So they could just make slightly less than that. This was a luxury that Western companies could not afford. But later, this system integration really enabled them to produce systems that are simply much more performant. So it was not just the pricing power and the logistics of it, speed to innovate or the ability to slash prices and still make money. It's also if you are not just doing your own manufacturing, but also rather than system integrating components that are produced by other parties, you are actually building all those technology blocks yourself from the chip out and then, otherwise, and then from within the chip itself in, in, in this latest stage, right, that I referred to earlier, you can start to unlock all these little pockets of efficiency that sit between the black boxes that you would otherwise connect, being a system integrator. And it boils down to, in order for these subsystems to be compatible, they need to talk over certain protocols, which are designed for interoperability, not necessarily for efficiency, and they're always a compromise. And then, when you have the luxury of owning the black boxes, the two black boxes you need to integrate, then you can open them and you can tweak them and you can optimize for whatever you're optimizing for and then unlock a tiny bit of efficiency here, tiny bit of efficiency there. And when you look at across the whole system, these little efficiencies compound.
Peter: 18:34
Give us some specific examples of where that manifests in drone performance, because as you describe this history, something that jumps at me is how is it that this performance advantage persists nearly 10 years later? If we look in the smartphone space, for instance, 10 years after the introduction of the iPhone, the hardware from competing smartphone manufacturers was all in the same league, right? They were all competing neck and neck. Whereas there's something about robotics or about drones or about this type of system that has enabled the first player and has enabled DJI to, not just have economies of scale, but to have what you're describing as a fundamental gap in performance versus what someone else could build using best of breed components that are available out in the rest of the market. You put those components together and you try and build a performant drone. It's going to fall short, and you're going to explain, hopefully, what that looks like today, and why that is, because I think that gets to the heart of the value of this integration, and it also gets to the heart of your first observation about the conventional wisdom that has been built up over the last few decades of how and why hardware commoditizes. Why is there a wrinkle in this space that challenges that conventional wisdom and leads you to a different conclusion? Let's get down to specifics and really unpack it.
Srdjan: 20:08
I think the parallel with, mobile phone, industry is a great one because both in, in small unmanned and smartphones, size matters, energy consumption matters and performance matters. And specifically thermals also matter, right? One of the, one of the interesting aspects where tighter integration can lead you to better outcomes than system integration, right. So I think it's a great, it's a great comparison and, In that world, in a mobile world, we have Apple versus others, just as we have DJI versus others in the small UAS world. And it just proves the point that, system integration will get you to a point but it will not bring you to the, the position of the number one player in terms of performance, right? And we're seeing the same thing play out in the smartphone world and in the UAS world. And I'm going to give you an example where this system integration, versus vertical integration really takes you to different levels, and that would be, low latency video as a sort of, specific use for a data link in a small UAS. In low latency video, and this is something that, that, we're very intimately, connected to because this is something that we've pursued and tried to compete with DJI in the FPV world, so we know that really well. In low latency video, there are certain quite different technology blocks that you need to integrate very tightly. And and in this case, system integration, like, it goes one level deeper. It's not system integrating a radio module and a video processing module. It's basically done on the chip level. So if you're not doing your own chips, you need to get a chip that's going to do your video processing, perhaps host to radio, then you need to get your chip, which basically is your baseband chip, right? And the problem with that, and it's that level of system integration. So it's like chip, not board level, but chip level system integration. And in this particular case, neither of those chips have been designed for this particular application. if you're really doing ultra low latency stuff, as we did in the FPV world, in the consumer FPV world, our, measure of latency was in milliseconds, single digit milliseconds. So, like, we need it to be below 30 milliseconds, glass to glass. When the photon hits the sensor to when the photon emits from the display on the other end, that needs to be 30 milliseconds or less.
Peter: 22:48
Just for clarity, that's important for the human pilot flying the drone to have that sense of instantaneous transparency through to what's happening with the drone. That's where any sense of delay just disappears. Is that what you're
Srdjan: 23:02
Yeah, I mean, to go to, to frame that further, like, flying, consumer FPV, either racing or freestyle, human is very tightly, embedded in the control loop. The drone is not stabilized in any meaningful way. You are keeping it level by throttling it more or less. So, so you really need, with your eyes, you perceive the response, dynamic response of the airframe and you're providing the control input. It's pitch and roll or it's yaw or it's throttle, which is, arguably the most sensitive part for most of the new pilots, right? Throttle control. So, with low latency video, the lower that latency is, the more tightly you can be embedded as a pilot. You can be embedded into that control, right? This is why latency is important. And, it was funny because when we were thinking about latency, low latency in the rest of the world would be below thousand milliseconds. And there we were counting single milliseconds in order to, acquire frame from the sensor, put it into memory, read it from memory, encode it, compress it, packetize, you know, again, memory write, and then again, make it into a packet, send it off to a radio, and then radio needs to be transmitted, right? So, we're counting milliseconds and, all of these chips are really made for a user experience which operates in a sub second world, half a second here or there, Great example is, encoders, decoders. Like, we're talking about SOCs that have video encoders and decoders. They are made to consume or originate media in an embedded device. Not necessarily to be super fast and super efficient at encoding video and sending it at high frame rates and counting latency in, in milliseconds. So, if you have any latency, you're gonna buffer. And, that you've problem solved. Nobody's gonna, nobody's gonna, notice. So, first of all, like, if, when you start selecting those SOCs and you start asking questions to chip vendors, okay, how fast, yeah, we have a, encoder that's capable of doing, X. Okay, how fast does it compress a frame of certain, resolution? They don't know. They don't care. They don't have that metric, So this tells you that, we're trying to use a chip for an application that was not even like an afterthought to folks. And then on the radio level, on the radio SOC level, like, if you want to be, affordable for consumer space as we want it to be, you need to go commodity and commodity in this world means, Wi Fi SOCs, 802. 11 as a standard is not optimized for that kind of latency. It's optimized and it's designed for, interoperability and coexistence. We need to have a number of 802. 11 APs and station devices that, that, need to coexist in the same space. So they, they need to be able to negotiate how they access the spectrum, and not stamp on each other. That's not what resilient, low latency, radio for unmanned applications is. It's all about, consistently and robustly being able to transmit video packets without interruption to the ground station. And again, you're trying to make that radio chip do something that it has not been designed for and all that firmware that's not been designed for. So like, you're having to make compromises. Or go off standard if you can, which you can't because, the volumes that we're doing are so small that all these chip vendors are not paying attention. So, you end up making all these compromises, which then result in a performance that's suboptimal. Whereas, if you had the full control over the entire black box that is the chip and the hardware encoder, and the black box that is the radio, piece, the 802.11 chip, you would be able to, go into that black box, tweak what you need to tweak, make the two black boxes talk to each other seamlessly and efficiently, and without losing a step. And that's how you get the performance system. And which makes the difference between DJI's video link and everybody else's video link. They had the opportunity to, go into those black boxes and access to those black boxes. Whereas rest of the world had to deal with wiring the black boxes, whether the wires are literal wires or they're traces on a PCB, it ends up being the same thing, right? That's a great example of how, being able to integrate tightly outperforms system integration.
Peter: 28:04
Especially in the absence of standards that emphasize the performance qualities that your market needs.
Srdjan: 28:10
Yes.
Luka: 28:11
Now, how do you balance developing proprietary protocols that you need to optimize the overall system, how do you balance that with the need to interoperate in, call it the modern battlefield or in a commercial ecosystem?
Srdjan: 28:25
So that's also, that's a great question. I think, the higher up you are in, the, you can draw parallels with the OSI layers, higher, how much higher are in, in those layers, it's easier to be interoperable and the lower you are to the bare metal, it's hard to be interoperable. And the reason I'm saying that is, you know, maybe you will want to do hardware that's, It's gonna be compatible with the Pixhawk universe, so your hardware needs to be somewhat interoperable, in order to be compatible with, with, Autopilot stack. But then again, you also want some proprietary stuff running there. so it's really, this is one of the challenges of building a hybrid, solutions that combine your proprietary stuff to, open source stuff, if that makes sense. So this is a very tough balancing act. And again, it leads to compromises, but the thing that's, that, suffers is that full compatibility.
Luka: 29:32
So given the benefits of vertical integration how do you explain that, outside of a few companies like Orqa like Skydio, like Autel, I don't know, perhaps there's some others that you would like to mention, but how come that there are relatively few companies who are pursuing this vertically integrated model?
Srdjan: 29:52
I think first of all, vertical integration in the small UAS world is super hard, and this is because there are so many constraints, space, energy, swap in general. And there's a wide variety of different technology blocks, really, very disparate technology blocks that you need to, command in order to basically be in a position to system integrate, right? So, ranging from radio on one end of the spectrum to, I don't know, propulsion on the other. Completely different problem sets, but also interrelated in how they can affect each other in a very tight and confined space, which is a small universe. So, I think that like the threshold to, to actually be able to, start system integrating is pretty high in terms of like, simply starting from the number of different teams that you need to have in order to, to get started down that path. And I think, the type of talent that's required for certain technology blocks in this group, that they could be considered exotic, from the standpoint of like the current talent pool that the West is operating on.
Luka: 31:06
Do you have any observations more broadly on this market dynamics and how it is dividing into different segments where one segment is companies that are taking off the shelf components and putting them together for custom use cases, as we will in this discussion call them system integrators versus some that are vertically integrated, and then also, a handful of others that are also vertically integrated, but also very focused on a specific mission, like Manna like Wing for instance. How do you make sense of this segmentation of the market and the dynamics playing out in the future?
Srdjan: 31:42
I do think that there's always going to be need for system integrators. If you think of the long tail of applications, a lot of these applications in the long tail are going to be much more efficiently and perhaps effectively served by specific system integrated solutions. And this is why I do believe in a long term persistence of component ecosystem, which will serve a diverse and innovative space of system integrators that will go after these, more or less exotic or like, all of these applications in the long tail. And I'm a firm believer in that. And, there's also, this angle of like the solutions that are addressing the head of the distribution, like, your Matrice or your Mavic, they're going to be great for 95 percent of applications or 90 percent of applications. And that, they're not going to be that great for the rest 10 percent of applications, right? So, there's also that aspect of how useful is your Swiss Army knife? So, in a way, this is like going back to semicon. This is why, all those chips that we end up having to use, are not that great for some, niche applications, like, low latency video, or radio, right? So, so, this is the same, this is the same type of dynamic. So I do think that there's going to be, a vibrant ecosystem of, system integrators that are going to be served by hopefully as vibrant ecosystem of component providers. there's going to be demands for interoperability. And, you know, to counter myself, this is going to be driving towards commoditization because you don't have to have this interoperability and, commoditization there I can accept happening. But if we talk about the head of the curve and really going after these most frequent, most, numerous applications, then in this space, we're going to have fierce competition in performance between the players that are able to vertically integrate, and not only to vertically integrate their supply chains, being able to compete on the economics, but also quickly integrate their R& D, being able to dig deep into those black boxes, integrate them tighter, unlock those pockets of, efficiency that will compound into a higher performance over the competition. And in the small UAS world, that means flying longer. That means, safer mission. That means easier operation, all these things that kind of, make an overall user experience.
Jim: 34:21
Can you give a, perspective of near term market size for the component manufacturers, for the system integrator, for the service delivery company. We just heard from Bobby Healy a couple of weeks ago, about the rough size of the service delivery market, and it was in the billions and it was very attractive. But as you talk about the OEM, the component companies, the system integrators, can you give a rough size of near term market opportunity for each? And how much does that steer people doing one versus the other
Srdjan: 34:53
This is pure guesswork, I would venture like order of magnitude, tens to hundreds of millions, but like, it's a wild
Jim: 35:02
for the system integrators?
Srdjan: 35:04
per year.
Jim: 35:05
What about for the OEM? That's it for the
Srdjan: 35:07
Oh, it's in the tens of billions. Like just look at DJI revenue. So, if they have like 60, 70 percent of the market, we're talking about 10 to 20 billion ish. And set to grow.
Peter: 35:20
Are there drivers in the technology landscape that are creating a wave of demand for low latency video and emerging applications that leverage it?
Srdjan: 35:30
Funny you should ask that, but I think, this is a super niche thing at the moment still. This low latency video comes from this niche consumer FPV space, which is where, the pressure cooker of consumer FPV where Orqa was born. But, outside of that, at the moment, like, your typical unmanned, remote operation scenario is not that, latency sensitive. Now, the lower the latency is the more dynamic remote operation scenario you can tackle. And, I think it's like one of those things where what's possible technologically, really defines what applications can be, but for the most, remotely operated or like, and even autonomy, right, Everybody's talking about autonomy, whether it's land based or aerial. In any massively fielded autonomous system, there's always going to be a percentage of edge cases where a human operator is going to need to intervene to get the bot out of the situation. Even in, in like, ideal world of autonomy, there's always going to have to be an option to remotely operate. Remote operation is integral part of the autonomy stack because you will need that. Even five percent of edge cases creates a necessity for hundreds of thousands of hours of remote operation. And then the latency of your video is going to allow you more and more dynamic situations, if that makes sense, right? But one thing I'm going to tell you what we were pitching back when we were kind of goggle company and low latency, you know, this was, you know, 90 percent of our stack, we were pitching this kind of remote reality as opposed to virtual reality, which is a technology enabling, an operator a immersive experience of remote real world setting, in a way that would enable the dynamic interaction scenario. So you, it would actually allow you to interact with remote environment. But I think that's still in, in an SF realm, in an SF domain. no,
Peter: 37:49
Yeah, in autonomous systems, as most of us have been thinking about them, and the community has been working on them, you'll have video as a sensor, but a lot of processing of that data will be done right there behind the camera, and then only a very small amount of data is transmitted anywhere based on it. In this case, it's looking at, well, if you could transmit the full video, even over an RF link, even over a long distance, and you could do it at this type of sub 30 millisecond latency, what are the implications of that? What does that enable? Whether it's a human on the other end that is consuming the video, okay, that's remote operations, we know about that. Or whether it is machines that are consuming the full video on the other end, is that a frontier that we're looking at? Does that have implications for how you coordinate massive numbers of automated systems and enable them to coordinate very closely together? Is this an enabler to that, or is this a nice to have?
Srdjan: 38:52
I, I think you're onto something there because like, especially for smaller autonomous systems, if you can displace that power hungry, and hot running compute from a small thing that flies perhaps, then, you would need a way to, to get that, sensory input at lowest possible latency. And, it's much cheaper to run a video compression workload or video pipeline then it is to run a autonomy processing pipeline on a tiny thing that needs to be autonomous. So, so, there's certainly, displacing the compute, displacing the brain kind of thing, which is also an added, benefit to, to like having low latency.
Luka: 39:42
So there have been attempts in the past for Western drone companies to displace DJI, or at least, come close to it in terms of market share, but none really achieved any meaningful, interesting success in scale. Is now the right time? Is now any different? How do you see Western drone companies emerging as a credible threat to DJI?
Srdjan: 40:05
It is. And I think that the rest of the ecosystem, rest of the world is actually getting a second chance. And I hate to say it, it's part of the unfair advantage that us as Western drone companies have, DJI does not have access anymore to increasingly large swaths of market and applications, and this is creating a vacuum in which, you know, Western companies that are lagging behind in terms of, the performance capability of their system, that they can, try to compete amongst each other in filling that vacuum. So I think,
Jim: 40:44
You're saying because of the ban, because of the ban, that's why the market's being limited. Okay.
Srdjan: 40:49
Yeah. And then, the ban is creating a space where all of us, Western drone companies can have an opportunity to build something, to, because we, as a collective West, we lost, we lost the first half of the game, of the match in aerial robotics. fair and square, DJI won for better or worse, like, you'll have people saying, oh, yeah, they were subsidized. Yeah, well, probably, yes, they were subsidized extensively, but, we cannot hide from the fact that, they just did a marvelous job at engineering those solutions. And we have a chance to catch up, like collective West to catch up, leverage technology supremacy that we have in other fields, such as semicon. Collective West still has an edge in semiconductors over China. So, we are now getting a second chance to compete in aerial
Jim: 41:50
So if you were going to put together the dream team, what would you do to make sure the West catches up as soon as possible?
Srdjan: 41:55
Oh, that's a great question. That's a great question. You mean like on, capabilities level
Jim: 42:01
Well, there's a vacuum we've, well, we created the vacuum, right? And so we need to be able to create a Western superior product to the DJI. Who would you assemble? What would you assemble in order to be able to put something together that's better and price competitive?
Srdjan: 42:17
I strongly believe in the possibility of the DJI of the West. I mean, this is, this is our, this is the premise that, that Orqa is building our future five, 10 years. We think that we can grow into a DJI of the West. But I'm a strong believer in DJI of the West having a strong presence both in United States and in Europe. And I believe that being able to serve those two markets, in, in full sense of the word, like being able to have customer intimacy, being able to have strong presence, and being able to, share identity. And then not, it's not just like, Orqa as European company opening offices in the States. It's also about being deeply embedded into the respective ecosystems. I do think that, strongly believe actually that, this is the way to build a DJI Out West. Have a very strong presence in both of these ecosystems on both sides of the Atlantic and, be able to, to really, serve customers on both sides of the Atlantic. And, when we talk about serving customers, there's going to have to be a, super tight integration especially if we talk about on the enterprise level, let alone defense, which is a story in and of itself, but even on the enterprise level, there's going to be, there's going to have to be a tight level of integration with all these aerial data collection devices and, how the data is being used and processed and decided upon, right. So, so, so, that requires, deep understanding of enterprises in their all specific shapes and forms on both sides of the Atlantic. Because, arguably there's a difference in how utilities operate in the States versus EU, for example, right? And then, you can look across all these other potential application areas. I'm not talking about how it's technically operating, but how they're organized, how they function and all these sorts of things. So I do think that this is like a strong advantage to a transatlantic player over any player that's going to be solely in US or solely in Europe, right? And learning from DJI Playbook, integration matters on the logistics point and pricing matters. And like, EU has done a tremendous job of reaching out to, more Eastern parts of the continent and integrating those economies and leveraging the beneficial costs environment. So, there's a good 10 to 15 years where this is going to be an advantage until we kind of level off. So there's going to be a substantial advantage of having a European, a specifically Central and Eastern European base, in being competitive on a cost level, both on R&D side and the production side. So I think this is also an important factor.
Luka: 45:09
Let's talk more specifically about the emergence of a credible DJI competitor. You mentioned a couple of catalysts. One being the, vacuum that's been created by the DJI ban that's been growing in, in scope and we'll see how that discussion ends, but we haven't talked about Ukraine, which is another great catalyst for the entire industrial base. In a world that looks much different than back in 2015 in terms of the maturation of the industry, the components, the supply chain, will the winning playbook be the DJI playbook, or will the maturation of the market and the technologies needed to put together a system like that, will that create an opening for another playbook?
Srdjan: 45:57
I think that, Ukraine and War in Ukraine is a massive, contributor to this, to this dynamic in unmanned, especially aerial system, small UAS and if you think about it, it's a perfect storm. On one hand, like you have these large swaths of market that are not accessible to DJI, primarily in the defense and security sector, primarily defense, and then later leaked out onto the law enforcements and security and critical infrastructure, et cetera. And then on top of that, you have this profound revolution in how defense thinks about UAS combined and, reinforced, right? So, so, before Ukrainian war, defense thought about UAS in terms of fixed wing systems, predominantly half a million, thereabout price point operated on a relatively high echelon of command. And what Ukraine is proving very graphically today is that the bulk of utility of unmanned actually comes from small UAS copters, multi copters, operated on the lowest echelon of command and deployed in massive numbers. And then, if you combine that, like massive need for small copters in the defense sector, and then, defense sector itself being sterile of DJI, like, this is like a complete no go zone, like, maybe there's a gray area in law enforcement, arguably, law enforcement agencies still could use DJI in Europe, but in defense, no way. Right. So the defense is like completely free of DJI competition for all of us other companies. So, these two dynamics are creating a perfect storm for a step up, massive step up in demand for the types of solutions that, would be provided by would-be DJI competitors and providing this initial momentum that will drive the, innovative cycle. So I think, even without the Ukrainian war, this dynamic and this kind of DJI West playbook became viable. The Ukrainian war just cranked it up to 11, really.
Luka: 48:23
So fair to assume that a number of companies will come in and rush to get into and fill that vacuum. How do you expect the competitive dynamics to evolve on each of the segments? Let's look at, the consumer, the defense, the enterprise.
Srdjan: 48:38
So for consumer, I think, we have to look at different segments of consumer. There's like a, retail, consumer, market, which is, basically still, open for competition and DJI is untouchable there simply in terms of performance, the price, pricing power, it's the obvious choice for the consumer. Now, if we talk about educational segment of consumer type products, that's a different story because again, especially in US, this is also an area where Western companies have this unfair advantage that we can leverage in order to get business, be able to reinvest in R& D and provide solutions that are going to be fully Western made. So, there's these two different sectors. And we as a company, we're looking at educational space where we can provide fully Western made solutions. And then we look at really some niches where we can compete on a purely technological and pricing basis. Now, enterprise is still, I believe, somewhat contested, although I've heard examples where, companies are really reluctant to go for DJI or Chinese solutions. So, this gives an unfair advantage to, to, Western companies. I think this is going to be contested, and it's going to be harder to compete with solutions that are not, as performant as DJI, however you want to measure performance. But I do think that, as opposed to consumer, you will still be able to sell a good enough solution. You know, you don't need to be better, you just need to be good enough to be able to sell. And there's like a large portions of that enterprise market where this is going to be the case. And then defense is, defense is purely like, it's going to be, and it's not going to be, it is right now the space where we're going to see most of, the business, happening for Western drone companies, in the small UAS or multicopter space, which is typically dominated by DJI. I'm not talking about Fixed Wing type platforms. We have a lot of great companies operating in that space across the board, talking about really small copters.
Luka: 50:53
What are the major cost drivers for developing a small UAS system? What we hear is that for performance that is, somewhat approaching DJI's performance, you can get that from a Western drone OEM for three, four, five times the price, depending on, on, the specific use cases, et cetera. Can you generalize the major cost drivers and how certain catalytic forces like the Ukrainian war and the broadening of the industrial base, how that might impact those prices?
Srdjan: 51:27
I think volume is the dominant one. I don't think that prior to Ukraine, there was a conceivable scenario where a Western drone company would be in a position to sell a lot of copters to government users. Now with Ukraine happening, what we're seeing is a lot of, Western defense, end users are waking up to the fact that copters are going to be needed and copters are going to be needed in large volumes and copters are going to be used in an attritable manner. You deploy them and you're going to expect to lose them within three missions. So you better stock up on them. And not just that, you better invest in a capability, churn them out in large numbers, should you need them. Now this is a major contributor to, to really willingness to buy, willingness to invest in this capability, and like in the capability to, to develop these technologies, right, and support, Western companies and giving them business and contracts, et cetera. So I think volume is the main, you know, this understanding that, small UAS, small copters, that's, if you look at Ukraine, like this is all Mavic and Autels are the workhorses of the battlefield on both sides. And we as a collective West do not have a solution that's comparable to either.
Luka: 52:55
You've been an early and leading player in the FPV segment, which now has emerged, or has grown from a very niche to volumes that dwarf any other segment. So can you talk about how you've experienced this evolution and kind of the trends that you're seeing and how you see the future?
Srdjan: 53:14
It's interesting because, before the war, we, had all these different strategies on how to popularize FPV because, you had to explain to people what FPV is. And for very unfortunate reasons, FPV has arguably become a household word now. And, you know, how that plays out, and I'm talking as a company that, that was started as a hobby company, consumer company, how that plays out and how FPV is perceived after the war, you know, it's anybody's guess. And I'm not entirely comfortable as FPV enthusiast, I'm not entirely comfortable with that situation. I don't know whether FPV is going to exist after the war, or, will we be able to fly FPV as civilians after the war. So I'm not entirely comfortable with that. But what Ukraine started is like, it showed how effective, consumer grade technology can be in a modern war. And that's the, I think the most impactful part of the whole dynamic that, that developed because, what's being used in Ukraine is essentially a hobby grade, technology that's incredibly accessible from the cost perspective, from the technological perspective, you need a bit of skills to build, and quite a lot of skill to operate, this type of vehicle, but nevertheless, super accessible. Like I cannot be expected to build a Javelin myself in my garage, but I can build an FPV in my garage. And with an FPV, I can probably hit a tank at a greater distance than I can with a Javelin. And I can bill it for like, what, 300, 500? And, Javelin is what, hundreds of thousands? So, this gives you an idea of how much disruption happened
Jim: 55:05
Can you just describe first person view? What is it? What's its value? What happens if somebody doesn't have an FPV capability for those who may not know as much about the market?
Srdjan: 55:16
So FPV stands for First Person View. It's what we used to call our hobby prior to the war. FPV today stands for a first person view, manually controlled, quad captor. That is a radio controlled, a drone that's capable of carrying an explosive payload and it's operated fully manually, like as you would a race drone, by a skilled operator. And, this is basically, you can fly a drone for, a couple of kilometers and you can, strike an armored vehicle, which costs in millions with an asset that's, you know, the, some people call it an, like a flying IED that, that costs like hundreds of bucks, to build. And, we've seen a lot of companies coming up, in that space, being a component provider, we could see a lot of Western companies coming into that space, but in reality, Ukrainian war in particular, although that created a lot of interest in the West and that, created a lot of companies that are looking in that space, realistically, this again, was dominated by Chinese companies who were well placed, to serve both the needs of Ukraine and Russia in providing either components or build systems. And I think, if you look at, how various Western companies performed and Orqa included, trying to compete in that space, like again, we collectively couldn't compete with the Chinese on, on, on prices. And I don't think we should, because longterm, I think that FPV is specific to Ukraine and it's a phenomenon that's, indicative of absence of other means of strike by both sides, and, whether, strapping an explosive ordnance to a copter, whether that's going to be a optimal way to do precision strikes, that's anybody's guess. And my guess is, you know, probably there are more optimal ways to perform strikes, kinetic strikes. But what I'm certain of is that copters are the optimal way to provide battlefield transparency, the type of which we're seeing in Ukraine right now. And, as much as I don't think that the FPVs as a strike capability are going to stay here post Ukraine, I'm 100 percent certain that copters, which may or may not be FPV and arguably don't need to be FPV, will persist as a principle way to provide situational awareness and battlefield transparency, in the future. I'm not saying that the fixing platforms are going to be done, but I'm saying that, we're still going to have these fixing platforms operated on a higher echelons, but we're going to see much more copters being operated on a very low command levels. The challenge is going to be really collecting all of the data and then processing it.
Peter: 58:35
What have you observed in terms of how FPVs are being employed in Ukraine? How has it changed since early 2023 to today? What do you see coming in the future, and what is it about that that makes you think that FPVs are perhaps unique to the Ukraine conflict and that they're not going to continue to evolve in this way for future conflicts that they'll be supplanted by something else. I think that there are people out there with, the assumption that this evolution, what we're seeing in Ukraine is going to continue and that Western militaries across the spectrum are going to be rearming with, massive numbers of FPVs and incorporating these into their tactics as a lesson from the war. But if you have a different view on how that's going to look, walk us through that.
Srdjan: 59:24
Oh, yeah. I think, if, if you wanna have the nitty gritty realities of use of FPV in Ukraine, I strongly advise, listening to War on The Rocks, podcast and Russian Contingency where, Rob Lee and Mike Kaufman are really, delivering direct from the battlefield experience of FPV operators. And this is very detailed and highly informative in terms of tactics. So, what we're observing, is basically there's, some cat and mouse development in terms of electronic warfare countermeasures and being able to deploy these systems, freely, so there's a big change between, what was working at the early stages of war to what's working now. We're seeing a lot of, evolution in terms of how the electronic warfare measures are deployed and how the, systems are evolving to, be able to, operate in the presence of, electronic warfare. So there's that. These FPV systems are being, operated in an increasingly larger numbers. And basically both sides have been very successful in adopting those and evolving the tactics of their use. But ultimately I think, again, this is a specific for Ukraine because, simply because barring, drones, FPV drones and drones in general, and the anti air systems this is a very low tech war, ultimately. And, these FPV systems are being applied in a, a bolt on manner. And whether I'm seeing FPVs that are being used in Ukraine right now being used by, uh, in this form by Western NATO member countries, I would say no. Just starting from the fact that once you power up the drone, you're broadcasting unencrypted video feed to whoever cares to tune in. Like, just start from that. And, this goes from the fact that because of the latency, and because of the EW resiliency, you're compelled to operate an analog video system, and there's no efficient way to encrypt or scramble analog video system, just that, that leaves you with having to broadcast your mission, to whoever wants to tune in in the vicinity. So, so there's that. Also, like, I don't think that we should be falling into a trap of thinking that these are super efficient systems simply by looking at social media. And this is one thing that, that really, repeated by the people who had the chance to really observe how these systems are being employed. Like the efficiency numbers are super, super low. because. If we look on the social media, we're just seeing the strikes, but, then there's a massive, positive selection, happening there, right? So you will not see unsuccessful strikes being posted on Twitter and, you know, all the FPV pilots, from before the war, they know that, whenever you fly, you turn on the DVR because, if you lose a drone, that's how you find it by looking at the DVR footage. You can imagine that if every FPV mission is filmed and the operator has the footage, then you could even, say that we have seen every successful strike that happened, right? And we can be fairly certain that most of the successful strikes have been actually recorded. And in the world of social media and viral dynamics, this creates a powerful echo chamber where we all think that this FPV is this super efficient new asset, which, you know, we really have no grounds on making that conclusion because we don't know how many unsuccessful ones there have been for every successful one that we see. So, this second piece is efficiency. Third piece is, FPVs are really hard to operate. It takes a lot of skill, and this was one of the barriers to entry to the hobby before the war, which we were facing as a consumer company, like it's really hard to get people to remain and then to persist in the hobby long enough so they can enjoy it. Like, you really need to struggle before you get the skill and you can actually enjoy flying because it's not that easy. And I have hard time imagining, FPV being deployed in a big army sense. I can totally imagine it being deployed in a highly specialized manner. So, I can totally imagine FPV being a part of a special operator's toolkit, like that skill set, right? So, you train for a sufficient time, you become a great pilot. And then this gives you almost superpowers, right? And then arguably this can be a, a Special Ops thing rather than a big Army thing. Then again, simply the logistics of, deploying large number of FPVs is a stretch to be, you know, diplomatic. I mean, I'm going to give you an example. Whenever we would go out and fly for fun, there was always problems with frequencies and 5. 8 band, which is where you, where your video operates is very narrow. So you effectively have eight channels. And if you go out with a group of friends, like five or six of you, it's a logistic nightmare to make everybody stick to their own frequency. Right, so you always have this situation you're flying and somebody plugs in and they're on your frequency so all of a sudden you can't see anything. So trying to extrapolate that experience to a chaos of war where there's not only your side trying to deploy these missions en masse. There's adversaries trying to deploy these missions, like it's a nightmare. it adds to that thing about how efficient it really is, this question. So, compounding these five, different aspects of, like, reality of FPV I don't think it being used massively for strike. What I am, again, I'm going to say it, what I'm seeing this Ukrainian war changing in a way, how, army uses, copters is this situational awareness, mass deployed, highly attritable platforms for, defense applications. And, as an FPV company, we get a lot of, oh, FPV is the big in Ukraine. you guys, must be selling hundreds of thousands. And I'm like, no, the only ones making money in this FPV space in Ukraine are Chinese companies that, that perhaps they were making ready to fly drones in maybe single digit thousands per batch. I can totally imagine a lot of these companies doing tens of thousands per month, easy. And that led, funnily enough, that led to the commoditization. the price per system is like 500 bucks.
Luka: 1:06:26
Isn't there an argument though that technology will continue to innovate and try to close some of the gaps that you were talking about, whether this is better use of the RF spectrum, better, digital data links that are also resilient to EW, better logistics, better deployments, better ways for humans and machines to interface. Isn't that a valid argument that this arc of technology evolution will continue and make FPVs relevant in the future? Or if not, how do you see that arc continuing to evolve?
Peter: 1:07:00
Yeah, and I would add increasing automation of the flight control itself. All of those things, what would you want to see in terms of changes across that spectrum of technologies that would change your conclusions?
Srdjan: 1:07:15
So, I think, one of the gaps, certainly is, spectrum management. It's also EW resiliency for the high throughput data links, and video arguably is a high throughput data link. It's very hard to sustain sufficient data throughput over an RF data link for digital video. If you think that even 10 percent package loss of a compressed video is total, video loss. That's, super hard to sustain even in a congested RF environment, let alone contested RF environment. So, so, that, that's going to have to change in order for, these systems to be viable and a prerequisite for like, I think wider adoption, FPVs is going to have to be moving away from freely broadcasted analog video, which all of the FPVs you're using both in defense and in the hobby are being used, are using right now. I can totally see a benefit in adding automation and autonomy to these systems. It's gonna reduce the barrier to entry for skill required to operate. It would not have to be a matter of dexterity and ability to, to really, fly and, skill it's gonna, it's gonna be learned and it's gonna, you'll be able to learn more efficiently. But also, and I'm going back to this point of whether copters are the right strike vehicle. And that's always this thing that I'm coming back to. And I know that this trope of swarm of drones is a very attractive one, but have you ever seen how drone light shows are being deployed? It takes a stadium to set them up. So you're literally setting up an acre of drones almost in order to deploy a light show. So this trope of having swarms of drones attacking, like you would be, you, there, there's a lot of logistics that needs to go in, in deploying a large number of drones. And if one thing that we've learned about this war is that in a transparent battlefield, you need to reduce your logistic footprint, not, so, so in, in a way, we're going to have to, innovate, like collective we, we're going to have to innovate heavily in order to come up with a smaller footprint for deployment swarms, if we want to make swarms a thing. And then it brings me to the question, is copter really a great platform for strike? Or are there more, like, compact, efficient platforms to deploy strikes at that distance with that type of payload. And again, I have to go back to saying, expressing an opinion that, I think that both Ukraine and Russia are using FPVs as a matter of necessity and accessibility. It's more of an accessibility thing, because all you need is 500 bucks of Chinese components, some carbon, and explosives, to be able to affect a, precision strike beyond line of sight. And that's it. and, couple that with virality of the footage that, we established is recorded for every mission by default, because this is what you do as an FPV pilot, you DVR everything. Like, when you add that two things together, like, you get this impression that, oh, it's a massive thing, whereas, reality might be completely different. But what we are seeing is, a lot of Western forces are, working to, acquire these capabilities to, learn how to deploy them, learn how to use them, learn how to defend against them. And I think this is going to be here to stay, but I think that, this is a starting point and it will necessarily evolve and what's the end state, the, final form of it. Anybody's guess. And I'm really not. even, arguably being an insider and understanding, this, deeply because, we were doing it before it was cool. I wouldn't dare to venture where it lands, like, what's going to be the final shape of the strike vehicle, and I'm not sure it's going to be a copter. On the contrary, I'm kind of willing to place a bet that it's not going to be. Where I'm willing, again, 100 percent to place a bet is like for, low level ISR situational awareness, there's hardly a better thing than a copter.
Luka: 1:11:55
Srdjan what's the message that you'd like the audience to take away from this discussion?
Srdjan: 1:11:59
Well, two messages, like, first of all, building small things that fly, it's super hard. And people should be aware that, although there is a massive, advantage of vertical integration versus system integration, we should not underestimate system integration because that in itself is also super hard to do right. So I think, ultimately if there's like one or two things, this would be that, this is hard, this is very hard. And, people that are doing this right now, are really, the barrier to entry is quite high.
Luka: 1:12:34
Excellent. Well, thank you very much for this insightful, engaging conversation. It was great to have you on.
Srdjan: 1:12:40
Thank you, it was, as interesting as always. And I'm not at all surprised that we're over time. Take care guys.