A conversation with Yuriy Filatov, head of the unmanned systems department at the headquarters of the Third Assault Brigade, was postponed due to Russian drone attacks on Kyiv. However, the topic of drones was precisely the reason for the conversation. Therefore, an ITC journalist contacted Phil immediately after the all-clear to discuss the ongoing “drone war”.
For over an hour, the representative of the Third Assault elaborated on the direction in which drones will develop, the current level of Russians and their flaws. In addition, the interviewee pondered whether there is a future in machine vision and the so-called “drone swarm”, and how to join the unmanned systems unit during mobilization actions. The conversation took place during a large assembly for the 3rd Assault Brigade on the acquisition of combat UAVs.
How did the war start for you?
The war started on February 24th with shelling, like for everyone. I live outside the city, so missiles flew over my house. Then there was TR (Territorial Defense) from the first days, and then I joined the Third Assault.
Did you start working with drones right away?
At the beginning of the war, it was not so much about drones as it was just chaos and confusion. It was necessary to organize somehow, to avoid playing guerrilla warfare. So the task was a bit different – to train people. Many came who had no experience with weapons. Therefore, the tasks, accordingly, were such that these people needed to be trained, prepared. And then there were trips to the same Kharkiv region, Luhansk region. When it became clear that this was a war of a slightly different format — not a war of individual companies or even battalions — and it was necessary to join larger units. And the Third Assault, perhaps, was the best choice.
Why specifically an unmanned system?
Because that’s my background – I’m from IT. I know microelectronics, I had drones long before the war, I understand how to manage and produce them. Accordingly, I joined my friends when I joined the unit – it was a UAV company. Practically, pilot systems were then the only component presented, and the first companies of unmanned systems in the Armed Forces were experimental.
No, just microelectronics – that’s my specialty. And then it remained as a hobby. But fundamentally, even if we look at FPV drones now, they are frankly and honestly toys from the store.
How is the system of unmanned units structured in the Third Assault? How would you describe it?
It’s already complex. I can say that we are leaders, definitely in the Armed Forces. In the Ground Forces too.
As part of the “Aratrash” assembly, we are supported by the Braveproject project. Together, we are raising 10 million for the “Most Massive Art Destruction of the Enemy”. They have an interesting approach to fundraising by providing additional value to donor companies.
It’s all of that together, we can say frankly. We now have metrics that allow us to measure our effectiveness, we know the numbers.
Can you name at least one figure, some kind of efficiency? For example, out of a hundred sorties, 60–70 are successful.
I can say that. For example, our FPV efficiency is around 65%.
And usually, how much can they be?
It varies in different units. But if you take a large sample, then it’s probably 30–40%.
Hard to say because I’m pretty sure that it varies depending on the units. I know we intercept a lot of theirs, but fundamentally they are also not standing still. To say that we have an absolute advantage, no. As soon as we stop developing somewhere, they catch up with us through volumes, state programs.
There are actually many factors at play here. Quantity turns into quality. And for that to happen, it’s necessary to build an appropriate system of resource management, a system for personnel support, and so on. It’s a large complex. And accordingly, this complex goes through stages of development. And when you have, conditionally, only a company of strike UAVs, that’s one story. A company – 100 people. When everything starts developing, when the corresponding units appear in battalions, then it’s a completely different story in terms of organization, management.
Now they are even making regiments.
But there are also aspects here, that the regiments that appear are specialized regiments of unmanned systems. And they must act, helping infantry, assault, or mechanized brigades. And there are pros and cons here because, from experience, it’s still better when the unmanned component is located as close as possible to the infantry in interaction with the artillery, with the infantry. And when all this works as a complex, accordingly, then the highest effectiveness, awareness. Because only units that are in the zone, which have their area of responsibility, are aware of enemy logistics, know where to strike, lay mines, where to expect the enemy, and with what forces. When some external unit arrives, it needs time to understand how the enemy operates, how to counter them.
You mentioned the Russians. You can read many different views on whether someone says they are ahead of us with drones, someone says we have the advantage. How do you see it?
Effectiveness is measured by criteria. It’s either the number of hits to the number of destroyed, or the total number of hits relative to other means of destruction, etc. There can be just different criteria. The front is quite different in different parts. I know that in our area, we restrain the enemy more due to unmanned systems.
Is this the Kharkiv area?
Yes. Our task, as unmanned systems, is to prevent the enemy from reaching the infantry.
That’s not the main thing. In fact, the main thing is the culture of application in general. Not only the experience of managing the drone itself is decisive. The main thing is everything in the complex because you need to know which equipment to apply, how to set it up, how to select it, which combat part to use, how to choose a position, how to set antennas, how to use additional various components. And only having all this, can we talk about the effectiveness of the crew or the entire unmanned systems unit as a whole.
How long does a person need to train? Is experience necessary?
It varies. We always have to understand that it’s not just one person because we are constantly recruiting and employing people, including into unmanned systems. And many come with the idea that they will be pilots, will operate FPV or something else. It doesn’t work like that. It works as a team. And it consists, not only of the pilot. Here there is also an engineer who assembles munitions, a master who sets up drones, an operator who will aim, turn the antenna, deploy a repeater, a calculation commander who will choose the target and help aim. And all these roles are equally important: if any one of these components doesn’t work, there will be no result.
The answer to your question is somewhat complex. It’s not about piloting a drone. Learning to control a drone can be done from the comfort of your couch by downloading an emulator; you can fly better than any athlete could. Navigate obstacle courses and fly through windows. But this doesn’t say anything about the effectiveness of the calculations.
You mentioned recruiting. Indeed, this aspect definitely differs in your organization – and the billboards are also eye-catching. Is the recruitment of pilots also different with you?
Yes. We have our own pilot school. It is located in Kyiv, easy to find. There we conduct training. You don’t necessarily have to become our pilot; perhaps someone might simply be interested in coming to learn for the future, or even to join a different unit. Also – even when someone enters the TSK, they can show our certificate and they will likely be directed specifically to unmanned systems units.
So, in our pilot school, we conduct lectures on drone building, on training. We have a range, an arena that emulates an obstacle course for drones. There’s even equipment that you can try to hit.
Perhaps it should be mentioned that your pilots focus on piloting; they don’t switch to other tasks, say, to infantry.
Actually, there is a strict prohibition against using pilots for other tasks. This ban is effective across all the Armed Forces of Ukraine. Why? Because pilot training is indeed a resource-intensive process, it takes time. And losing a pilot would be more costly for everyone.
Many people have passed through your unit, many different cases have occurred. Recall something from real events that impressed you, perhaps from the work of your pilot.
You know, war is such a thing where every day is filled with events. For example, yesterday several units of armored vehicles tried to storm us, there was a tough fight. And each situation is perceived differently because at some critical moment a successful hit can stop the assault and so on. On one hand, it’s all very vivid, on the other, over these years the perception has somewhat dulled. So, to recall something striking right off the bat, I can’t say that I can.
And the Russians, specifically the pilots, have they become more professional? How do you see it?
First and foremost, we see technical progress, they are indeed evolving. They have significantly strengthened the components where they were behind us, they are catching up with us in the use of FPV drones. The Russians no longer use such blunt strategies, for example, frequencies, as they used to.
Still, at this time, do we still have the advantage in evolutionary development?
Probably, yes. And there’s a rather simple explanation for this: we are quite diverse, in any garage any team can start assembling drones. They all come out different, but from this diversity, we can always choose in a broader range. And this is our strength. It allows us to develop, to stay technologically a step ahead, because we have a lot of talented engineering teams working in various directions.
With the enemy, it’s more centralized. When they, conditionally, make some model, it immediately reaches the entire front and in large volumes. But the variety is much less. And these things, accordingly, affect the speed of updates, the speed of development and innovation. Because if something stops working for us, we simply choose something else within two weeks.
The inertia of their state machine, state orders, is much greater. Accordingly, their speed of innovation is slower. Therefore, for us, it’s also very important to always maintain a compromise between state orders and the speed of innovations – as quick feedback from combat units to the manufacturer as possible, with no unnecessary intermediaries.
“Aratrusha Collection” from Braveproject. We are raising 10 million for “The Grandest Art Destruction of the Enemy”.
It will definitely progress in several directions. And this is very good because fiber optics has its advantages and disadvantages.
What are the disadvantages of drones with fiber optics? The advantages are clear – electronic warfare doesn’t work.
The disadvantage is quite simple. Actually, any additional load reduces the useful load on the drone, when it carries an additional coil of fiber optics, or a larger drone is needed by itself, accordingly, a larger battery. So today, fiber has limitations in range or there will be a smaller combat part.
Is there an understanding of how to combat drones with fiber optics?
There are ideas. These drones are used only in a few units, so the counter-technologies are also just emerging as a response to the new threat. So I think it’s a bit early to talk about it out loud.
Regarding the development of FPV drones, there are two paths: either to reduce the cost of the product or, conversely, to increase the cost of the technology – machine vision, swarming, and so on. In which direction will FPV drones most productively develop?
The FPV drone is good in the form factor it has. It is effective because it is quite mass-produced, and again, the economy of war dictates the direction of development. Accordingly, we can make something, conditionally, from a drone to a cruise missile, which will cost corresponding money. This automatically means that we will have enough resources, conditionally, for three missiles. But this is not the weapon we need. We really need something mass-produced, as cheap and effective as possible.We know, understand, at what distances it must work, what tactics of its use are, and accordingly, for different purposes, there will be different, more specialized means.
We will have a mass-produced, cheap drone that we can use on the front line to provide sufficient cover against assault groups. There will be drones against dugouts. There will be more specialized drones against armored vehicles, against tanks, etc. There will be even more specialized techniques, it will be long-range – there are completely different requirements there. But it can already be said that there will be specialization of these means. Teams will have different drones to choose from, depending on the objectives.
And regarding machine vision and swarming – what about these technologies?
Swarming is a hyped topic, everyone talks about it, but no one can say exactly how it should happen, what is meant. That they all fly together? Well, good, they flew. And then what? What will they do? Machine vision has already been used on various drones, has undergone a path of evolution. It’s a working topic. It’s not such a mass product.
And why not?
It absolutely makes no sense. It will occupy a niche of some specialized equipment and will be used for special tasks. For example, in harsh conditions of electronic warfare, when other drones cannot fly in, using a drone with guidance just solves this task: to knock down EW, disable the first vehicle, block it, etc.
I also want to talk about bombers. Do you use them at all in units? Is there any evolution? What does it consist of?
Firstly, I want to remind you that we are collecting for 3 OSHBr just on large drone bombers for two leading groups of pilots. Regarding the bombers themselves – we had R-18 drones from Aerial Reconnaissance at the beginning of the full-scale war. There were also several developments. Of course, now they are completely different drones, they have also evolved.
Also, bombers have received their specialization because we divide light drones, and heavy bombers have also developed. And even within this large class of heavy bombers, they have been divided into short-range and long-range bombers. We are currently raising funds for the latter.
What is their effectiveness? How are they used?
Range up to 15–20 km. They are already solving a completely different layer of tasks, that is, their main task is not to operate along the front line, but deep into enemy territory, destroying their artillery, mortar crews, logistic depots, equipment in protective shelters, etc. Thus, 20 km is already a normal brigade level, and basically, it shows their quite high effectiveness.
There is also a trend with bombers: if they are elevated higher, then electronic warfare will not work and the shells will be guided?
There are currently developments regarding guided munitions. This is a separate direction. It is quite an obvious path of development, as the methods of warfare are becoming more complicated. Solutions are emerging for UAVs regarding specialized munitions that can be controlled during release.
So drones are turning into mini bombers.
In fact, that is the case. Therefore, the development of munitions is also progressing in turn. Yes, developments in intelligent or guided munitions – that is also currently in the works. We are communicating with many teams about this. And I think that one way or another, we will see this quite soon on the battlefield.
Collection “Aratrash” from the project Braveproject. We are raising 10 million for “The Most Massive Artillery Demolition of the Enemy”.
To be honest, we haven’t seen bombers from them, except for “Mavics”. But lately, we have noticed their use – they are trying to use large bombers as well.
It’s always a complex of actions. Electronic warfare will always be part of these measures, as long as it is possible to suppress and disrupt enemy control of the system. But not only that. It’s no secret that we are also quite effectively intercepting enemy “winged” scouts using relatively inexpensive FPV drones. I think that we might also move in this direction. At least, the enemy is already countering our bombers using FPV. Trying.
I’m looking at the shooting down of scouting drones. Yes, very successful, but it won’t be long before the Russians install a rear-view camera and the drone will start maneuvering, dodging FPV. How will technologies develop here?
They have already done it, they installed the camera, they try to evade our FPV. It’s action and counteraction, it always happens that way. Why are long-distance “scouts” all in the form of a wing? Because a “wing” itself is a much more efficient means in terms of energy expenditure for movement. Because a “wing” creates lift and in the air uses the engine only to create this flow. But there are limitations in “wings” in the ability to maneuver: they are much more rigid. It will not be so easy for FPV to evade a “wing”, so let’s see.
There is a trend – anti-aircraft drones, some have already started using them, at least experimentally. Can we also use anti-aircraft drones against “scouts” and against bomber drops?
It’s worth trying, but it’s not that simple there either.
What’s the complexity?
The drone is the smallest, last piece in this puzzle. Actually, the simplest part in the whole complex of actions regarding the interception of “wings”. Accordingly, there will be various nuances there.
Again about anti-aircraft drones. This is also quite a hyped topic. What do you think, does the technology have a future? How will they be used? Also in specific cases?
Interceptor drones are actually an example of how not to do it. Because here we were one of the first brigades that started doing it systematically. And we did it quite successfully, for quite a long time.
And quietly.
And quietly, yes. Until unnecessary hype and publicity did not make it. This is the case when it’s better not to hype, but to just do your job. We trained neighboring brigades, trained almost the entire sector. And everyone started doing it. And taught everyone else without any fuss. We ourselves revealed our cards and effectively lost the advantage. That should not have been done.
You mentioned about the state and private sector. Many talk about the supply of drones. If you read posts online, it ranges from betrayal to victory – in one line. Still, this million drones, two million from the state – are they being produced? Are they delivered to the front or not? Then, what volunteers periodically collect, does it supplement state supply or completely replace it?
You can notice that the same experts may comment on everything – from ballistic missiles to the number of drones. That’s nonsense. I think only the Ministry of Defense of Ukraine knows about the number of drones purchased. Commenting on the number, a million or two, can only those people who are related to this. Basically, compared to even half a year ago, the supply has increased, at least what we receive. However, it is still not enough. We use much more than we get from the state, but the trend is positive, we really feel – the supply has increased. Is it enough? No, it’s not enough. How much is not enough? I would say that now we buy about 70 percent, if we talk about FPV, through various channels independently.
That’s a huge percentage.
But here you also need to understand. It’s us, because we place a big bet on drones. We use them a lot because it’s the deliberate policy of the unit commander. FPV – that’s what we can buy with money from communities, state money, volunteer money, from collections, etc. This is what we can easily produce, easily buy, and apply quite effectively.
With private suppliers, it’s easier for us because we order and receive in relatively small batches. Then we rework them at our expense as needed. We have situations when we receive some products, some drone, or system, and it’s non-functional, already outdated. And we have to take and at our expense, spending our time, organize large workshops that will modify and adapt the received drones to real conditions.
Does the unit have its workshop that modifies drones?
This is already part of the specificity of unmanned systems in general. This applies to UAVs and ground robotic systems. There is a need to have our engineers who can quickly respond, change something, because circumstances change.
You said that a lot of assistance is specifically volunteer.
Where do our supplies come from now? Apart from state sources, it’s money from communities through special subsidies and collections that we conduct. This is how we specifically meet our needs. Why is this important? Because here we are really quite flexible in what we need, and we can order directly from the manufacturer exactly what is needed. And knowing, considering the experience here and now, we can do it much better. Therefore, I also advocate the market for unmanned systems.
But when we face monopolizing factors, some centralized procurements, when, conditionally, you can precisely pour money into one manufacturer, accordingly, such things actually kill this market. Why should someone do something if he will get the contract anyway? He got, for example, a contract for a year, and he can even not listen to what the units will tell him because he already has a contract.
Do such volunteer collections help you be more efficient?
Yes, and they support this ecosystem of innovation and everything else much more than any state programs. As grateful as I am to the state, money still drives the economy and economic processes. Flexibility and opportunities – that’s what volunteer collections give us.
The “Aratrush” collection from the Braveproject. We are raising 10 million for the “Largest Enemy Art Crush”.