FPV drones are small, highly maneuverable unmanned aerial vehicles used for reconnaissance and short-range strike missions. They are fast, difficult to fly, and often used in kamikaze operations or as communication relays. Due to their high mobility and relatively low price, they have become one of the key tools of modern tactics on the front lines, but at the same time, they have a number of technical limitations — radio horizon, vulnerability to electronic warfare, and the need for experienced pilots.
NORDA Dynamics is a Ukrainian defense tech company that grew out of a volunteer movement. Today, its automatic drone guidance and optical stabilization systems are actively used on the front lines, and the team recently raised $1 million to expand development and increase its staff.
AIN.UA spoke with NORDA Dynamics CEO Nazar Bigun about the journey from a volunteer movement to the production of tens of thousands of drone auto-delivery systems, how the company is using $1 million in investments to expand its team, what product it is currently working on, why the team received the fastest feedback from the front line in just 26 minutes, and how Ukraine is forming a new military technology industry.
How did NORDA Dynamics appear?
I started looking for like-minded people with whom I could do all this together, because it was clear that I couldn’t handle all these requests on my own. That’s how I met Dmitry, our COO, and a few other people. And we created a volunteer organization called “Help a Hero.” Basically, everyone there was from IT, engineers, and our focus remained the same — various devices.
At some point, the topic of FPV drones became quite hot, and we received several requests to deliver FPV drones to our military. We started looking for where to buy them. We were told, “Pay for the drones, wait 3–6 months, and you’ll get your drones, because we have a huge queue.”
We realized that in 3–6 months, we would probably learn how to make them ourselves. Then we found two more future founders of NORDA. One of them had been involved with drones as a hobby for quite some time, and the other was from a volunteer organization. We set up a small volunteer drone production facility, which is still operating and is now not only producing drones but also making about 200–300 FPVs per month.
Since we already had production facilities, we began to delve into the subject and investigate why efficiency on the front lines was so low. At that time, it was around 8–10%, meaning that one out of every 10–12 drones reached its target.
We identified the main problems: radio horizon, inexperienced pilots, and electronic warfare (EW).
The logical solution to these problems is the so-called last mile system, or automatic target guidance system. We started looking again for where we could buy such a system, and at that time we found only one very early MVP solution in Ukraine and several very expensive foreign ones. Then we realized that there was a market, demand, and a problem that needed to be solved, and we started developing our own system.
The hackathon organized by the Ministry of Defense of Ukraine and Brave1 in early 2024 was a turning point and a very strong boost for us. We went there to learn about mature solutions, because ours was still at the MVP stage, and we were just sending the first samples to the front. But we won that hackathon, and it turned out that there was nothing better in Ukraine. Since then, we have been receiving a lot of calls from the military, volunteers, drone manufacturers, and investors. It was a very strong impetus for our development.
About the products
We currently have two products that are actively used on the front lines: a last-mile system and an optical stabilization system for repeaters, which can hold the drone’s position very steadily. We have already delivered more than 50,000 auto-delivery systems to the front lines and have grown to 40 people.
Underdog — an automatic guidance system, or Terminal Guidance, or Last Mile Targeting System — is essentially a small computer installed on a drone controlled by a pilot. When the pilot sees a target, he locks onto it (points to it), and our system begins to guide the drone toward that target. Then there are two options. The first is that the pilot switches to a conditional autopilot, and our system controls the drone without needing to communicate with the operator. It will independently guide the drone to the target.
Underdog module
The second option is that the pilot locks onto the target, our system tracks it, and when communication with the pilot is lost due to radio horizon, electronic warfare, or other reasons, control is automatically transferred to our system. It then guides the drone to the target. This system works on aircraft-type and quadcopter drones day and night.
We are currently increasing the level of autonomy by integrating AI models that will perform automatic target search. This is the next stage in the development of the systems.
StableLink is an optical stabilization system. Essentially, it is a small box with an equally small computer and a digital camera that attaches to the bottom of a quadcopter. The drone can then take off and land automatically. A repeater, for example, is attached to this drone, and it maintains altitude and azimuth. This frees up the pilot, because he doesn’t have to watch this drone at all and can focus on the kamikaze.
StableLink module
In addition, this has significantly reduced the cost of the carrier, because previously the military used DJI Matrice for repeaters, which is a drone that costs around $10,000, and now they use ordinary copter drones, which cost around $1,000.
We have also adapted this system so that it can automatically lower the drone, drop the mine, and automatically rise. In other words, it is used for autonomous mining.
Our systems work with 99% of the drones used on the front lines. However, the system does not work with DJI, Mavic, and so on, but there is no need for that.
Read more: Russia’s GPS and GNSS interference disrupts flights: airlines and authorities respond
How is the issue of system stability resolved in conditions of GPS jamming or lack of signal?
This is the next stage of development, because our system relies on a visual sensor. That is, while the pilot is flying, he does not need GPS, because he is controlling the drone. Then he points to the target and the drone automatically flies to the target. But for longer missions, mid-deep strikes, you can’t just fly there because you won’t have communication the whole way.
For this, we need what is called GPS replacement or GNSS denied navigation. We are currently working on our own development, as well as integrating several solutions from our partners that rely on either visual odometry and map matching or beacon navigation.
About competition
Nowadays, it sometimes seems that every garage is developing an autonomous driving system. In reality, there are probably about 5–6 major players who are actively developing their systems. But there are many more junior solutions.
Many teams are working on similar solutions, and there are several companies that are significantly ahead. Therefore, it seems to me that the problem of auto-delivery or the so-called last mile has already been solved.
And so when I see young teams trying to implement it again, I try to convince them that there are many unsolved problems. And it doesn’t make sense to work on this if you don’t have a unique algorithm, and you can’t do it better than other teams have done.
What are your advantages over your competitors?
In my opinion, a significant advantage is that we have a very large network of military personnel with whom we cooperate. And since we are a volunteer group, we have a certain amount of trust from the military, which helps us get better feedback and communicate more effectively.
We were also the first to scale our solution to tens of thousands on the front lines. While we had thousands of systems on the front lines, everything worked, and we received excellent feedback. When we scaled up to tens of thousands of systems, we received some negative feedback. We identified many different problems and edge cases rare, non-standard, or extreme scenarios of system operation that are at the limit of its normal conditions of use on the front lines, which took us about eight months of continuous improvement to resolve.
Underdog and Stablelink modules
Right now, our competitors are in the process of scaling up to tens of thousands. This is a path you have to walk with sweat and blood because there is no way to skip it. And I believe that this is our advantage in terms of time.
About the international market
We are currently only operating in the Ukrainian market. We have several contracts in the process of being signed, and together with other teams, we are working to lobby for the simplification of export controls. We want everything to be transparent, with minimal bureaucracy and no corruption.
We understand that our technologies are now significantly ahead of any other technologies in the world, except for those of our enemy, which are also developing rapidly. And this is our great advantage in global markets. This is something that can greatly increase the revenue generation of all teams involved in defense in Ukraine and help Ukrainian defense gain momentum.
Currently, the export procedure is very slow, and this is what is blocking us from starting to sell abroad, but we are working on it.
What, in your opinion, are the main challenges for developers of autonomous control systems for UAVs in conditions of active electronic warfare?
The first is the lack of qualified personnel.
The second is working with cheap hardware. Since we have limited budgets, we always have to keep in mind the limitations of the hardware we use. This is a complex engineering task.
Plus, when it comes to artificial intelligence, and not just computer vision and mathematics, there are many edge cases where machine learning does not work as it should. In addition, there are problems with masking, weather conditions, and so on. This is also a very interesting task, but at the same time, it is a major obstacle to making all drones super-intelligent.
How is the product tested in field or combat conditions?
We have a testing team that goes out into the field almost every day, flies drones, and tests them. We also have partner units with whom we conduct beta testing. They understand the importance of developing such technologies, they are very open, and they don’t get angry if something doesn’t work as it should. They provide very constructive and important feedback.
An interesting thing, and a competitive advantage of Ukrainian companies over foreign ones, is the feedback loop. For example, the fastest feedback from the front line from the moment we finished developing a feature to its first use was 26 minutes. That is, we made the product (functionality), sent it to the pilot, the pilot downloaded it to the drone, flew it, and tested it. This is something that does not exist anywhere else in the world.
I really want everyone to understand the price Ukraine pays for being able to test technologies in this way.
How is your solution for the military financed?
We work directly with the military in terms of support and implementation, but our main model is B2B. That is, we essentially work with drone manufacturers, integrating our systems into their platforms. Our main communication is between our technical department and the technical departments of drone manufacturers.
Accordingly, drone manufacturers pay for our system, and then drones with our system go either through a defense procurement agency, directly to procurement departments, or through volunteer funds, etc.
What is the future of autonomous systems for drones?
The message I am trying to convey, especially to our Western allies, is this: we are developing a monster here in Ukraine, but we have no choice.
In my opinion, the monster is a swarm of autonomous drones. And the side that gets this technology first will have a new nuclear weapon. Moreover, it will be a much more effective weapon than nuclear weapons, leaving no radioactive contamination behind and not destroying the entire infrastructure; it will be possible to simply clear the area in a matter of minutes, leaving the environment almost intact.
And I think this technology will be developed no later than five years from now. And maybe even sooner with the acceleration of progress and the development of LLM (large language model), as LLMs become smarter, help write code faster, and so on.
Another important stage of development, or an important consequence of this development, is that there is currently a shortage of pilots, and we are doing everything we can to enable one pilot to build missions with 10–20 drones. We understand that it is impossible to completely remove humans from the front line of defense, unfortunately. But I see development in minimizing human involvement so that it is more likely to be engineers who control the robot.
About the latest funding
We recently announced the closing of a $1 million round. We are very fortunate because we found not just investors who gave us money, but like-minded partners who believe in us, our mission, and what we do. Their main goal, like ours, is to make an impact on the front lines, with money coming second.
We need these funds to significantly expand our team, which is currently happening. We are a very R&D-intensive company, with a large percentage of our expenses going to R&D.
Furthermore, we have begun more active development of another product, which I cannot talk about yet, but which solves very important problems for our country.
What are your plans for the future?
We see our mission as follows: we solve specific problems on the front lines that exist today in such a way that all our solutions and products form a single ecosystem.
Accordingly, our plans are to conceptually develop a system for autonomous drone control and simple, high-quality mission planning for these drones, as well as to provide feedback during mission execution.
We believe that this is all very important because carrying out a mission without being able to see the result or get some kind of feedback or report is a slow process, one that greatly slows down development. Therefore, the ecosystem will increase the level of autonomy of drones.
Powered by WPeMatico
https://en.ain.ua/2025/11/06/interview-with-norda-dynamics-ceo/