Early UAV Detection: How Target Designation Works

Zero seconds to react: why detection decides

Every conversation about protecting a site from drones starts not with effectors but with time. Without a detection system the guard force has zero seconds to react: the attacking aircraft is noticed when it is already over the target — and more often, people learn about it from the explosion. No countermeasure works if the target is discovered at the moment of impact.

An early UAV detection system changes that equation. It pushes the detection line out to the approaches and turns a surprise attack into a managed incident: the target is found, classified, the crew receives designation and readies the interceptor. The clock still counts in minutes — but now the guard force actually has those minutes.

The threat sets the context: modern attack UAVs reach ranges of up to 2,500 km, fuel-and-energy, power and industrial facilities deep inside the country are exposed, and damage from a single hit is estimated at 50–500 million rubles. At those stakes, drone detection is not an option but the first, mandatory layer of defense. Our product is described on the early detection system page.

360-degree thermal imagers: the eyes of the system

The system's primary sensor is a cooled 360-degree thermal imager on the autonomous observation posts. Choosing thermal imaging is fundamental for thermal drone detection in real conditions:

• Round-the-clock coverage. A target's thermal contrast is as visible at night as in daylight — and attacks most often come in the dark.
• All-weather performance. Haze, dust and cluttered backgrounds degrade optics far more than the thermal channel.
• Passivity. A thermal imager emits nothing: it cannot be direction-found and bypassed by its signature the way active sensors can.
• Full 360-degree view. A post watches every bearing at once — the attacker gets no blind sector to sneak through.

The posts come in an arctic-grade build and run on autonomous power, so they can be pushed out to lines where there is no infrastructure. An important architectural detail: the posts only collect and transmit video. All the analytics live at the control point — this keeps the posts simple, makes replicating them along the perimeter cheaper, and lets the brains of the system be upgraded in one place.

AI classification: copter, fixed-wing, bird, aircraft

Video from all posts flows to the AI server at the control point. Its job is to answer the operator's key question: what exactly is flying? The server classifies the target into four classes:

1. Copter — multirotor aircraft, including FPV;
2. Fixed-wing UAV — long-range winged aircraft;
3. Bird — the main source of false alarms for any electro-optical system;
4. Manned aircraft — airplanes and helicopters that must never be engaged.

Two numbers define the classifier's practical value. First — time to decision: 25–30 seconds from the target entering the field of view. Second — the false-alarm target: no more than 5%. Balancing them is the core engineering problem: a system that yells drone at every seagull quickly trains the crew to ignore alerts, while a system that thinks in whole minutes burns the very time reserve it was installed to create.

Classification is the filter that turns a stream of frames into a decision: copter on such-and-such bearing — engage, or bird — stand down. That filter saves the defense's scarcest resource: the crew's attention and time.

Target designation: from frame to launch

Target designation is the output of the whole chain: confirmed information about the target that the crew uses to aim the intercept asset. The path from frame to launch looks like this:

1. A thermal post picks up an aerial target and streams video to the control point;
2. The AI server classifies the target within 25–30 seconds;
3. The operator confirms the threat — a human stays in the decision loop;
4. The crew receives designation: bearing, target class, motion parameters;
5. The operator aims the interceptor and launches.

The key principle: the system does not shoot by itself. It takes over everything routine — continuous horizon watch, recognition, filtering out false targets — but the decision to engage stays with the operator. This division of labor allows a minimal crew that enters service quickly: nobody has to stare at a screen around the clock anymore; the job is to respond to confirmed threats.

Paired with the interceptor: the full loop

Detection alone does not stop an attack — it creates the conditions for an intercept. That is why the early detection system was designed as one loop with the VOLKODAV 2 interceptor drone: it hands designation to the crew, the operator aims and launches the interceptor, and from there the aircraft's onboard AI tracks the target through the thermal camera feed and rams it — no GPS and no explosives. The interceptor itself is covered in the announcement VOLKODAV 2: Drone-vs-Drone Interceptor with Onboard Homing.

This pairing also closes the main weakness of small arms: a shotgun works at tens of meters and only against a target already over the site, whereas the detection-plus-interceptor loop meets the target on the approaches — with forward-deployed crews the intercept perimeter of the complex reaches 10–12 km around the site. Why small arms are unreliable against an aerial target is examined in Why a Shotgun Won't Stop an Attack Drone, and the kinetics-versus-jamming comparison lives in Kinetic Interception vs Jamming (EW).

Complex architecture: the 16+ km reference

For large facilities the detection system scales into a full complex. The reference configuration for a 16+ km perimeter includes:

The configuration for a specific site is derived from an audit: perimeter geometry, threat axes, buildings and terrain determine the number of posts and crew positions. For smaller sites the logic inverts: a package of the detection system plus interceptors is deployed for a single guard unit — following the principle of one unit, one intercept kit.

Conclusion

Early detection is the foundation of any drone defense: it converts the surprise of an attack into a time reserve. The 360-degree thermal imagers deliver a passive, round-the-clock picture on every bearing; AI classification separates a real threat from birds and manned aircraft in 25–30 seconds with a false-alarm target of no more than 5%; and the crew receives ready-made designation to launch the interceptor.

Finding out whether the system fits your facility takes one step — leave a request: we will discuss your site's geometry, the threats and the package composition. The technical product descriptions are on the detection system and interceptor pages.