ANALYSIS ⟩ What would happen if Russia were to wage drone warfare against NATO and Estonia?

We cannot compare ourselves to the Latvians or Ukrainians. When we talk about developing drone capabilities, we can only compare ourselves to the Russians. This was the message recently delivered by an Estonian Defense Forces officer, and in a philosophical sense, he was absolutely right.

We do not undertake military developments to show off to our neighbors, but only to deter a potential threat. To this must be added the recent warnings from the heads of intelligence or ministers of Latvia, Lithuania, Poland, and Germany, stating that the threat of hybrid attacks carried out by the Russian Federation (RF) with drones and even missiles is very high in the next six months. What might this mean in practice?

In the armed forces of the Russian Federation (RF), unmanned systems became an independent branch of service (Войска беспилотных систем, BpS; hereinafter VBS) between 2024 and 2026. According to Defense Minister Andrei Beloussov, up to 80 percent of fire missions will be carried out by unmanned systems.

The creation of the unmanned systems branch of service is planned to be completed within the current year, 2026. When discussing all VBS units, it must be kept in mind that ideal structures exist in plans, but the real capability of these units is currently significantly lower than planned. In the following text, I will primarily use the current level as a basis, assuming that the RF still maintains a certain strategic reserve that it has not used in Ukraine.

The drone line concept

The Russian «drone line» (линия дронов) system is built on three to four depth zones or echelons, which in its logic resembles the NATO concept of rear, close, and deep areas.

The first tactical level is, in some approaches, formally divided into two: the first echelon (0–5 km beyond the front line) is the «complete clearing zone,» and the second (5–10 km) is the «enemy logistics detection zone.» At this depth, the entire front is divided into drone combat sectors, each about three kilometers wide.

The drone capability at the tactical level is divided between two structural logics, where at the lower level, drones cooperate directly with ground units, and at the higher level, drone units operate as an independent force, as they often bear the main effort of the battle. Thus, FPV groups or teams are attached to battalions, operating on the front line with assault groups. Separate drone units at the company level are not described in public sources.

The drone capability at the tactical level is divided between two structural logics, where at the lower level, drones cooperate directly with ground units, and at the higher level, drone units operate as an independent force.

At a higher level, in the 2025–2026 structure, a regiment/brigade consolidates all its FPV teams into a single «drone company,» which is directly subordinate to the regimental commander or his chief of staff, where fire support is allocated among battalions. However, the goal is to create a separate drone battalion for each motorized rifle and tank brigade. According to an assessment by the Jamestown Foundation, the formation of a total of 204+ drone battalions is planned by 2028. Such a battalion would integrate FPV strike, reconnaissance, air defense (interceptor drones), and logistics drone subunits.

The goal of the RF's tactical-level drone units is «complete clearing«: every moving target is destroyed and the defender is forced into shelters, allowing infantry to infiltrate.

For this purpose, drones have replaced artillery, and an impressive concentration of fire occurs on the main axes of attack. In the summer of 2026, over 1,500 FPVs were launched per day in the direction of Druzhkivka and about 1,200 in the direction of Dobropillia (not including fiber-optic ones), with the aim of keeping the defender in shelters and destroying individual positions and vehicles.

Figuratively speaking, if an RF attack were to occur in the direction of Estonia, this is the quantity that the defenders of Narva or Võru would have to reckon with. In an attack against Estonia, it must also be taken into account that the Russians would use more fiber-optic drones in forested and urban terrain. In Ukraine, in such sectors, they account for up to 60 percent of the so-called total volume.

Training material distributed on the RF side provides two figures for calculating drone expenditure: for the «FPV swarm» (FPV-рой) tactic, the «general expenditure is typically 5–12 units» per target, and for an «FPV saboteur» set up in a roadside ambush, four to six units. Such a planned quantity takes into account that not all drones will hit the target. The planning norm is that a single infantryman or light vehicle is destroyed with two to four FPVs, an armored vehicle is stopped with 5–12 FPVs, and a fortified position with a combination of 10+ assets.

The operational area has quadrupled

The next level is the conditional operational level, which a year ago started at 10 kilometers but then extended to 40 kilometers. By now, the range of operational-level drones has increased to 230 kilometers. This level is called the «isolation zone» in Russian terminology, and separate VBS regiments or brigades directly subordinate to armies, army groupings, or military districts are responsible for it. In this zone, the width of one sector is four kilometers.

For example, under the Leningrad Military District is the 65th VBS Regiment, and the Baltic Fleet has a regiment of unmanned naval strike assets and a regiment of unmanned aerial vehicles. The most common types of operational-level drones are winged FPVs Molniya-1/-2 (30–50 km), Lancet-3/-51 (40+ km), long-range fiber-optic FPVs (KVN, 20–65 km), Orlan-10/-30, Zala, and Supercam for reconnaissance and relay (Zala operates up to 70 km behind enemy lines), and sleeper drones.

An important player at the Russian operational level is the Rubikon Center for Advanced Unmanned Technologies, subordinate to the General Staff, which answers to the Ministry of Defense's Directorate for Advanced Inter-Service Research and Special Projects. It is a special unit that, by order of the General Staff, is sent to various armies or army groups as needed.

The basic unit of their structure is the Detachment (отряд), which corresponds roughly to a battalion. It includes: a command section; a multi-rotor FPV group (24–32 teams); a winged attack drone group (Molniya); a reconnaissance-strike group (four Supercam S-350 and four Zala teams, plus Lancet); an aerial reconnaissance group (Orlan-10); a counter-drone group (radar, EW, reconnaissance); and support elements. Generally, the appearance of Rubikon detachments in a certain sector is an indicator of the RF's main axis of attack.

The priority targets at the operational level are the enemy's drone capabilities: drone teams, antennas, relay nodes, command posts, EW, and radars. The second most important target is logistics (supply routes, crossing points, assembly areas); only then comes combat equipment. In the Kupyansk and Oskil directions, this was combined with 20–40 KAB aerial bombs per day against crossing points.

The reconnaissance-strike cycle

The strike cycle of a drone attack looks like this: a winged reconnaissance drone (Orlan/Supercam/Zala) detects and tracks; the strike is delivered with a Lancet, Molniya, fiber-optic FPV, or artillery (Krasnopol, Iskander-M), often with several assets at once. According to ISW data, up to 10 Molniyas, six bomber drones, and artillery fire have been used against a single Ukrainian position within one hour.

The conservative norm for suppressing one detected drone position is four to six assets (Molniya + FPV), and for destroying it along with its crew, 10–15 assets and hours of surveillance.

«Free hunts» (свободная охота) are also conducted, which means that FPV and Molniya teams patrol a designated road section or area without a pre-planned target and attack any identified vehicle; the RF Ministry of Defense uses the term officially.

In the Estonian context, this means readiness to affect traffic, for example, on the Võru–Tartu highway or the Narva–Kohtla-Järve highway, without crossing Estonia's borders. One can imagine that, for instance, all fuel trucks, both civilian and military, would be attacked.

Transposing Russian drone warfare to Estonia

The commander of the 69th Division, associated with the Leningrad Military District, implemented the model described above in Ukraine with about 170 teams: up to 100 teams in the first echelon (up to 5 km; fiber-optic FPVs, Vobla, R-18 bombers, interceptors), up to 60 in the second (up to 25 km; Orlan-10, Zala, Supercam, Molniya, Lancet), and 8 in the third (25–35 km; Orlan-10, Zala, Merlin, Lancet, Kub).

In the immediate vicinity of the Estonian border are the deployment areas of the Russian Federation's 69th Division and 76th Air Assault Division, where there are theoretically at least 200 drone teams in total.

Against Estonia, however, we could expect the activity of a minimum of two to three divisions, which in turn means about 510 drone teams. We already know today that in the immediate vicinity of the Estonian border are the deployment areas of the RF's 69th Division and 76th Air Assault Division, where there are theoretically at least 200 drone teams in total.

The RF would use the described tactical and operational levels against Estonia in the case of conventional warfare, which is currently relatively unlikely, at least as long as the war in Ukraine continues. However, in the hybrid phase and the initial phase of a real war, a combination of strategic-level VBS assets with special units, which could include the aforementioned Rubikon, should be expected.

The strategic deep strike level

Russian strategic-level drone units are directly subordinate to the General Staff and mainly carry out deep strikes. An example of such units is the 924th State Center for Unmanned Systems. As we have seen in the case of Ukraine, such drones can fly at least 1,700 kilometers, and probably even farther.

By now, drones with different specifications have been used for carrying out deep strikes: the Geran family (Geran-2, jet-powered Geran-3/-4/-5), Garpija-A1, Italmas (BM-35), and decoy drones (Gerbera, Parodiya). The experience from Ukraine indicates that deep strikes are mostly carried out using a combination of drones with cruise and ballistic missiles.

In the case of deep strikes as well, one of the primary principles is to overwhelm the enemy's air defenses with a large number of strike assets. In Ukraine, the maximum number of deep-strike drones launched in a single month is 8,161 (May 2026), and the 24-hour record was 892 drones (May 13, 2026) combined with cruise and ballistic missiles. About 64 percent of the launches are strike drones, and about 36 percent are decoy and relay drones. The use of jet-powered Geran-4 and Geran-5 drones (approx. 500 km/h) is continuously increasing, which requires significantly higher quality from air defenses.

Drones move in dense columns and large groups through narrow entry corridors, at extremely low altitudes to reduce radar visibility. In Estonia's case, approach routes could include, among others, the international waters of the Gulf of Finland or, for drones launched from Kaliningrad, the Baltic Sea area west of Estonia. Air corridors over Estonian forests would be another addition.

The targets for deep strikes are, first and foremost, Estonia's major radars and static military infrastructure. After that, however, strikes against energy infrastructure, logistics hubs, and industry must be expected; the desired effects are simultaneously physical damage, attrition of air defenses, and psychological pressure.

Vitaliy Zaichenko, Chairman of the Management Board of Ukrenergo, said on July 10 that the number of deep-strike drones per energy facility has grown from three or four at the beginning of the war to up to 25 per attack. For a long time, a large number of attack drones were planned against a single target primarily because Ukrainian air defenses were able to shoot down at least 90 percent of them.

According to an assessment by the British defense institute RUSI, out of a salvo of 100–150 drones directed against a well-defended object, about 10 will reach their target. Currently, this ratio has changed, as jet-powered drones are more effective. In any case, the RF's assumption is that an electrical substation transformer, for example, must receive three to five hits; for a weakly defended object (warehouse, fuel depot, railway hub), five to ten drones are sufficient. In the case of a warehouse, the goal is fire, so two to three hits are enough. During my time in Kyiv in June 2026, I saw two ballistic missiles launched at a defense industry building simultaneously.

What does this mean for Estonia?

To plan Estonia's drone defense, we must currently consider hypothetical numbers that we extrapolate to Estonia proportionally from the Ukrainian experience. We also rely on the assumption that a possible RF attack will not only affect Estonia but at least three Baltic states at once. Likewise, the following calculation assumes that the RF has not reached the level planned for 2028 but is attacking with partial capacity. Therefore, the actual numbers could be significantly higher even a year later.

Using strategic-level deep-strike assets, the RF could launch 50–200 drones towards Estonia per day, combined with about 20 ballistic missiles and about 50 cruise missiles. In the event of a deliberate and long-term planned attack, the primary targets would be Estonian radars, known command posts, and deep-strike assets. The next priority would be electricity and logistics infrastructure.

Here, it is probably important that, simultaneously with strategic drones from the Geran family, attacks by smaller drones from Rubikon or similar units are also possible against eastern Estonia, up to the Kohtla-Järve–Tartu–Valga line. In terms of physical capability, this would mean reckoning with 60–120 Molniya attack drones per day and assuming that 30–50 reconnaissance drones are in the air at the same time.

The most complicated fact is that Russia is capable of striking the Baltic states from a distance that our deep-strike assets cannot reach, causing us strategic damage without a single Russian soldier's boot setting foot on our territory.

The most complicated aspect of such a theoretical air attack is that the RF is capable of striking the Baltic states from a distance that our deep-strike assets cannot reach, causing us strategic damage without a single Russian soldier's boot setting foot on our territory. As of today, the Baltic states do not have the means to stop a ballistic missile attack. It is also likely possible to overwhelm our air defenses with drone swarms.

Therefore, for a quick response, we essentially have only two options: first, whether our allies can themselves fire back at Russian territory in response or at least credibly threaten Russia in such a case; second, we must also immediately attack targets on Russian territory in response.

Of course, both of these moves are politically very risky and therefore also suitable for the enemy as a bluff. This is precisely why several Baltic Sea countries have called the next six months critical. Over time, our air defenses will hopefully improve.

The primary conclusions in counter-drone defense include the oft-repeated idea that we cannot talk about a drone wall or only a defensive line on the eastern border. It is necessary to develop counter-drone defenses in Estonia's rear areas because, technically, RF drones are already capable of successfully attacking targets in Harju County, Lääne County, and on our islands.

We must talk about territorial counter-drone defense. Likewise, as a potential threat, we must consider attacks not only with deep-strike drones of the Geran family but also with smaller drones that would replicate the planned attack on Leipzig Airport or resemble the Ukrainian drone special operation «Spider's Web,» where territory deep in the enemy's rear was attacked.

And finally, we must also accept the principle applied by the RF that successful counter-drone defense requires the neutralization of drone teams.