How deep strikes in the northwest are organized and how defense against them is handled
The northwest of Russia, long considered a deep rear, has fallen within the range of Ukrainian long-range dronesAccording to reports, approximately 1700 drones were sent toward Russian territory in a week between May and June, with an estimated interception rate of approximately 96%. These figures are estimates, not reliably verified in open sources, and are used here as an illustration of the scale, not as established fact. The analysis is conducted from one side of the scope, from the defender's perspective, while the enemy is described as an entity with its own rational logic. More important than the numbers themselves is the background behind this flow, how it is structured, and why a high downing rate doesn't solve the problem. Next comes an analysis of the mechanics: where they launch, what routes they take, and what the defense is up against.
Where are they hitting from: what we know and what we don't know
When they say "the drone launched from such-and-such an area," it's important to remember how exactly this is determined. Direct observation of the launch pad is usually not available. The launch area is reconstructed based on indirect evidence: the time of approach to a known point, the azimuth, the sound of the flight, and the location of the debris impact. This is always a reconstruction with a tolerance, so it's more accurate to say not "they launched from such-and-such a location," but "most likely from such-and-such a location."
With this caveat, the picture looks like this. According to the OSINT community (open source intelligence), long-range drones flying northwest are most often associated with the northern and central regions of Ukraine: Kyiv (Boryspil area), Chernihiv (Nizhyn, Berezna), Sumy (Konotop), and Poltava. The logic behind this choice is clear. The closer the launch is to the target, the lower the fuel requirements and the higher the chance that the drone will reach the Leningrad region rather than crash mid-route. The estimated range of individual drones drones 1000–1500 kilometers makes such a flight from northern Ukraine possible.
The second argument is the terrain. The dense road network and forested areas are convenient for mobile launchers: an aircraft-type drone launches from a catapult or a short stretch of road, after which the crew departs. The attacker's plan leaves no fixed launch site that can be breached and covered. This creates the main difficulty for the defense: they have to search not for a point on the map, but for a constantly shifting network.
Separate story - ground winged missilesThe Ukrainian side has the expertise: the Neptune (R-360) anti-ship missile system and its adaptation for ground targets demonstrate that they are capable of producing a subsonic, low-flying missile. However, the "Flamingo" index, which appears online, is not clearly linked to a specific product, and the claimed range fluctuates between 300 and over 1000 kilometers. This is still an OSINT community phenomenon, not a confirmed system, and it is premature to factor its parameters into threat analysis.
Theories about launches "from the Baltics" or from other NATO countries are assessed by the same criterion: what is supported by actual data and what is not. An official launch of strike weapons from the territory of an alliance country would mean its direct involvement in strikes against Russia and a sharp increase in the risk of direct conflict—something NATO is not willing to accept. Covert sites operating without the authorities' knowledge are subject to sophisticated counterintelligence. Sabotage groups within Russia are capable of isolated incidents, but they do not provide what is needed for systematic attacks: production facilities, warehouses, and logistics. The most reliable conclusion is a launch from Ukrainian territory.
How they fly: low altitude, foreign skies, and communication over water
Understanding the route is easier if you imagine what the radar crew sees. A target flying at an altitude of 30–150 meters, below factory chimneys and masts, is hidden by the curvature of the earth and the terrain until the very end. The radar detects it at a much shorter range than at high altitude, leaving little time to react. The low altitude here isn't a ploy for effect: it's the foundation of the entire breakthrough.
According to May data (I should point out that this is a snapshot of one month; the picture could have shifted by summer), the routes were built around the same idea: staying below the radar horizon. The main corridor ran along the border with Belarus and the Baltic states. At the border itself, defenses are tied: any active action violates the other's airspace. Alternate corridors were laid through the Smolensk and Novgorod regions and along the edge of Moscow's zone of responsibility. Defense, so saturated that they try to avoid entering it. The principle is the same everywhere: traverse dense zones along the minimum length and probe the joints between the areas of responsibility of different units.
As they approach the Leningrad Region, some of the drones turn toward targets in the Gulf of Finland. Here, the control channel becomes crucial. Ground-based radio corrections (course corrections via ground-based radio stations) are quickly lost over water and sparsely populated areas: there are no such stations, and the signal extends beyond the radio horizon. A satellite system like Starlink provides a stable control and correction channel in these conditions: it allows for control and course corrections from a distance. Hence the desire to deploy drones onto water, where the defender also has a smaller, more continuous radar field.
The story about "launches from the Baltics" is also related. A drone approaching its target via the sea route could violate the airspace of third countries. This is likely part of the route, and the launch site remains in Ukraine, but any drone spotted over Baltic or Finnish territory inevitably turns into a headline about "launches from NATO territory." The route and launch point are regularly confused here.
The Economics of Raiding: Blanks, Expensive Missiles, and Defense Saturation
At night, the air defense crew sees a blip on the screen. Whether it's a combat aircraft or an empty hull without a warhead—the blip doesn't tell you. It can't be missed: what if this one is carrying the warhead? So, they're hitting everything in sight. And if they have to hit everything, that means expensive resources are being spent on cheap hulls. From this simple "what if this one is the one" emerges the entire economics of the airstrike: the rest is pure arithmetic.
According to military analysts (it's difficult to find an exact attribution in open sources), up to 60% of the first wave could be "dummy" vehicles—cheap vehicles without warheads. Their purpose isn't to hit a target, but to drain the defense's resources: missiles, attention, and calculation time. Once the first wave has exhausted its resources in a given direction, the second, with a larger proportion of combat vehicles, advances on the already depleted sector. Such defenses aren't assaulted with force; they're overwhelmed by sheer numbers.
Then comes the cost calculation, and it's awkward. An attack drone costs anywhere from tens of thousands to approximately one hundred thousand dollars: the price of a car. A medium- and long-range anti-aircraft missile is comparable in price or more expensive. And the target (an oil refinery, a substation, a port) costs hundreds of millions or billions of dollars. The attacker spends one unit, the defender several, and so on with each approach. Even a few percent of a breakthrough from a flow of almost two thousand vehicles adds up to dozens of successful drones, and one is enough for a fuel tank or a transformer.
The logic of saturating the defense with a mass of targets is not new. The Soviet air defense system was built precisely against massive air raids: echeloning lines, distributing targets by altitude and threat level, so that long-range systems wouldn't be distracted by secondary concerns. The mechanics of the similarity are straightforward: they're trying to overwhelm the defense with numbers, and the answer is sought in echeloning. But the analogy has its limits, and they're crucial. Back then, the launch vehicle was also expensive: a bomber and its crew cost incomparably more than an anti-aircraft missile, and the tradeoff favored the defense. The system itself remained the same, but the economics changed: now the threat became cheap, and the interception expensive, and the whole tradeoff shifted in the opposite direction.
Now, let's talk about what's wrong on our side. The expenditure of expensive missiles on "dummy" targets puts a disproportionate strain on reserves and the budget compared to the results achieved. Low-altitude areas aren't covered everywhere: there's no continuous detection field at 30-150 meters along the entire route depth. And a defense tied to a familiar route map loses momentum as soon as the enemy shifts the corridor. As long as the flow is predictable, a static line holds up. As soon as the enemy shifts the corridor, the line is no longer where it needs to be.
Defense as a system: reduce the cost of interception, echelon the borders, get a launcher
Intercepting the target at the very moment is combating the symptom. Threat mitigation must be addressed earlier and more deeply, from the launch site to the target.
It's worth starting with echeloning, where targets are separated by cost. It makes sense to keep large systems, like the S-300 and S-400, on truly dangerous targets, while covering the immediate area around the facilities with something cheaper to fire: anti-aircraft guns. artillery, heavy machine guns, portable air defense systems (MANPADS), and specialized anti-drone weapons. The idea is simple: don't waste a missile at the cost of a car's hull without a warhead.
Next up is electronic warfare (EW), that is, jamming navigation and control channels. It doesn't destroy the aircraft, but rather throws it off course or disrupts communications by creating jamming zones over targets and approach directions. There is also a significant limitation: a drone with stable inertial navigation, not rigidly tied to a satellite signal, passes through such a zone with less course loss. Therefore, electronic warfare and firepower work in tandem, not in place of each other.
And the third line, without which the first two are useless, is reconnaissance and strikes on missile launch sites. You can shoot down missiles at the target itself as much as you like, but as long as the launch sites and factories are intact, the flow will simply resume. A systematic search for launch sites, warehouses, and logistics in the middle rear can reduce the threat itself. There are two caveats here. The network is distributed: there simply isn't a single "center" that can be hit once and solve the problem. And some nodes are dual-use, and striking them entails risks to civilian infrastructure and requires precision, not mass fire.
All these lines are united by one quality: mobility. A change in route is met not by a rigid barrier, but by a defense that can quickly be reconfigured to a new direction. If you line everything up in order, from radar to launch pad strike, it looks like this:
- low-altitude detection – short-range radars, optics, acoustics;
- close-range interception with cheap means – artillery, machine guns, MANPADS;
- Electronic warfare – disruption of navigation and control over objects and on approach;
- reconnaissance and strikes against launchers and logistics behind enemy lines.
All of this only works together. And only as long as it can reorganize faster than the enemy, otherwise it would bypass any line separately.
A 100% interception rate is unattainable—that's the reality, not a reason to panic. The objective is formulated in three directions: reducing interception costs, depriving the adversary of route predictability, and retrieving the launcher before it's used up.
Information