The face of the Ukrainian conflict continues to change, and many analysts and commentators alike are stuck interpreting the battlefield with an outdated model. Others stick to deficient generalities, based on vague cliches about drones, ignoring the subtle nuances at play on the frontline. Let us take a look and examine where the real war stands today, with a focus on answering the ultimate question on everyone’s mind: can Russia still “decisively” win this war which, in the eyes of many, is trending toward an entropic drone stalemate in perpetuity?
i.
To answer that question, we must examine the current on-the-ground realities, rather than rehashing out-of-date tactics from last year or the year before. One example of popularly repeated-but-dated information, is that artillery still accounts for ~90% of casualties, or thereabouts. While neither is it true that drones inflict 90% of casualties as some pro-Ukrainians swear, it is likewise no longer true that artillery dominates to such an extent as it did even a year ago, let alone further back. It’s difficult to determine the exact percentage, but at this point it would not be unreasonable to suggest that somewhere between 40-60% of deaths are drone related.
This is inferred by a variety of methods:
- Direct quotes from frontline units. For a long time only the Ukrainian side contended that drones were their primary means of fire damage, but this was understood because they lacked in other weapons systems compared to Russia. However, now even many Russian units are reporting that drones are outweighing other systems on their section of the front.
- Direct video evidence. We see less and less footage even from the Russian side of artillery destruction, and disproportionately more drone strikes. This is particularly the case with the advent of fiber optic or ‘optical fiber’ drones, which increase the success rate of hits exponentially.
- The sheer scale of drone production on both sides has grown far beyond any other weapon systems. For example, while Russian artillery and glide-bomb production may increase by 20-30% per year, drone production is seeing parabolic increases of hundreds or even thousands of percentage points year-on-year.
An example, various sources claim Ukraine produced 20,000 FPVs per month in early 2024, and now produces over 200,000 per month in 2025, an increase of ~1000%.
Ukrainian producers were delivering about 20,000 of the dish-plate-sized quadcopters a month at the start of 2024, but increased investment and better-organized supply chains and manufacturing processes spiked output to 200,000 aircraft a month in January 2025, Havryliuk said.
Russia is said to be seeing similar figures. The scale is so staggering, that most people are not capable of comprehending it, and are stuck in obsolete paradigms of the war.
I’ve posted the photo before, depicting a Russian EW unit’s trophies of disabled AFU drones just on one small portion of the front:
Today a photo emerged claiming to show the web of fiber optics cables stringing over the battlefield—reportedly after a morning frost, which made the thin wires more visible:
Granted, the sheer scale of drone production, usage, and success may be vastly overstated. For instance, even if both sides produce 100-300k drones a month as claimed, they also both admit that the vast majority of drone strikes are unsuccessful, with the systems either brought down by EW or simply missing the target.
Let’s say 300k+ drones are produced per month, as now claimed, by the Russian side, with only 10-30% of them succeeding in some way, even if it’s a glancing blow which does not disable the target. That’s about 30-90k hits per month. Artillery is fired at a rate of 10-20k shells a day, or 300-600k shots per month, by the Russian side. If we assume a similar 10-30% do some damage to a target, we can educe that anywhere between ~30k to ~150k artillery hits are being registered per month, which does not include various other systems like aerial bombs, etc.
Judging by those numbers, it’s easy to see that drones could plausibly account for at least 20-30% of scored hits, if not much more, given their higher accuracy. It may be better to break it down by type of target: tube and rocket artillery and aerial bombs likely account for the vast majority of damage to infrastructural targets like weapons depots, ammo dumps, fortifications, workshops, stationary equipment, etc., while drones may account for a proportionally high amount of infantry kills—like I said, not necessarily the majority, but perhaps 35-65%. A huge portion of videos we now see feature not only FPV kills of infantry, but large hexacopter and agricultural ‘agro-drones’ dropping bombs on dugouts, etc. Drones have also weakened opposing artillery due to their increasing range, which now regularly allows them to roam 15-20km behind enemy lines—and even much farther in the extreme examples—which is precisely where most artillery systems work. This forces artillery systems to retreat out of range and be ineffective, with only the minority portion of systems with superior range able to consistently function along some fronts.
And there are new kinds of drones appearing all the time—to show one example on the Russian side, the Molniya-2, a kind of hybrid cross-OWA-FPV:
⚡️ Crews of the Molniya-2 strike drones of the Center group destroyed a fortified firing point of the Ukrainian Armed Forces in the Krasnoarmeysk direction.
Not only has production of fiber optic drones skyrocketed on both sides, but “machine vision” AI-powered drones have been increasingly as well. Here’s one recent example of a Ukrainian one, which appears to have missed, but only by an inch:
Footage has emerged of the use of the new Ukrainian attack drone UAS SETH, which uses AI machine targeting. In appearance, the drone resembles a smaller copy of the Geranium, but has an optical guidance system with automatic acquisition and target acquisition and is designed to destroy objects in the frontline zone. Looks like it needs work, because it missed.
And another more effective one, called the Shrike 10CV:
Ukrainian specialists, on the other hand, increasingly find all kinds of new Russian drones with this AI ‘machine vision’ on the front:
Recently CEO of Anduril Palmer Luckey bragged about how his company’s Altius-700M drone, which according to him features fully autonomous hunter-killer mode, has already been widely used in Ukraine.
These days the most successful drones are modular, and can be adapted to a variety of EW conditions and general tasks. Russian forces have been rolling out a highly modular copy, which can change cameras depending on needs, and most importantly, the antenna itself, allowing it to operate on different frequency bands to out-maneuver whatever EW frequencies the Ukrainians are favoring in that particular section of front:
Note they mention the drone also has AI machine vision in the final section of flight, whereby it can autonomously hold the target in case it is jammed. In fact, the developer mentions they are working on enhancing the AI capabilities even further, integrating a topographical autonomy that will allow the drone to hunt its own targets in an unknown environment, presumably after understanding the surroundings via a kind of terrain mapping (TERCOM) integration.
UGV, or ground bots, have also been increasing in vast numbers on both sides, mostly DIY or cheaply made variants.
Ukraine has virtually stopped using naval drones in a ‘kamikaze’ role, instead deploying them as mothership carriers for FPVs which attack Russian coastal targets around Crimea and the Kinburn Peninsula. In a newly released video, one can even see how Russian Pantsir-S1 missile systems are unable to hit the tiny, maneuverable drones:
It should be noted it is a good sign they fired at it, which means the Pantsir radar is at least picking the small cross-section craft up, but the missiles simply were never designed to hit such tiny, jittery targets. A new class of Pantsir-SMD mini-missiles made specifically for small drones is still being developed and rolled out.
One of the last refuges against drones now being extensively utilized by both sides happens to be a rather primitive solution: the creation of anti-drone net corridors to secure the entire length of important supply routes. Russians have now systematized the installation of these on various fronts, with engineering troops specially outfitted for the task, as can be seen below:
And Ukraine is doing the same—here’s a spliced video of two new Ukrainian supply routes:
It brings up the question of why the world’s most powerful and advanced countries cannot come up with a solution that effectively neutralizes these drones. EW (Electronic Warfare) was meant to be the silver bullet, with Russia as world leader in this intricate art—but it turns out, fiber optic and autonomous AI drones completely negate the jamming side of EW.
There is DEW (Directed Energy Weapons) like microwave emitters which can easily fry the electronic motherboards on a whole swarm of small drones at a time. The problem is, these systems are hugely expensive because they require massive amounts of energy directed in a tiny cone, which would never be able to stop swarms coming in from every side—and above—as is now common practice on the front.
One of Ukraine’s top radio-electronics experts Serhiy “Flash” Beskrestnov recently ridiculed the very idea after someone posted the following video:
His comments:
Another activity of microwave adherents. The video is gaining views, many consider it their duty to forward it to me: “This is a miracle solution against drones on optics.”
We tried one magnetron and three. Already at 3-5 meters the drone does not care about this radiation.
Yes! Such technology of work against drones is possible. To do this, you need to collect a lot of power in a narrow beam. So much so that at 20-30 meters on the semiconductors and boards such a strong induced RF voltage is created that the elements will fail.
And such power requires energy.
I already told you that the only weapon of the Russian Federation “Ranetz” of this type travels on a missile tractor!!!!!! and is charged for one shot for 20 minutes!!!!!! And all this to shoot down a drone at a hundred meters.
Summary: he says his team already tested such solutions, and past 3-5 meters, the microwave beam no longer affects the drone. To have a stronger beam, like Russia’s experimental ‘Ranetz’ system (pictured below), he explains you need massive generators of power, claiming the Ranetz charges for 20 minutes just to take a shot:
In short: such solutions could work for protecting highly sensitive sites, valuable HQs, or stationary C2 nodes, etc., but absolutely cannot be mass produced and distributed at scale along an entire 2000km frontline aswarm with drones.
Similarly, the ‘net tunnel’ solution seen earlier is great for protecting the rear, where your engineering corps can spend time constructing such apparatuses in relative safety. But along the contact line, where assault troops must advance through ‘no man’s land’, such contrivances cannot be put up in time.
One major way to mitigate FVP dominance that both sides have been utilizing more and more is the countering of enemy midsize surveillance UAVs by hunter-seeker FPVs. The FPVs themselves cannot be stopped, but they are not effective without an intermediary “spotter” drone which first identifies targets for them. The reason FPVs themselves do not generally act in ‘free range’ hunting mode, is because they have very limited battery life, since their construction is a very finely-balanced tradeoff between size, payload, speed (which is very important), and range. To maximize payload and speed, you necessarily must reduce flight time, otherwise you must carry much larger batteries that reduce the other important attributes.
By taking out the longer endurance ‘spotter’ drones—which typically are larger, fly slower and higher, and are easier to detect—you can ‘blind’ the FPVs working in tandem with them, or at least greatly complexify their job. This method is effective in mitigating the damage, but has enormous room for improvement, given that it’s a fairly nascent combat role, relatively speaking. In particular, the detection of enemy surveillance drones is where the art lies, requiring skilled technicians manning “listening posts” full of complex spectrum analyzers and mid-range radars.
Besides that, the other solution both sides have adopted is of course the atomization and distribution of troops now so notorious on the front: small groups of units, often eschewing their sluggish heavy armor in exchange for fast mobile transports like scooters, ATVs and four-wheelers, motorcycles, etc. One recent interview with an AFU officer said that on his frontline, he can attest to the fact that “Russians on 4 ATVs are more effective at advancing and securing positions than on 8 BMDs.” Note that a BMD is a powerful IFV (Infantry Fighting Vehicle), and as such the statement is a true testament to how vast the changes to the battlefield have become. Lines of heavy armor are now often—or even usually—sitting ducks for FPVs of all kinds, but a lightning motorbike sprint can wedge a fireteam into a hedgerow landing.
ATVs on the Front Lines – Indispensable Helpers in the Combat Zone
This equipment provides soldiers with everything they need: from food and ammunition to emergency evacuation of the wounded.
Maneuverable and fast, ATVs allow for the rapid delivery of cargo to the front lines, the transfer of troops and the execution of urgent tasks in the most difficult conditions.
ii.
Now that we’ve set the stage, let us discuss where the war specifically can go from here, from the point of view of the Russian side which still seeks to prosecute successful advances in order to capture the remainder of Russian territory.
The biggest focal point in such discussions has revolved around the attrition to Russian armor, vehicles, and weapon systems in general. Pro-Ukrainian analysts have made a cottage industry out of studying Russian tank and APC storage depletion via satellite photos, as well as collating daily hits into vast spreadsheets.
The current numbers from their top experts look like this:
The charts represent the time period of August 2023 to February 2025. On the left chart you’ll note that the total number of strikes on Russian vehicles appears to have vastly increased. However, the total strikes proportionally done on strictly armor like tanks and IFVs has actually decreased, while civilian vehicles, bikes, and ATVs have skyrocketed.
There are two ways to interpret this:
- Russia is running out of tanks and APCs and has switched to using civilian vehicles, donkeys, horses, etc. This is the preferred explanation for pro-Ukrainian experts for obvious reasons.
- Russia massively started to roll out civilian vehicle usage as a deliberate change in tactics in response to the exponential growth of drones.
Certainly, the true answer lies in a mixture of both—however, one major aspect propagandists leave out is that the Russian army was for a long time contending with internal problems related to civilian vehicle usage. Essentially, upper command greatly hampered their use, restricting units from using “donated” vehicles which were not “officially” registered, provided, and approved.
Many viewed this as a result of the Russian MoD’s embarrassment: not wanting to be upstaged by civilian crowd-funders and organizers. But last year, most of these issues were finally resolved and it resulted in the floodgates of civilian vehicles being opened and sanctioned by command. This is one of the chief explanations for the sudden influx of civilian vehicles seen in the charts above. Obviously, however, the other explanations still hold merit: that Russia continued to refine its tactics, particularly the ‘atomized assault’ or “minoring the major” as some analysts have dubbed it, wherein small unit tactics replaces large-scale ‘maneuver’ warfare. Civilian vehicles, in essence, became battlefield ‘consumables’, and the vast amount of them are destroyed unoccupied, i.e. where they are parked or sitting camouflaged somewhere.
That’s not to say that Russia isn’t losing lots of valuable equipment, but recent figures have pointed to a slow down in tanks destroyed to a level that could approach sustainability:
Above shows recent months averaging about 60 tanks per month, which is about a 2 per day average, or ~720 a year. The worry is that Russia only produces 200-300 brand new tanks per year, while the rest are refurbs, so these are the two numbers that need to align eventually, lest Russian refurb stocks run out.
Part of the long term strategy I had predicted long ago, was that Russia would eventually wind down heavy armor usage, intermixing it with civilian vehicles, specifically to the point where those numbers even out, so that Russian production of new tanks breaks even with losses. Even so, now that tank losses have slowed down as can be seen above, refurb stocks would have several years to go before depleting, which means Russia has several years to potentially setup new production lines to build brand new tanks.
To highlight an important point again: most Ukrainian data analysts assume all Russian tanks taken from storage are sent to replace losses, and as such equate 1:1 with losses. In reality, Russia has built several new entire armies for the new military districts, which had to be staffed and outfitted, and much of the gear can logically be expected to be sent there.
For instance, here’s German Inspector General of the Bundeswehr Carsten Breuer claiming Russia built the new Leningrad Military District specifically to “threaten the West”:
He says Russia has doubled the size of its army, its structures, created the Leningrad and Moscow military districts, which he claims are so big and powerful that their threatening nature is readily visible to the Germans. If even a fraction of that is as true as he says, it means Russia must have heavily staffed these new districts, each having their own combined arms armies.
But does the new drone threat mean the modern battle tank and associated systems are ‘obsolete’? No, they still all have their part to play, as long as a generalized infrastructure exists within the military that is able to minimize and blunt the danger of drones as much as possible. A successful future military is one that moves away from outdated extremes, chasing perfect “silver bullet” systems to take out “all drones”. Instead, the winning philosophy will be one of curbing the drone threat as much as possible, while accepting that some percentage of attrition is now built into the calculus of waging successful military campaigns.
And even curbing the threat cannot come from one or two vectors, but from the totality of integrated systems, minimizing the damage at every possible level of the military structure. Intelligence must play a role, which means strengthening integration of ISR throughout the command structures; but also training, equipment, EW, anti-drone hardware or ‘add-ons’ to vehicles, etc.
The long-promised Russian Arena-M APS (Active Protection System) was recently spotted on a T-72B3M on the front for the first time, after being seen on production lines a month ago:
It’s uncertain if they will try to tune it against drones, though this is unlikely, given that drones fly too slow. Programming the system to discriminate slow-moving objects would be dangerous, as the radar could cause the system to fire at birds or other ‘false targets’, which could injure nearby troops. But we’ll see—it’s possible the sudden roll-out could be drone-related.
iii.
The question remains: as drones continue rising proportionally in use against other weapons systems on the battlefield, can Russia maintain forward momentum, or is the front fated to grind ever more to a standstill, as advancement becomes impossible without major losses?
Firstly, let us say that the current trend of atomized small-group assaults will obviously continue, as will the use of civilian vehicles, which will only grow in proportion, partly because of various battlefield expediencies, and partly due to Russia not being able to afford outsize armor losses as in the first half of the war. Likewise the importance of asymmetric tactics like now seen in Kursk, with Operation Potok—the infamous pipeline operation which bypassed the lethal drone killzone via underground safe passage. Of course, to some extent none of this is new—just recall World War One’s Battle of Messines, where 10,000 German troops were allegedly blown up by an underground explosion laced by British sappers, as seen in The War Below. But it’s simply to say that increasingly ‘innovative’ tactics will have to be utilized to bypass deadlocks. The most common of these is simply waiting for the weather to become unfavorable for drones in the way of heavy fog, rain, etc., and then attacking. The Russian side very often relies on this method to make advancements, including generating their own smokescreens as was used even in the battles for Sverdlikovo in Kursk, at least by one account.
