30 vs 2 drones a day: the difference between battalions isn't drones — it's who was allowed to think Author: Дністер Published: 2026-06-19T03:01:16.000Z Language: en URL: https://neurodrift.org/en/blog/odyn-batalion-trydtsiat-droniv-inshyi-dva/ Original (Ukrainian): https://neurodrift.org/blog/odyn-batalion-trydtsiat-droniv-inshyi-dva/ Tags: drones, war, asymmetry, technology, adaptation Original source: https://neurodrift.org/blog/odyn-batalion-trydtsiat-droniv-inshyi-dva/ An OSINT dissection of why two near-identical units near Pokrovsk produce drone loss rates that differ by an order of magnitude. The answer isn't in the hardware and isn't in the enemy — it's in an invisible procedure that one unit follows and the other considers an insult. From Semmelweis and Boyd's loop to Colonel Budnikov and Rubicon. ----- Two wards, the same year Vienna, 1847. Two maternity wards in the same building, separated by a corridor. In the First, women die of puerperal fever at an average rate of 18 in every hundred — in certain months, more than thirty. In the Second, fewer than two. Same hospital, same medicines, same year, same weather outside. Women being wheeled in beg not to be put in the First; some would rather give birth on the cobblestones outside — and survive more often than those who made it to a bed. The First ward is the only place in the city where it is safer not to arrive. Ignaz Semmelweis, assistant in the First, searches for the difference for years. He goes through everything: bed placement, the position during labor, even the direction the priest walks through the ward to the dying, ringing his bell. The answer comes through a death. His friend, the pathologist Jakob Kolletschka, accidentally cuts himself with a student's scalpel during a dissection — and dies of the same clinical picture as the mothers. Semmelweis understands: the difference between the wards is that in the First, doctors and students come to deliver babies directly from autopsies, without washing their hands. They carry death on their fingers and call it experience. He orders hands washed with chlorinated lime solution. Mortality in the First drops from eighteen percent to under two — to the level of the ward next door. Proven not by theory but by washing. And here is where the real story begins: colleagues were not grateful. The head of the clinic, Johann Klein, read the procedure as a personal insult — the implication being that his doctors had been killing women with dirty hands. It was simpler to keep killing than to admit you had been killing. Semmelweis is driven out of Vienna, hounded for years, and ends his life in a psychiatric institution to which he was committed — where, by a grim irony, he is beaten by the guards and dies of blood poisoning. The very same kind he spent his career preventing. I keep this story in my head every time I read a dispatch from around Pokrovsk. Because out there right now is the same building with two wards. One battalion loses close to thirty FPV drones a day, burned by enemy electronic warfare. The one next door, on the same sector, against the same EW, flying the same drones — loses two. And the headquarters of the first writes in the combat log a phrase I recognize as Johann Klein word for word: what can you do against an enemy like this. My stake here is personal. I've spent years building systems — information systems, business systems, whatever you like — and I know from my own experience exactly what a collective looks like when it attributes its own inefficiency to external circumstances. It is the most comfortable lie in the world: it blames nobody inside, and therefore never ends. And when I see that the same lie now costs not money but front-line territory and burned-out soldiers, I need to understand its mechanism down to the last bolt. And that mechanism is factory-grade. Imagine a battalion not as a unit but as a production line. Input: drones, operators, fuel. Output: targets hit. And as on any conveyor, sooner or later defects begin rolling off the line — airframe after airframe goes down before reaching the target. This is where the entire difference hides. On Toyota factory floors, a cord hangs above each section — the andon. Any worker who spots a defect has the right to yank it and stop the entire line until the cause is found. In a bad shop, the worker who yanks the cord gets a dressing-down for wrecking the plan — so they silently let the defect through and pray someone else notices. One battalion near Pokrovsk lets the operator yank that cord. The second punishes the yanking with exclusion from the tea break. Because EW is childbed fever 2026: invisible, soundless, burns you on the way to the target. The question isn't whether it can be washed from your hands. The question is who on this line has the right to stop it — and who gets thanked for stopping with an incident report. EW is childbed fever 2026. And Semmelweis is the officer the brigade calls a paranoid for changing frequencies every day. How many actually don't arrive Numbers first, because without them this is small talk. And the most honest figure here doesn't come from a general's briefing but from a peer-reviewed front-line dissection. According to RUSI (Watling and Reynolds, February 2025), 60 to 80 percent of Ukrainian FPVs do not reach their target — and RUSI specifies directly what drives that spread: the sector of the front and operator training. Stop on that phrasing. A twenty-point spread in percentage the institute attributes not to the make of the airframe but to who flies it and how. The range itself is no longer a hardware indicator. It is a shop indicator. What this means on the conveyor is visible from a second figure. The OSW Centre for Eastern Studies (October 2025) counts: one brigade burns through several hundred FPV drones per month, while the real need is around 2,500. The front fights on twenty to forty percent of what is required — and not because the drones were never built. In 2024, Ukraine, according to Zelensky, produced up to 2.2 million UAVs (the Ministry of Defence cited 1.7 million, the General Staff — 1.3 million delivered to troops; the entire discrepancy is in what you count: manufactured, contracted, or delivered to the trench). However you count — it's millions. The factory runs full shifts. It's just that on the other end of the conveyor stands an enemy jammer, pulling product off the line faster than it can be boxed. Now for what happens inside the shop — and why the finished number cannot be trusted here. The most detailed public sortie log was kept not by a headquarters but by an operator: Jakub Jajcay, a former Slovak Armed Forces officer who fought for six months inside a Ukrainian drone team in Donbas and then sat down and counted his missions in an essay for War on the Rocks. His arithmetic is not a think-tank dataset but a record of one unit over one period; the exact percentages cannot be extrapolated to the whole front, and I don't present them here as fact. But one detail deserves to be held up to the light, because no general's report will volunteer it: a portion of the airframes were killed not by enemy jamming but by their own Ukrainian jamming station — switched on without notifying the operators. Picture that on the assembly line: you're guiding an airframe to the target and a neighboring shop kills it — by switching on its own jammer and telling no one. Not sabotage. Just the left hand of the conveyor jamming what the right hand launched, and both then signing the same incident report about 'difficult electromagnetic conditions.' It is worth stopping here. Two million two hundred thousand — that is not a number of shortage. It is a number of saturation. Ukraine crossed the production barrier in 2024. That means procurement is no longer the bottleneck: drones exist, they are sufficient, the conveyor doesn't stop. The difference between 'battalion A loses two airframes' and 'battalion B loses thirty' lies entirely beyond the factory gate — in organizational process, not the production line. Once production is solved, the only remaining level is training. This is the 'why now' — and the answer is not technical. And here enters the figure one must be honest about, because it is widely abused. You've surely heard: 'Ukraine loses 10,000 drones a month.' That figure was first cited by RUSI in May 2023 and has circulated as an eternal mantra ever since, as if nothing changed in three years. It is folk statistics. Since then, both production and losses have grown by an order of magnitude; presenting 2023 as today is the equivalent of treating COVID with a plague manual because 'patients coughed there too.' I cite it only to stamp it 'expired' immediately — because the load-bearing thesis of this piece is precisely that whoever refuses to update its data is the one losing. Go back for a moment to those twenty RUSI points. That is not a range of hardware quality — it is the difference between two shifts on the same conveyor. The first shift lets 60% of output through as defective; the second, 80%; the airframes are the same, but the foreman is different. In the gap between those numbers there is no new technology. There is a person with a procedure — our Semmelweis, hidden inside the margin of error and invisible to everyone until someone bothers to count. Below: the two shifts side by side, in two columns. Ward A — the "adapted" unitWard B — the "what can you do" unit Changes working frequencies daily, sometimes hourlySits on "stable" frequencies because "it worked yesterday" Every downed drone: debrief — where, when, on which frequency was it cutEvery downed drone: a line in the loss report Maintains a map of enemy EW built from its own lossesNo enemy EW map; just a sense that "they're jamming over there" Tests fiber optics, autonomous guidance, new countermeasuresWaits for "a solution from above" Operator experience feeds up and comes back down within hoursOperator experience burns with the operator Losses: single digits per dayLosses: dozens per day Two columns. One building. The same enemy across the line of contact. The entire difference is in the right column, and not one item there is about hardware. The right column isn't jammed by EW. It is jammed from the inside. !Two wards through a glass partition: on the left an operator writes the frequency of a downed drone on a map (two downed), on the right an operator sits without a map beneath a pile of… The same building, two wards down the corridor. On the left, a downed drone becomes a line on the map — on the right, a line in the loss report. And the officer who suggests 'washing hands' on the radio is listened to in one ward and shown the door in the other. The firefly is already going out under someone else's palm. The person with a face on the other side This difference has an antagonist, and he is not abstract. His name is Guards Colonel Sergei Budnikov, and he commands what the Russians called the 'Center for Advanced Unmanned Technologies "Rubicon."' Created by a Belousov ministry order in August 2024. An RFE/RL investigation (September 2025) counted seven known detachments of roughly 130–150 operators each; by November, the Financial Times put Rubicon at around 5,000 fighters, each detachment specialized by airframe type. It strikes logistics and equipment 15 kilometers behind the line, played a notable role in Russia's recapture of the Kursk region. How many kills exactly — a separate question: the figure 'over 13,800 strikes' (as of December 2025) was picked up by analysts from the Lostarmour tracker — a pro-Kremlin OSINT project — so read it as a claim from an interested party, not a verified total. An American volunteer fighting for Ukraine put it simply: "The game changed when they arrived." And here is why Budnikov is the perfect antagonist — the genuinely unsettling kind. He is not a genius. He has no secret better weapon. Rubicon takes the same drones and the same EW that lie scattered across the front on both sides. Its advantage is purely organizational: it concentrated expertise into one fist, logs every sortie, and drives experience up and down the loop within hours, not months. This is not a center for advanced technology. It is a center for advanced learning, with an intimidating name on the door. Which is precisely why it cannot be beaten by buying a more expensive airframe: whoever keeps a journal will also learn to shoot down the expensive one. And now — the most uncomfortable part. The real antagonist of this piece is not Budnikov. He is beyond reach; he is on the other side. The real antagonist is his mirror image in our uniform: the staff officer who has a report about thirty drones burned in a single day placed on his desk, and who, without looking up, says 'well, what can you do against an enemy like this' and goes to make tea. That is also a center for advanced technology. Only the technology is alibi. The internal 'reverse Rubicon' — the same 'Center for Advanced Technology' with a placard on the door, only behind the door they kill mothers with unwashed hands and yet maintain a flawless journal. Data is collected here with equal tirelessness — only not on sorties but on reasons nothing needs changing; and in that register not a single gap, not a single smudge, neat handwriting throughout. That is the foreman who spotted the andon cord above the line and welded it shut so no one could stop the plan. The conveyor can never be halted now — and this is presented as a sign of discipline. But calling this incompetence is too simple, and wrong. The staff officer who writes 'what can you do' is usually not stupid. He is making a rational career choice. Because if he acknowledges that Semmelweis was right — that frequent frequency rotation would have cut losses tenfold — this is not merely a methodological correction. It is a retroactive admission that all thirty airframes yesterday, and all thirty the day before, burned because of his procedural decision. Klein was not defending dirty hands. He was defending his own unimplication. The same logic applies here: it is simpler to write 'what can you do' forever than to sign, once, under thirty pairs of charred operator gloves and the list of names of the operators who went with them. This is not helplessness — it is self-protection, formatted as an incident report. This is not a center for advanced technology. It is a center for advanced learning. Which is precisely why it cannot be beaten by buying a more expensive airframe. Why identical machines produce different results: Boyd's loop Military theorist John Boyd once described why the faster decision wins a dogfight, not the faster plane. He called it the OODA loop: Observe — Orient — Decide — Act. Whoever cycles through it faster than the adversary sets the tempo, and the adversary is forever reacting to yesterday and can't catch up with today. The unit that loses thirty drones a day has a broken loop. Each downed airframe doesn't return to the system as a lesson — it returns as a damage entry. 'Observe' exists (we can see they're falling). 'Orient' is already missing (why here specifically, on which frequency, at what time of day — nobody compiled it). 'Decide' and 'Act' degenerate into 'request more drones.' The loop closed not on learning but on restocking the supply depot. This is a conveyor whose andon cord someone welded shut: defects roll off the line continuously, and nobody has the authority to stop it to find the cause. The result is the world's most expensive automated feeder for enemy jammers — it doesn't even need reloading: the headquarters delivers the feed by itself, shift after shift, recording in the journal with even handwriting how much it served up today. !Two battalions at the same table separated by a sandbag: on the left an operator feeds a downed drone to the map as a lesson (the loop closes, two losses), on the right… On the left, a downed drone returns to the map as a lesson — Boyd's loop closes on two losses. On the right it returns to the pile of thirty, and the senior is already pushing a new one into the operator's hands: the world's most expensive feeder for enemy jammers. The firefly on the edge is already dark — clenched in a fist, because someone decided a signal in the dark was more dangerous than defeat. A scene from The Imitation Game comes to mind here. Alan Turing's team at Bletchley Park resets all its overnight work every morning — because at exactly midnight the Germans change the Enigma settings, and the entire previous day's progress turns to garbage. Colleagues break the cipher by hand, message by message, and every day they don't finish in time. Turing understands: you can never win by hand, because the enemy resets you every night. You need a machine that changes along with the enemy. The direct rhyme to EW: a unit that changes frequencies and tactics every day is Turing's machine. A unit waiting for 'a stable solution' is one that cracks yesterday's Enigma by hand every morning and genuinely wonders why it can't keep up. The enemy resets you every night too. The only question is whether you've noticed — because he is not waiting for you. Psychologist Martin Seligman in 1960s experiments showed another face of this coin. Dogs shocked regardless of their actions were confined so escape was impossible; later, when the cage was opened, they no longer tried — they lay down and endured the current next to open doors. He called it learned helplessness. An institution learns the same thing. A headquarters that has written losses off to 'external circumstances' thirty times no longer looks for internal levers — it no longer sees them, the way the dog no longer sees the open door a step away. 'What can you do against an enemy like this' is not a situational assessment. It is a symptom. A diagnosis of learned helplessness at unit level, neatly formatted in administrative language and filed away. What already works against jamming (and why it's not about hardware) The skeptic will say: maybe the problem really is in the technology — EW is genuinely fearsome. Yes, it is. The Russian mobile Shipovnik-Aero complex classifies a drone in roughly 25 seconds, severs the command channel to within a three-degree azimuth — and, if parameters match, intercepts control: substitutes the home-point, paints a false navigation field, and sends the drone wherever the station's operator wants it. Range around ten kilometers, works against two targets simultaneously. Meaning your drone doesn't just fall — it turns around and flies obediently off to surrender, sincerely believing it is heading home. The hardware is genuinely diabolical: it doesn't kill the airframe, it recruits it. But there's a stronger version of this objection, and it's not about hardware. Perhaps battalion A simply got a better sector — flatter terrain, thinner EW coverage, a lucky grid square. If the difference between thirty and two is a geographic lottery rather than a procedure, the entire argument collapses into a pretty metaphor with no operational application. This objection deserves a direct answer. Jajcay's data and RUSI's analysis are not about comparing sectors — they are about dispersion within identical conditions: same sector of the front, same enemy, same drones, different operator and procedure. RUSI states it precisely that way: the 60–80% spread is explained by training and sector — and 'sector' is listed second there, training first. If the sector decided everything, a better sector couldn't be transferred to a neighboring unit; but the tactic of rotating frequencies can be. Which is why the difference is reproducible, not accidental. And it is equally telling that the answer has already been found — and it is again not 'a more expensive drone.' A fiber-optic airframe trails a cable thinner than fishing line: a physical link that is impossible to jam in principle, because there is no radio channel to break. Range 5–20 kilometers, prototypes up to fifty. Accuracy high even where the airwaves are dead. Price of the fiber: from 500 dollars, often more expensive than the drone itself; the downsides are real — weight, inertia, weather sensitivity, and the cable itself leaves a trace. Defeating an invisible jammer costs a spool of wire visible from space. But it works where EW wins everything. The other branch is autonomy. Kamikaze drones with Auterion guidance locked onto the target a kilometer before impact and did not release it when jamming was switched on during the approach — and burned a column of Russian tanks festooned with their own EW, in a situation where, according to participants, 'everything was jammed.' A human operator at that moment would have gone blind. The autonomous airframe did not: the Ghost Dragon neural network from the Estonians cross-references local landmarks against a map the way a navigator cross-references a chart, and guides the drone with no GPS at all. There is nothing to jam — the brains are riding inside. Now look at the logic of all three solutions — fiber optics, autonomy, daily frequency rotation. None of them is 'more powerful.' Each simply changes the rules so that the enemy's brute force bounces off. This is the Holtzman shield scene from Dune: the shield lets through a slow blade and deflects a fast one, so the winner is not the stronger blow but whoever adapted their tempo to the enemy's defense. EW is a shield that catches everything that hits fast and dumb: more drones, more expensive hardware, frontal assault. Only whoever changed approach gets through — not force. Whoever hits faster and dumber is feeding the enemy's shield with their own drones. And this is ultimately not an engineering thesis but an organizational one: the technology to change the rules already exists and sits on shelves. Those who have a shorter loop take it. The rest wait for orders — and will wait until the next briefing about thirty burned machines. Your drone doesn't just fall — it turns around and flies off to surrender, thinking it's going home. The landscape as a black box: a mirror for you There is a detail that gets to me every time. April 2025, fields near Pokrovsk and Kostiantynivka. By eyewitness accounts, among the most densely fiber-cable-covered patches of earth in the world — thousands of thin threads dropped by drones lying on stubble and hanging from bare branches. In the sun the field glints as though a giant spider passed through and forgot to clean up. This is a physically visible landscape. Now the inference: you can read from this landscape who flew and how much. Adaptation and failure leave the same glinting trail on the ground — but whoever learned to read their own tracks is alive and reading still. And this is not only about the field. It is OSINT methodology, and it is a mirror for anyone who imagines their decisions are invisible. From fiber density, from the frequency of Rubicon strikes, from the published RUSI and OSW reports, an outside analyst can reconstruct which unit adapted and which didn't — without access to a single combat log. The traces of your decisions are always visible from outside. The only question is whether you read them yourself — or only the enemy does, having already drawn conclusions. Military expert Maria Berlianska named the root of the difference directly, without anaesthetic. Her prescription is organizational, not technical: build 'Anti-Rubicon' hunter teams targeting enemy drone crews, copy adversary tactics (the Russians unselfconsciously copied Ukrainian FPVs, fiber optics, and air-defense concepts), and embed 'technology managers' in brigades — people whose sole job is to collect, analyze, and distribute the best techniques, both homegrown and captured. Call it what you like — in factory terms, this is a shift supervisor for quality control, the person with the authority to yank the andon, stop the line, and circulate across all shops exactly why the defect was rolling off. The Ukrainian systemic answer already exists — it is the 'Drone Line,' assembling elite regiments (the 20th 'K-2,' the 414th 'Madar's Birds,' the 429th 'Achilles,' the 427th 'Rarог,' 'Phoenix') into continuous 'kill zones' 10–15 kilometers behind the line of contact. What percentage they take out there — I won't assert without a primary figure, and the figure isn't the point. The point is that systemic adaptation is visible and measurable — exactly as visible as its absence. The wheel has turned full circle. But the hardest line in all of Berlianska's diagnosis is a single sentence that belongs on a placard in every headquarters: "It's a system. And on most Ukrainian fronts we still have chaos." That is the entire diagnosis. Not 'the enemy has better drones.' System versus chaos. The ward that washes its hands versus the ward that considers doing so an insult. The thinnest thread: when the whole system is one person Here I need to kick out my own thesis's legs before the reader does. Because it's easy to picture the 'adapted' battalion as a flawless hero — and it has one flaw, and it is fatal. Ask simply: where does all this adaptation live? In the enemy EW map built from their own losses. In the habit of changing frequencies every hour. In the feel for which sector today has a dead ether. And almost always, this map lives not in a documented procedure but in the heads of two or three veterans — that same 'paranoid' whom ward A prizes. The adaptive unit is often not a system. It is a loop closed on a specific individual. And on a battlefield, individuals get rotated, wounded, burn out, or die. And here is where the factory analogy stops being pretty and starts cutting. On a good line, knowledge doesn't live in the foreman's head — it is extracted into a standard operating procedure: a document where every step of producing this part is written down, so tomorrow a newcomer can do it, not only the veteran who 'feels the lathe.' Toyota calls this standardized work for exactly that reason: the person is mortal, but the line must survive a change of personnel. A unit whose adaptation lives in the veteran's head and nowhere else is a workshop where a single pre-retirement machinist holds the entire technology. While he's in place — records. He goes on rotation (best case) — and the new shift steps onto the conveyor without even an instruction for switching the machine on. The enemy EW map goes with him. The new crew starts from a blank page — and within a week discovers it is losing thirty machines again, because it is learning what the predecessor already knew but never wrote down anywhere. This is precisely why Berlianska's 'technology manager' is not a bureaucratic nicety but a load-bearing beam. His real job is not catching lessons. It is extracting the lesson from a person's head into a procedure — while the person is still alive and on position. Without that, even ward A is not a system but a lucky streak with a face — one that will be shot away along with the face. A loop closed on a person breaks precisely where the person breaks — and the enemy doesn't even have to wait: it suffices to wait for the rotation. The thinnest thread on this field is not fiber optic cable. It is undocumented experience belonging to someone who might not show up for the next shift. Who burns and who writes the report In all this machinery there is one dimension easiest to avert your gaze from — and the one that most precisely explains why the system doesn't correct itself. It is neither technical nor organizational. It is personal. The FPV operator in unit B goes out on a sortie in a drone costing from a few hundred to a few thousand dollars. The airframe lifts off. Shipovnik-Aero captures the channel in twenty-five seconds. The drone turns around and flies into captivity — or falls on its own people. The operator at the console watches his drone vanish into an electromagnetic void: no target, no return, not even the spectacle of an explosion. Only a dead screen. He makes an entry in the log, goes to sleep, and tomorrow launches another. Thirty times a day — he burns out with the machines, only more slowly. The staff officer who writes 'what can you do against an enemy like this' that same evening bears no personal consequences. His career does not burn with the drone. His night does not shorten with each failed sortie. He does not buy a new airframe out of his own pocket. He simply draws his alibi in even handwriting — and tomorrow receives a salary that bears no relationship to how many lines fill his loss log. Semmelweis showed: the difference between the wards is a difference in who pays the price. The mothers paid with mortality. The doctors paid only with inconvenience. As long as inconvenience remained smaller than the price of acknowledgment, hand-washing did not become the norm. The same is happening in battalion B today: the price of organizational blindness falls on the shoulders of those who hold the controller — not on those who write the report. The tool: a childbed fever self-test A mirror without a tool is a sermon. So here is a checklist — factory acceptance inspection for your own line. It is not about drones: substitute your unit, your department, your company, your life. Five questions. Answer honestly — this is that rare quality control where the only defect you can pass is the one you pass past yourself, and the bill arrives regardless. The most expensive defect here is the confidence that the conveyor is calibrated and touching it would be a sin. Question for yourselfWard A (washes hands)Ward B (childbed fever) When did we last change a procedure because of our own failure?This weekI don't recall / "it works" Where does the experience go when someone burns out or screws up?Becomes a lesson for everyone within hoursDisappears with the person Who is personally responsible for making sure lessons travel upward?A role exists with a name attached"It sorts itself out somehow" How do we explain recurring losses?"Our procedure" (looking inward)"Such an enemy / market / client" (looking outward) What do we call the person who paranoidly changes and checks everything?Valuable; they're listened to"Pain in the neck"; they're routed around If you answered in the right column three times or more — you have childbed fever. Doesn't matter whether you are fighting near Pokrovsk or compiling a quarterly report: the mechanism is identical — an invisible procedure that one collective follows and another considers an insult. And here is the question with which this checklist hits hard. Your reflex of 'don't touch the conveyor, it's calibrated' — whose interest does it serve? Sometimes genuinely yours: not every line needs to be torn apart every morning; constant stoppages also kill production. But that exact same reflex is your enemy's or your competitor's best ally, because while you guard what's 'calibrated,' he resets his Enigma every night and waits for you to come on shift with yesterday's frequencies. The entire trick is to tell these two cases apart honestly, not in hindsight. Because the worst of it is that childbed fever is invisible from inside until someone on the outside counts your threads in the field. From inside, ward B doesn't feel like a place where people are dying. It feels like a place that just got unlucky with its enemy. The blade Semmelweis died in a psychiatric institution of blood poisoning — the same rot he had spent his career fighting, beaten by those who should have listened to him. It wasn't the disease that killed him. It was the institution for which hand-washing felt like an insult. He was right twenty years before being proven right — and that didn't save him, or the mothers of those twenty years. Being right is not a life jacket. At the bottom it sinks with you. So don't comfort yourself with the thought that 'the truth will prevail by itself.' It doesn't prevail by itself — it sits on the shelf next to the fiber optics and the autonomous guidance, available to both wards, and it is taken by whoever has a loop that is an hour shorter. Colonel Budnikov is not waiting for history to vindicate him. He is simply cycling OODA faster than your staff counterpart who is finishing his tea over a briefing about thirty burned machines and writing in even handwriting: what can you do against an enemy like this. The difference between thirty drones and two is not a difference in technology, not in money, and not even in the enemy. It is the difference between whoever washes their hands every day and whoever considers doing so an insult. Between whoever gave the worker the right to yank the cord and stop the line — and whoever welded the cord shut to protect the target. Childbed fever does not ask what you believe. It counts only what you do with your hands — and leaves a trail in the field visible from space. Look at your hands. Look at your line. Someone is already counting your threads in the stubble — and he is not on your side, and he is not waiting for the end of the shift. Sources: Jakub Jajcay, "I Fought in Ukraine, and Here's Why FPV Drones Kind of Suck" — personal sortie dissection from one unit (including airframes taken down by their own jamming station without operator notification). Specific success percentages are one author's count for one unit over one period, not independently verified and not extrapolated to the whole front in this text — War on the Rocks: https://warontherocks.com/i-fought-in-ukraine-and-heres-why-fpv-drones-kind-of-suck/ "Game of drones: production and use on the Ukrainian battlefield", OSW Commentary #694, 14.10.2025 (one brigade burns through several hundred FPV/month vs. estimated need of ~2,500; ~2.2 million UAVs per Zelensky's statement) — OSW Centre for Eastern Studies: https://www.osw.waw.pl/en/publikacje/osw-commentary/2025-10-14/game-drones-production-and-use-ukrainian-battlefield-unmanned Discrepancy in official 2024 UAV production figures: Zelensky — up to 2.2 million; Ministry of Defence (Umerov) — 1.7 million; General Staff — 1.3 million delivered to troops (manufactured / contracted / delivered — different counting bases) — Kyiv Independent: https://kyivindependent.com/ukrainian-drones-made-up-over-96-of-uavs-military-used-in-2024-defense-minister-says/ J. Watling & N. Reynolds, "Tactical Developments During the Third Year of the Russo-Ukrainian War", RUSI, 14.02.2025 — 60–80% of Ukrainian FPVs do not reach their target (non-delivery metric, not 'non-destruction'), depending on sector of front and operator training — RUSI: https://static.rusi.org/tactical-developments-third-year-russo-ukrainian-war-february-2205.pdf "Ukraine Losing 10,000 Drones Per Month to Russian Electronic Warfare" (RUSI estimate, originally May 2023 — outdated, marked here as folk statistics) — UAS Vision: https://www.uasvision.com/2023/05/24/ukraine-losing-10000-drones-per-month-to-russian-electronic-warfare/ Counter-UAS 101 / Electronic Warfare (Shipovnik-Aero specifications: ~25 s classification, 3° azimuth, home-point interception, ~10 km range, two targets) — drone-warfare.com: https://drone-warfare.com/counter-uas/electronic-warfare/ "Ukraine's Autonomous Killer Drones Defeat Electronic Warfare" (fiber optics, fiber cost, Auterion, Ghost Dragon) — IEEE Spectrum: https://spectrum.ieee.org/ukraine-killer-drones "Center for Advanced Unmanned Technologies 'Rubicon'" (Col. Sergei Budnikov; founded by Belousov order, August 2024; logistics strikes up to 15 km; role in Kursk region; ~5,000 fighters estimate — Financial Times, November 2025) — Wikipedia: https://en.wikipedia.org/wiki/Rubicon(UAVunit) "Inside Rubicon, The Elite Russian Drone Unit…", RFE/RL, 17.09.2025 — original investigation: seven known detachments of roughly 130–150 operators each — RFE/RL: https://www.rferl.org/a/russia-drone-rubicon-secret-ukraine-war/33532804.html The figure "over 13,800 strikes" (as of 18.12.2025) originates from the Lostarmour tracker — a pro-Kremlin OSINT project, circulated secondarily through analysis; this is a claim from an interested party, not a verified total — drone-warfare.com: https://drone-warfare.com/research/russia-unmanned-forces/ "Ukraine must build anti-Rubicon hunter units…" (Maria Berlianska: 'technology managers,' copying tactics, 'it's a system — and we have chaos') — Euromaidan Press: https://euromaidanpress.com/2025/08/23/ukraine-must-build-anti-rubicon-hunter-units-to-target-elite-russian-drone-crews-says-expert/ "The Drone Line: what is it, and can five units stop the Russian advance?" (10–15 km kill zone; K-2, Madar's Birds, Achilles, Rarог, Phoenix) — Ukrainska Pravda / Defense Express: https://www.pravda.com.ua/eng/articles/2025/05/05/7510582/ Ignaz Semmelweis (Vienna 1847, mortality 18%→<2%, Kolletschka, Klein, psychiatric institution, death from sepsis) — Wikipedia: https://en.wikipedia.org/wiki/IgnazSemmelweis Martin Seligman, learned helplessness (1960s experiments) — Wikipedia: https://en.wikipedia.org/wiki/Learnedhelplessness