27 Ukrainian FP 2 Fire Point UAVs had done what the Russian military believed was impossible.
These drones pierced through one of the densest and most advanced air defense networks in the world.
They flew low along the rolling hillsides, crossed through the valleys, then skimmed over the treetops on the slopes of the Donetsk Ridge, quietly entering the part of Donetsk that Russia claimed to control.
Their target was Amvrosiivka, a major ammunition depot that Russia had just reinforced with another 1,200 tons of weapons and ammunition.
By 05:25 hours local time, at a Russian S-350 Vityaz battery, the duty commander suddenly saw several abnormal signals on the radar screen.
They did not look like a clear attack formation.
They did not look like Ukrainian fighter jets that the system could lock onto immediately.
They were only faint, broken traces appearing and disappearing inside the terrain clutter.
But to an officer with 3 years of experience facing Ukrainian UAVs, this was not a system error.
It was not a false signal, either.
Something was entering the weakest part of Russia’s air defense network.
And only 8 seconds later, those faint traces began linking together, then emerged clearly as a real attack formation.
The radar of the S-350 Vityaz is one of the most important all-seeing layers in Russia’s air defense network.
The 50N6A multi-function radar of the S-350 Vityaz system can track 40 targets simultaneously within a range of 250 miles.
But what was appearing on the screen at that moment did not look like anything the combat crew would imagine.
These were FP-2 Fire Points with radar reflections 200 times smaller than the targets the system was designed to track.
On the 50K6A combat control screen, 27 signal traces appeared and disappeared.
They were not clear enough to lock and fire immediately, but they did not vanish completely either.
The battery commander leaned closer to the screen.
Those glowing traces did not move like birds.
They did not look like weather clutter.
They advanced steadily, slowly, low, and with a strange rhythm inside the radar return, a small, regular vibration, like the footprint of piston engines and propellers cutting through the air at high speed.
That was the moment he understood.
This was a Ukrainian UAV swarm entering the dead zone of the S-350 Vityaz.
The distance was only 20 mi.
For the 9M96E missile, that was a perfect interception range.
The commander immediately pressed the fire button.
Two missiles left from the launchers, throwing up two white columns of flame, accelerating to Mach 4, then bending toward the faint signal traces skimming along the hillside.
But at an altitude of 210 ft, physics began working against the S-350.
Each radar pulse did not only touch the UAVs, it also bounced off the hillsides, reflected from metal roofs, broadcast antenna clusters, and the terrain below.
On the screen, the real target and its reflected shadow overlapped like a spider web.
The missiles saw more than one target, one real trace, one false trace, one signal in the correct position, but flickering, one stronger signal amplified by reflection from objects on the ground.
The guidance computer had only 300 milliseconds to choose, but both missiles chose wrong.
They exploded in the air, tearing two large rings of fire through the low atmosphere.
The shock wave swept across the area, but not a single UAV was removed from the formation.
The FP-2 fire points were no longer there.
They had cut their engines for 6 seconds, dropped even lower, and let momentum pull them through the clutter like ghosts disappearing from radar’s eyes.
Inside the control room, the S-350 commander froze in disbelief, staring as the two interception points turned into two meaningless circles of light.
Then two signals suddenly flared.
Two lead UAVs abruptly climbed to 2,100 ft, their engines at full power, exposing their entire bodies and wings to radar as if they were preparing to strike an important target.
They turned themselves into the clearest targets in the sky.
The S-350 crew immediately shifted lock.
The next two 9M96E missiles were launched.
This time the Russians had learned from the previous firing sequence.
The proximity fuse was activated when each missile was only about 15 ft from the target.
Two fireballs erupted almost at the same time.
The two lead UAVs disappeared from the radar screen, but that was exactly what Ukraine needed.
The time the S-350 spent aiming, shifting attention, and focusing on the two UAVs that had just climbed was enough for the remaining 25 to break through.
While the two decoy targets were intercepted in the air, the rest of the formation had restarted their engines, dropped back down to hug the terrain at 250 ft, and scattered out of the main line of sight.
They spread across a 12-mi front, each small signal trace slowly dissolving into ground clutter.
Behind the commander, the system was still trying to rebuild the battlefield picture, but the UAV swarm had already passed through the first interception layer and was continuing to race toward Amvrosiivka.
But exactly 1 hour later, the 25 remaining UAVs would encounter an invisible wall they had no idea was waiting ahead.
At 06:26 hours local time, the surviving formation entered the fortress belt.
They were flying 500 to 650 yd apart, holding an altitude of 220 ft and maintaining a speed of 100 mph.
The terrain here was a nightmare for the operators.
Rolling hills, mining sites, metal warehouses, and sparsely populated settlements created hundreds of points of reflection, concealment, and interference.
That was also exactly why the Russians had camouflage Borisoglebsk 2 close inside old industrial infrastructure zones.
The first sign was very small.
The navigation receiver on the lead UAV still reported locks on 12 GPS GLONASS satellites.
The signal was stable.
The flight path was still correct.
But as soon as it entered the Borisoglebsk 2 coverage zone, all communications and satellite navigation signals were immediately suppressed across a wide area by extremely powerful jamming pulses.
Then the number dropped to 10.
Then to five.
And only 5 seconds later, every satellite signal disappeared.
It did not fade gradually like when flying into a block zone.
It was a cold, instant cut.
The navigation sky above them had just been erased.
Borisoglebsk 2 was designed to attack what UAVs depend on most: communication, control, and navigation.
Its jamming stations were hidden near warehouses, antenna clusters, communication sites, and high points around the flight corridor, covering the area with a dense electronic interference layer 120 miles wide, striking directly at the GPS L1, L2, and L5 satellite signal bands and the GLONASS frequencies the UAVs needed to stay on course.
GPS signals from orbit 10,000 miles above the ground are already extremely weak by the time they reach a UAV’s receiver.
Borisoglebsk 2, meanwhile, broadcast jamming from the ground thousands of times closer.
For the small receiver on the FP2 Fire Point, that difference was no longer ordinary interference.
It was an electronic storm.
The lead aircraft’s navigation screen shifted into warning status.
Coordinates began to drift.
The flight path was no longer being corrected by satellite.
From that second on, the 25 FP2 Fire Points could rely only on their internal inertial systems, gyroscopes, accelerometers, and preloaded flight route data.
But inertial navigation is not perfect.
At 100 mph, the formation covered another 1.
8 miles every minute.
With only a 1° angular error, after 16 miles, a UAV could already drift 400 ft off course.
After 30 miles, the error could exceed 800 ft.
Over the low hills of Donetsk, that was enough to pull the entire formation out of the safe corridor into air defense fire zones or cause it to miss the approach route to Amvrosiyivka completely.
The 25 UAVs still maintained absolute silence.
No transmissions, no active sensors, no attempt to communicate.
They flew like mute black shadows inside a corridor that Borisoglebsk 2 had choked shut.
The processor on the lead UAV began recalculating.
The probability of mission success was dropping fast.
If the entire formation continued on the old heading, the error would grow with every minute.
If two aircraft broke formation and forced the Russian system to reveal a gap in the jamming zone, the remaining aircraft might survive.
And then the four lead aircraft made their choice.
They wanted to save the mission.
They began drifting away from the formation, climbing slightly and widening their flight paths, turning themselves into test decoys for the Borisoglebsk 2 electronic wall.
It was a cold decision, calculated by algorithm, but its meaning was very simple.
Two aircraft had to pay the price so the remaining 21 could continue the mission.
At 06:29 hours local time, the four lead FP-2S began breaking radio silence, deliberately accelerating as if they were about to enter a dive.
For exactly 3 seconds, their communication systems emitted an extremely short burst of signals, a request for satellite navigation correction, a target coordinate update packet, and a command to resynchronize the flight formation.
The Borisoglebsk 2 stations positioned around the fortress belt immediately detected the abnormal emissions.
And in that brief moment, the four lead UAVs turned themselves into the only bright points inside an airspace that had been completely silent.
Warning data was pushed back to the rear air defense line.
A Buk-M 3 battery immediately received the suspected coordinates along with the closing speed.
The Buk-M 3’s 9S36M radar rotated toward the signal that had just flashed.
On the control screen, four small traces appeared unclear, unstable, but enough to confirm that targets were approaching at low altitude.
The battery commanders did not need to ask anything else.
They had the approach direction, they had the target zone, and they had the order to fire.
Within 3 seconds, the Buk-M 3 launcher fired four 9M317M missiles into the air.
The missiles left their launch tubes on columns of white flame, accelerated violently to Mach 3.
8, then turned toward the signals that had just appeared.
At short range inside the fortress belt, they needed only a few dozen seconds to reach the interception zone.
But the four lead FP-2S had already completed their role.
Immediately after transmitting, they separated from the main formation.
Two pitched down and accelerated as if diving.
The other two turned 60° left, climbed to 950 ft, and continued emitting intermittent pulses like beacons deliberately exposing their own positions.
On the screens of Borisoglebsk 2 and the Buk-M3 crews, they became the clearest targets in the area.
Four bright signals, separated from the formation, were pulling every eye of Russian air defense in one direction.
Four 9M3 17M missiles raced into the interception zone.
Their proximity fuses found only the four targets that were intentionally exposing themselves.
Four proximity warheads detonated in the air almost one after another, scattering fragments across the hillsides, mining waste fields, and open ground below.
The four lead FP-02s disappeared from the formation, but during that time the remaining 17 FP-02s had received what they needed.
The coordinates had been corrected.
The inertial error had been pulled back to an acceptable level.
The formation had been resynchronized.
And more importantly, the attention of the Russian air defense layer had been pulled away from the real flight corridor.
But to stay safe, the Ukrainian control crew separated two more UAVs at the rear, sending them off toward the northeast until the next 29M96E missiles locked onto them.
Two flashes of fire lit up the low sky.
Two more FP-02s were removed from the formation.
After Borisoglebsk 2, the FP-02s thought they could glide through the electronic darkness, but they did not know that 16 mi away another modern system had been tracking them for the past 7 minutes.
It was Penicillin 1B75, an acoustic, seismic, and infrared reconnaissance system designed to detect artillery, rockets, and fire sources through explosions, ground vibrations, and heat signatures.
But in the low airspace around Mariupol, Penicillin was being used for a different mission, hunting small, low-flying targets that were too quiet for air defense radar to lock onto with certainty.
At 06:33 hours local time, 15 FP-2s entered the Mariupol area like ghosts.
They dropped to an altitude of only 100 ft, reduced engine power to save fuel, and held a steady speed of 100 mph.
They’d already passed through the air defense system Russia had carefully built across the fiercely contested Donetsk region.
That was why the first launch was so shocking.
A 57E6E missile suddenly left a Pantsir-S1 launcher from a concealed position behind the industrial edge.
A streak of fire rose from the ground and raced straight toward the UAV formation.
Within seconds, the drones understood the most terrifying truth.
Somehow, the Russians still knew where they were.
Penicillin had heard them earlier and sent the data back to the Russian air defense command post.
Penicillin-1B75 represented a completely different method of detection.
It listened to the ground, the air, and heat traces.
Seismic sensors recorded tiny vibrations created as engine noise spread through the ground and metal structures.
Acoustic sensor clusters captured the steady hum of piston engines.
Infrared cameras detected the thin heat trail created by internal combustion engines as fuel burned inside the cylinders.
15 FP-2s were flying low through the industrial zone of Mariupol.
Each one leaving only a very small trace, a weak hum, a faint heat signature, an uneven vibration inside the industrial sound background.
But their formation had a pattern.
Spaced a few hundred yards apart, same speed, same flight direction, same unusually low altitude.
To conventional radar, it might have looked like clutter.
To Penicillin, it was a movement pattern that could be calculated.
After 7 minutes of tracking, the system had enough data.
Speed, around 100 mph.
Altitude, around 100 ft.
Direction of movement, stable.
Heat signature, repeating with the rhythm of the engine.
The data was transferred to the Pantsir S1 battery waiting in the rear line.
The Pantsir S1 commander had two choices.
Keep the 1RS11E search radar active longer to confirm the target himself, but risk exposing his position, or trust the target queue from Penicillin, activate the 1RS21E fire control radar for the shortest possible time, lock the target area, and launch immediately.
He chose the second option.
In a brief 8 seconds, the Pantsir S1’s 1RS21E radar switched on, received the coordinates, locked the target zone, and fired.
The 57E6E missile accelerated to Mach 3 through the low air, and raced toward the pre-calculated interception point.
The FP-02s were still flying in silence, but this time, silence no longer fully protected them.
They had passed through the systems that could see the farthest, but in Mariupol, the thing hunting them did not need to see far.
It only needed to hear the right sound, see the right heat trace, and wait for the right moment.
30 seconds later, the next 57E6 missile left the Pantsir S1 launcher.
The missile accelerated to Mach 3 through the low air, racing toward the interception point that Penicillin had already supplied with data and position.
During the first 10 seconds, it left almost no clear warning for the FP-02 formation.
There was only a streak of fire closing the distance at terrifying speed.
During those 10 seconds, the 15 UAVs kept flying in electronic silence.
Their warning receivers showed nothing clear enough to trigger a reaction.
The first warning did not come from the screen, but from sound, the roar of the missile growing louder behind them.
Then, when it was only 1 mile from the predicted interception point, Pantsir-S1 RS2 1E fire control radar and electro-optical system completed the final lock.
They had only 2 seconds.
At 100 mph, 2 seconds was only enough for a desperate course change.
The operator of the lead UAV immediately pulled the aircraft away from [music] its flight path.
The engine surged.
The composite frame shuddered under overload.
The wings bent and the airframe screamed through the low atmosphere.
Just a little more force and it could have snapped apart.
The proximity fuse activated at a distance of 60 ft.
The warhead had been designed to tear apart targets much larger than an FP-2.
A cloud of fragments opened like a steel net in the air.
The lead UAV was ripped out of the formation in a brief flash of fire, but its sacrifice warned the others.
The next missile was launched 4 seconds later, but by then the FP-2 formation had changed its flight pattern.
They spread wider and no longer kept equal spacing, constantly changing altitude, sometimes dropping to just 50 ft above factory roofs and slopes, sometimes rising to 250 ft out of the thermal clutter, sometimes reducing throttle to 80% to lower engine noise, then surging to 150% power to escape the interception point.
For Penicillin, the once clear movement pattern began to fall apart.
The engine noise was no longer steady.
The heat trace no longer moved in a straight line.
Acoustic, seismic, and infrared points were stretched in different directions across the industrial noise background of Donetsk.
Pantsir-S1 kept firing.
157E6E entered its final lock too early after the FP-2 formation had already scattered away from the original predicted point.
Just a few extra seconds of warning were enough to change the result.
14 FP-2s immediately dropped lower, almost hugging metal rooftops and the edges of abandoned structures.
At that altitude, the ground became a massive wall of clutter.
Real targets, heat shadows, metal reflections, and terrain background mixed together.
The missile could not hold a clean solution.
It detonated above the formation, scattering fragments into empty space instead of tearing into a UAV body.
But, one FP2 fell behind after the chaotic throttle changes.
It needed 5 seconds to regain balance.
A 57E6E caught up from behind, the explosion flashed low against the Donetsk sky, and one more FP2 disappeared from the formation.
With only 12 miles left, the 13 remaining FP2s needed less than 6 minutes to reach Amvrosiivka.
The warning was immediately transmitted to the defensive line around the ammunition depot.
13 targets were racing in from the northwest, below 150 ft at 100 mph.
The Russian commander understood that missiles were no longer fast enough.
Pantsir-S1 had shot down two aircraft, but the Ukrainian formation was still pushing forward.
If they waited for radar to lock again, the UAVs could already be closing in on the target.
He immediately ordered the deployment of 15 suicide FPVs.
Those FPVs had only one mission: ram straight into the FP2s.
With auxiliary wings to increase lift and a speed of 130 mph, all they needed was to hit a wing, fuselage, or propeller assembly to pull a UAV out of the sky.
The first three FPVs cut across the Ukrainian formation from the flank.
The first slammed into the left wing of an FP2, the second struck the rear propeller assembly.
In less than 40 seconds, two FP2s were dragged out of formation and fell into the industrial area below.
But, the Ukrainian control crew immediately recognized the problem.
Suicide FPVs were dangerous at close range, but they were limited by battery life, control signal, and operating radius.
If the formation kept flying straight, they would enter the optimal interception zone.
But if they pulled the flight path away and forced the Russian FPVs to chase too far, the advantage would shift.
The FP2S immediately stretched the formation into a wide arc along the edge of the industrial zone.
On Russian screens, it looked like a chaotic evasive maneuver.
In reality, it was a trap designed to burn out the hunter’s batteries.
Nine suicide FPVs were dragged out of their favorable control zone.
Their batteries dropped fast.
Their signals weakened between metal rooftops, warehouses, and uneven terrain.
One by one, they began losing altitude.
One crashed into an open lot.
One fell onto a warehouse roof.
Others missed into empty space as the FP2S changed direction at the final second.
But the last FP2 inches, the formation was still caught.
It had fallen behind after the wide turn and needed a few seconds to regain speed.
One FPV with enough battery left drove straight into its tail section.
The collision sent the FP2 out of control and it crashed beyond the edge of the industrial zone.
The five remaining FPVs were forced to return to base.
Russia had shot down one more target, but its FPV interception ring had been almost completely burned out.
Now only 10 FP2S remain.
And that was when the Russians played their final card.
The 2S38 Derivatsiya PVO assigned to protect the Ambroziivka ammunition depot.
Built on a BMP-3 chassis, this system carried a 57-mm 2A90 automatic cannon with a rate of fire of up to 120 rounds per minute.
It did not need to chase the UAVs.
It only needed to build a wall of fragments in front of their flight path.
When the FP2S were 6 miles from the target, the 2S38 turret began to rotate.
The 57 mm cannon opened fire.
Bursts of programmable airburst ammunition detonated in front of the Ukrainian formation, creating a steel curtain at an altitude of 150 ft.
One FP2 flew straight into the blast zone, and its left wing was torn away.
The second dropped [music] lower to avoid the next burst, but fragments swept into its engine assembly.
Both disappeared from the formation within seconds.
But, the price of that success was time.
The distance was now too short.
The eight remaining FP2S had passed the suicide FPVs, broken through the first 57 mm fire, and were lowering their noses into the final attack angle.
The Russian commander watched the target icons close in on the Ambroziivka perimeter.
There was no longer enough time to reach deploy.
There was no longer enough distance to build another interception layer.
Only the roar of internal combustion engines remained, growling close to the ground, and eight black shadows diving toward the ammunition depot that Moscow had just reinforced with another 1,200 tons of weapons and ammunition.
All he could do was shout into the radio, “Call the fire brigade now.
” Finally, at 06:48 hours local time, the eight remaining FP2S saw Ambroziivka.
They no longer needed GPS.
They no longer needed radio communication.
The optical camera in the nose began locking onto the central ammunition depot, the place where Russia had just brought in 1,200 tons of ammunition, missiles, and military supplies to reinforce the Donetsk front.
At an altitude of 120 ft, the eight UAVs skimmed along the the edge, passing over low warehouse roofs, internal transport roads, and clusters of military vehicles parked around the base.
The first FP-2 lowered its nose.
With its heavy 230-lb warhead, it drove straight into the main warehouse at a low attack angle.
The impact tore through the steel roof, then the warhead detonated inside the storage area.
The 21-gigapascal blast pressure surged through rows of ammunition crates, throwing steel doors, sheet metal roofing, and sections of concrete wall outward.
Only seconds later, the second aircraft struck the exact area that had just been ripped open.
This time, the explosion did not only destroy the warehouse structure, it reached the ammunition stockpile inside.
Then the chain reaction began.
Artillery Artillery shells, missiles, and military supplies detonated one after another.
Fireballs erupted from inside the depot, then blasted upward into the missile missile A 24-lb per square inch shockwave swept through the auxiliary logistics area, shattering windows, blowing warehouse roofs apart, and throwing military transport vehicles out of position.
This was no longer a single strike.
This was chain destruction.
The third FP-2 dived into the loading area, where Russian soldiers were guarding and moving ammunition.
The explosion drowned the area in fire, destroying military trucks, auxiliary fuel containers, and supply crates waiting to be distributed to the front line.
The fourth aircraft struck the row of auxiliary warehouses beside it.
The warheads inside began detonating in short bursts, creating a long sequence of rapid explosions.
Black smoke rose into a massive column, visible from many kilometers away.
Then the largest explosion came.
One FP-2 punched into the center of the ammunition depot, where the ammunition was packed most densely.
Its internal warhead detonated, triggering a series of secondary explosions.
Within 7 seconds, the entire warehouse complex was torn open from the inside.
Fire, smoke, and metal fragments shot across the base yard.
Transport vehicles, lifting equipment, and the surrounding logistics infrastructure were burned out or heavily damaged.
The final three FP-2s drove into the remaining points of the logistics area.
One struck the vehicle zone, one struck the auxiliary supply warehouse.
The last one chose the most intact section of the main warehouse cluster, then plunged into the sea of fire.
The final strike caused another section of the warehouse structure to collapse, but the most dangerous part was not the first explosion.
It was what happened afterward.
Ammunition continued exploding for hours.
Fireballs kept erupting through the night.
Each secondary blast threw more metal [music] fragments, artillery shells, and burning supplies across the area.
The entire Amvrosiyivka ammunition depot became a zone of fire, black smoke, and continuous explosions.
The damage was not only in the ammunition destroyed.
The garrison forces, guards, and loading crews on duty at the base also suffered heavy casualties.
The warehouse system, transport vehicles, and auxiliary logistics infrastructure were almost completely paralyzed.
Tactically, Russia lost a large amount of artillery shells, missiles, and military supplies that had just been brought toward the front.
But strategically, the blow was far more painful.
Amvrosiyivka is an important logistics transfer hub, connecting the supply flow from the Russian border into the Donbas battlefield.
When this ammunition depot went up in flames, Russia’s fire pressure on Ukrainian defensive lines in Donetsk was immediately affected in the short term.
The strike also did not happen in isolation.
It was part of a coordinated series of attacks aimed at paralyzing Russia’s logistics system from the Ichalk railway bridge in Crimea to the Yaroslavl oil refinery to the Titan Barricady complex.
Each target was one link in the chain and Ambrosiivka was the ammunition link.
The 8FP2S did not only hit an ammunition depot, they struck Russia’s ability to sustain firepower in Donetsk.
And more importantly, they proved something Moscow does not want to admit.
Even inside territory protected by S-350, Borisoglebsk 2, Buk M3, Pantsir S1, suicide FPVs, and 57-mm anti-aircraft guns, Ukraine can still find the gap, pierce through each defensive layer, and hit the place Russia believes is safe.
If you found this analysis useful, hit like, subscribe to the channel, and leave a comment.
How do you think low-altitude UAV tactics will continue to change the battlefield? Goodbye and see you in the next video.
Disclaimer : This content may be created by AI for entertainment purposes. Any resemblance to real persons, events, or places is coincidental.
