Why aren't swarms of cheap consumer-level drones a viable anti-aircraft tactic?

Question

In the Russia/Ukraine war, when it comes to cheap, consumer-level drones, I've only heard of them used for reconnaisance, not for anti-aircraft defence. But given that:

• In civil aviation, a single rogue drone is often sufficient to shut down an airport due to the danger it poses
• Bird strikes are very dangerous, and presumably a drone strike is no different
• Large swarms of drones can be relatively easily controlled by modern software to form coordinated patterns
• The drones are cheap (in the hundreds of dollars each), and even if one drone scores a strike the rest can still be reused
• The drones can carry other payloads, e.g. small explosives, always-on cheap consumer laser pointers, liquid gallium to weaken the airframe if they score a hit, etc

... I would expect swarms of cheap drones to be a very effective way of interfering with invading aircraft, if not planes then at least helicopters. Yet one never hears of such tactics being used - why?

Answer 1

If a drone has a one-in-five-thousand chance of bringing down an aircraft, that makes it a serious danger on a civilian airport, but useless as a weapon for attacking military planes. And if you want to increase your chances by flying a thousand of them, then you need to think about how to control such a swarm, which is going to cost a lot more than the price of the cheap consumer drones.

Answer 2

Great. But what advantage does this offer compared to just designing a missile with an expanding rod fragmentation warhead to take out an aircraft?

While your comments are true that small drones have caused damage or even total destruction of a manned aircraft, keep in mind these accidents generally occur at low altitudes, typically less than 400 feet (130 m) in the vicinity of airports and the collisions are pretty rare as compared with the total number of flights which end safely every day.

Drones also have an absurdly limited operating envelope, as a compare to guided anti aircraft missile, which means aircraft can simply climb above an area if there is a perceived threat of collision with these things.

Typically anti-aircraft defenses employ a large screen of shrapnel emitted in the vicinity of an aircraft in order to destroy it. It is much cheaper to design a warhead that way, than build a swarm of drones to try and do the same thing.

Answer 3

Payload. Inexpensive quadrotors have a very limited carrying capacity. I have a HolyStone HS720E. ~$300. A measured 200 grams lifting capacity, and that with reduced control in flight, and much reduced range and speed.

Coordination. Getting several of them up into the path of something is non-trivial.

Still new. Eventually this may be a thing. But not just yet.

(added) Altitude and range. Go above 300-400 feet, precise control and positioning becomes a LOT harder. You cannot see your own little UAV. The aircraft you're trying to bring down is only at that level near his runway/airbase. So you need to be near that as well. On his land.

Answer 4

Consumer drones are small and very slow, so the only viable way to use them for anti-aircraft purposes would be to have a swarm of them covering a certain area, thick enough that an airplane would fly through the cloud and be damaged by the impacts. Military aircraft are built tough so there's no guarantee this would work, but let's assume it would.

The issue is that you need a lot of drones to make that work, and I mean a whole lot. A Stinger missile has an effective range of close to 4km (allegedly, it's probably less than that). Doing a Fermi approximation, let's assume you want to replicate the coverage of a Stinger, so a radius of 3000m, assuming a spacing of 1m between drones. The area of half a sphere is 2πr², plugging in a radius of 3000m you would need about 56,550,000 drones. Say you could make them explode when an airplane comes close and only need one per every 4 sq meters, then you only need 14 million drones. Say they cost \$300 per drone, that's \$4.2 billion to cover that area, as opposed to a Stinger missile which costs $38,000. Even if my figures are an order of magnitude off it's still an incredibly expensive system. Even if it's just hundreds of thousands there's still a huge logistical challenge to transport, deploy and maintain such a system.

Answer 5

Possibly - but you'll need to pull the enemy down to the effective altitude of cheap drones. They are already doing that.

Deny the high sky, part 1: high altitude.

The preferable method is to use long-range, long-reach, very large SAM systems with 200-300km range, well in the rear. The extant systems of this type are huge, "4 missiles per semitrailer + separate RADAR trucks + separate command post" type deals; the missiles are 0.8m diameter and 10m long, basically flying telephone poles. This includes the Soviet S-300 and successor S-400 systems (no NATO equivalent), and warship-based S-300 and AEGIS/SM-2 system. These control the high sky, but cannot control the middle and lower sky because they are positioned well in the rear, and lower altitudes are below the system's horizon.

In the Ukraine theater, Ukraine and Russia both possess the S-300 truck mounted system, Russia has the evolved S-400, and there are no warships with such systems anymore.

The less preferable method is to exhaust your adversary's supply of precision guided bombs and missiles by making it necessary, either by absorbing the hits to valuable infrastructure, or by using disinformation tactics to make them waste precious precision-guided weapons on bogus targets. I'm figuring on quite a lot of the latter, given Ukraine's demonstrated history of doing a lot with a little. Once they are down to World War II tier dumb bombs, to hit a given target their bombers must come well below 20,000 feet (6000m).

Deny the high sky, part 2: medium altitude.

Next, the objective is to deny the medium altitudes below 20,000 feet (6000m). This can be done several ways. First, use medium SAMs such as Buk, Tor or Pantsir. By putting everything - RADAR, command and missiles on the same truck, there's a downside in missile size, which limits range to medium - but a massive gain in agility, so they can "keep moving" to avoid counterattack. As such, they can be much farther forward with less of a horizon.

Second, use MANPADs like Stinger or SA-7 which have far shorter range (under 10,000'/3000m altitude) but are yet more portable, and more importantly use infrared seekers, so anti-RADAR missiles don't work on them, and it won't light up the threat-warning receiver on aircraft. The aircrafts' only warning is seeing the puff of smoke. And flares don't work if you don't deploy them. The MANPAD shooter has an effective field-of-view of "the tree line", buildings or other obstacles, so their access to the sky is somewhat less than 180 degrees. As such, the aircraft flight path is in that arc only for a limited time. The higher the aircraft flies, the longer it is in that arc and thus subject to attack. This amount of time matters. because MANPAD shooters take a moment to grab the weapon and get ready.

An aircraft operating in these zones is now denied the lower "good bombing" altitudes of 300-20,000 feet (100-6000m). So to manage both threats, the aircraft must fly very low. It defeats the medium range SAMs by dropping into the RADAR "ground clutter", literally below the RADAR. It defeats MANPADS by shortening the time it is in the shooter's field of view to where there is not time to alert, get to the MANPAD, shoulder up, power up, aim and shoot. But, this has the effect of denying the enemy medium and low altitudes, and forces them to hug the tree line. If you don't want to die, you better like the taste of leaves in your teeth.

In the Ukraine theater, everybody has these systems. On March 29, 2022, Ukraine claims not to have held a clinic in Belogrod on penetrating a modern, networked SAM system.

To see these systems at play in simulation, with discussion, check out the "Grim Reapers team" video on Youtube where they attempted to "fact check" the Belgorod raid. The conclusion that if you fly "not perfect" the SAMs get you, but a more experienced player (who really liked the taste of leaves) made it in and out (even overflew a SAM site too low and close for it to engage).

Let's change the meaning of "drone", though.

I know "quadcopter" is synonymous with "drone". Forget that. Why?

• Any 'copter type aircraft is hideously fuel-inefficient compared to a fixed-wing because it's "suspending with thrust" instead of actually flying. For many uses, that's a worthy trade-off, i.e. the ability to hover. But the efficiency is bad - loitering surveillence aircraft would fly like gliders and probably look like gliders. Does the glider community have any rotary-wing gliders? No.
• The quad/multi copter needs many engines and thus electric drive, which locks it out of using fuel for longer duration flight. A fixed-wing can go either way. And R/C airplane sized engines are plentiful.
• Payload capacity is much higher on a fixed-wing, so a meaningful warhead or sensor package could be carried.
• Speed - which matters both for intercept, and a second thing.

The second thing is the ability to have a simple basic airframe with swappable mission modules. Some drones might carry a warhead. Others might carry a RADAR or other feature that needs a lot of electricity. Rather than complicate the basic drone design by adding a generator to some engines, do as the military does and build the module with a generator called a RAT (Ram Air Turbine). Instead of belt drive, airflow drive. Less to go wrong. But the RAT requires airspeed beyond what is possible for a quadcopter, but is easy for a fixed-wing. In fact, I would even prefer a RAT for the basic electrical needs of the drone. A RAT is just a generator attached to a prop, in fact every quadcopter comes with 4 of them lol. BLDC motors work as generators.

And they could fly high but they don't have to because the above measures have pressured the enemy into the narrow zone below 400' AGL. In fact, having the interceptors loiter high would give them the ability to dive to gain energy/speed.

Nothing against a quadcopter - they're a good fit for the task size and attention span of hobby and task-focused drone flyers. But a sustained air surveillance or combat air patrol needs much more endurance than that, so for your task I'd absolutely say fixed-wing is the way to go.

Now, network them.

The general idea I see is to have some drones tasked with surveillance and others tasked with intercept. Surveillance could be as easy as listening - using the sound of aircraft to locate them - the drones have accurate position sensing and accurate clocks, and report contacts to each other via radio. That makes it easy to triangulate targets. Digital signal processing of that sound could identify right down to model of aircraft. You can't outrun Motorola, so the intercept drones are getting a picture of the track of the enemy even while it's a long way off. They have time to position to "get in its way".

There would be technical tricks to resolve regarding sensors, geolocation (since the enemy can be counted on to mess with GPS/GLONASS), and telecommunication - you might actually have to read Hedy Lamarr's work on spread-spectrum... not using a single frequency subject to jamming or RDF. See also Qualcomm's work with CDMA cellular, or Metricom's lovely-for-the-time "Ricochet" - a wireless mesh network that did not need the backbone -- Ricochet modems could talk peer-peer. It might be as simple as 25-year-old Ricochet modems in very drone.

Another thing this network could do is dull the senses of pilots to MANPAD launches by having $20 decoy rockets all over the place designed to mimic a MANPAD launch. They would listen to the network and fake out just as the aircraft is overhead. At worst the pilots use up their flares on decoys so they're Winchester when they meet a real one. At best the pilots treat everything as a decoy and are blindsided by the real McCoy.

Once the hardware is in place, the performance of the system is mostly down to software.

So yes, possible. Very possible. Because you've reduced the battlespace to 2-D, or really a line of a particular width.

Answer 6

What looks like a simple task at first, becomes quite impossible, when you start comparing the exposed area of let's say a fighter-jet or a helicopter to the humongous volume of airspace around it. You would have to know exactly where to place your swarm in order to effectively intercept the hostile aircraft. I assume that you are referring to a mine-field like design. In that case, the same problems apply to your drones, that apply to ground based mine fields: You have to know where to place them. Just with the difference of having to deal with roads (less possible area) instead of airspace (a lot of volume).

Assuming you have soooo insanely many drones, that you could actually reliably hit hostine aircraft, you would have to equip every drone with a proximity fused IED or other explosive payload, which is heavy. It's not like you could take down an aircraft with a firecracker. Assuming a drone might hit an engine is not worth the thought. As I mentioned before, even the entire surface area of a military aircraft is insanely hard to hit - concidering the placement issue mentioned -, not to mention hitting a single engine intake (disregarding the fact that many russian military aircraft like for example the SU-27 or the KA-60 have twin-engine architectures and can deal fairly well with a single-engine-failure).

Laser-pointers... don't do any damage. They say, a laser-pointer might blind the pilot for a second, but even in civil aviation this is merely a hypothetical problem. And here, the "hitting-problem" also gets to another scale: Now you are trying to hit an aircraft with a 2mm-dia beam. "Improbable as hell" is an understatement.

And gallium, same problem. Hitting is nearly impossible. And the potential damage is minimal. It might work well under lab conditions with a lot of galium on an exposed aircraft wing, but not in action on a battlefield. The airframe of an aircraft is protected by an oxide-layer and paint. No way a very slow reaction could ever harm the aircraft (if it even stays on the surface while the craft is in motion)

All in all, one multi-ten-thousand "cash" AA-missile is way more effective at reliably taking an aircraft down, than hundreds of drones over a city (if someone could even create such a swarm in a realistic period of time, or even pay something like that. One or two drones are cheap. Hundreds of them to hit one craft... SAMs are more economic!)

Answer 7

The sky is huge with regards to volume. And planes are fast. I imagine you would need thousands upon thousands of them to be put in front of a path of a fast moving aircraft so that the plane can not just easily go around them.

But for a second, lets imagine a world where this is possible. Drones would need to be:

1. fast... able to arrive to the ever changing path of the aircraft.
2. have a long lasting battery that would let it arrive from its base to the calculated trajectory. This might be several miles.
3. stupid cheap to be able to have the thousands of them available.
4. maybe not as important if it can get into the right parts of the plane... able to carry munitions.

I suspect a better way to arrive to this might be from starting at loitering munitions. Loitering munitions can typically remain in the air patrolling. Then when they go into action, they turn into a missile of sorts. So this is a mixture of drone tech (for the loitering part), then missile tech for the final part. Maybe over time this can evolve into something that starts becoming more drone and less missile over time?

Answer 8

Consumer drones are a danger to aircraft that are landing or taking off.

It may be possible to shut down an enemy airport with a swarm of drones, if it's being used to ferry in supplies and personnel. It would be bad for morale to use such an airport without clearing the drones first. Their low ceiling prevents using them higher up.

Combat aircraft can choose their flight altitude to suit the conditions and the threat profile. In general, flying low is done for two purposes:

• Get obscured by terrain to penetrate long-range air defenses.
• Deliver precision strikes that achieve objectives with fewer flights and less collateral damage.

In asymmetric warfare, low altitudes are already dangerous, since they can be reached by low-tech AAA and man-portable weapons such as MANPADS. These are greater dangers than a risk of collision with a drone.

If necessary, the opponent can choose to fight from higher altitudes. This will decrease the precision of non-guided munitions and require more flights to achieve the same objectives. Guided munitions can be used from altitude.

Rotorcraft don't have that choice. They almost always fly at low altitude, due to how they get their lift and to their operational role. However, combat helicopters have their intakes protected, and most of them are armored.

They don't fly fast enough for a small drone collision to be very dangerous on its own, and small drones don't carry enough payload to effectively deal with the armor. MANPADS, on the other hand, and sometimes even RPGs, are quite effective against low-velocity, low-altitude targets.

In conclusion: technically possible, but the only viable targets are airports that aren't well-patrolled on the ground.

Answer 9

Let us check the specs of an example drone, DJI Air 2s:

• Top speed 68 km/h
• Max climb/descent 6m/s
• Range 18km
• Weight 595g

That's pretty good already, but in order to damage a flying aircraft, the weapon has to catch up with it. These speeds, especially the climb speed, are so low that's not going to happen unless the target is a helicopter in stationary flight. Even this would require the helicopter to stay still long enough and close enough to where the drone is. Basically, it would require the drone to be already airborne, and the aircraft to fly almost straight into it.

In addition, if the drone flies into a helicopter... I wonder if it would be able to actually touch it. I mean, there's this big fan on top of the helicopter, blowing air at pretty high speeds, more than 100km/h straight down. Since this drone is specified to operate in winds up to 38 km/h only... I think the drone would just be blown away by the rotor downwash and never hit the helicopter. So it would have to attack from above and get sucked into the rotor blades, maybe try to damage them. But since it is tiny, would probably get through without being hit at all. Okay, maybe try to get sucked in by the tail rotor instead, while tumbling in the airflow from the main rotor and the turbine exhausts...

That doesn't bode well for success.

I think this type of drones will eventually be used, but for ground targets only, with a small explosive charge. It should be possible to use computer vision to target a moving tank's sensors, or get it to fly into a building and blow up someone important, for example. This would need quite smart software, because both GPS and the radio link from the drone to its operator will be jammed, and a drone is too small and cheap for accurate inertial guidance.

Besides that, as someone said, "When you merely wish to bury bombs, there is no limit to the size." Therefore, IEDs seem a much more viable option for a low-tech military.

Answer 10

Because the usual off the shelf drones have extremely bad loiter times and you could not rely on chance to hit an airplane but would need fancy targetting algorithms (and probably feed the algorithm with data from ground based radar).

With off the shelf drones, usually you get quad copters. They have short battery lifes. And without a lot of custom software you can be absolutely sure that there is no better way to hit an aircraft than chance - maybe arranging them in curtains intersecting possible flight routes, but thats still square kilometers with at least one drone per square meter to get any useful effect.

I think it's quite a plausible idea, with a purpose-built swarm of military drones (sure as hell not n-copters, more likely small planes with the capability to ride thermals autonomously) designed to be sucked into jet engines. But then you would need to be able to anticipate an aircraft before it arrives, drawing the swarm in close around it's flight path. Most likely this means you need radar coverage of the airspace, some system to mark detected planes as targets and then send the drones a kill order. And the drones would need to have some autonomy, because it's easy to jam a signal such that low power drones have no chance to receive or send data. And the drones would need some way to head for the engine and get sucked in, because that's where you can do a lot of damage with little mass, that might imply machine learning algorithms (because the engine intake can probably not be targetted with a simple heat seeker).

So... cutting edge military technology with serious moral implications especially before there is enough experience with it to make it "safe". No way to do it with off the shelf drones.

Answer 11

I'll take "drones" in this case to mean quadcopters. Consumer quadcopters don't have very long flight times- 20 or 30 minutes maximum. The battery charge time for the type of battery common to toy- and hobby-grade quadcopters is often an hour or more. It's just not feasible to sustain a fleet large enough to be considered for an anti-aircraft swarm. Further, unless you plan on making each aircraft autonomous (consuming time, money, manpower, and supplies), each aircraft will have to be controlled via radio frequency, which is susceptible to jamming. Finally, as others have noted, the kill probability is extremely low relative to other weapons systems, giving a low yield for a highly maintenance-intensive system.

Answer 12

Generally speaking, civilian radio equipment is not up to the task of operating in a battlefield electromagnetic environment. In the U.S., at least, the FCC makes quite clear that civilian devices need to be specifically designed to be overwhelmed by law enforcement/military devices ("accept all interference" is the specific terminology I recall). Once current generation aircraft are in the area, you're extremely likely to lose contact with the swarm unless you have non-radio means of directing their activity. Pre-programmed deployment patterns, or just brute-forcing and trying to fly the swarm visually from the ground (your ground station is likely to be robust enough to break through jamming), suffer from the costs associated with complex pre-development or a general lack of acuity.

In short, these types of systems are ambush predators, and highly unlikely to be effective against aircraft in flight at any reasonable speed, since they will struggle to meaningfully climb to intercept altitudes in the kind of time frame that you are alerted to enemy aircraft vectoring towards your launch site.

Trying to use them against hostile airfields runs afoul of operational range limitations of these devices - but would otherwise be your best option for weaponizing these systems. (Bird strikes are dangerous because they happen at low altitudes, and in otherwise vulnerable flight regimes.)

Remember that MANPADs are rocket-powered, and still have very short useful intercept ranges because of the relative speeds involved.

So between the extremely low odds of successful interception and the difficulties of controlling civilian equipment in the presence of radio jamming, it's just not likely to be a fruitful avenue for a military to use against a peer-opponent.