Hacker News Reader

Comment by ajross | Hacker News Reader

> I see this changing soon. We will soon have small cheap drones that will take down nearby spotter drones and the balance will change yet again.

That's hard to imagine. You can do the spotting job with a $400 consumer device. No piece of autonomous weapons delivery hardware in the modern world exists with a price tag less than 100x that amount. Drones are simply too cheap.

If there's a reasonable countermeasure, look to ECM (though frankly modern broadband wifi is remarkably resistant there too, given all the "jamming" present in modern coffee establishments!) and not drone killers.

View on HN

What I've imagined for the past decade or so, are spotter drones that are about the same size as the drones they are design to shoot down. These spotter drones could be built much like WW1/WW2 aircraft, just much smaller, with top speeds between 50-200knots for the smaller versions.

As a weapon, they could carry weapons similar to submachineguns, shotguns or light marchine guns, depending on the size of the target. Add an AI of about the same level of sophistication as the "autopilots" in cars, and they should be able to identify and shoot down hostile drones relatively cost efficiently. Produced at scale, the smallest versions (designed to shoot down comercially available drones, for instance), would not have to cost all that much more than the drones they are designed to shoot down, if produced at scale.

I'm imagining at least 3-4 size categories, lets say. 1) Tiny, able to shoot down most quad copters. Top speed of 50-100knots. Price needs to be low enough to allow tens of thousands to be produced. Could potentially also double as anti-infantry weapons by strafing. 2) Small. Able to shoot down switchblade-sized drones. Top speed of maybe 200 knots. Can be produced in numbers of thousands. 3) Medium. Sutable to hunt small UAV's, such as Buyraktar. Top speed of several hundered knots, maybe just below Mach 1. Somewhat more expensive than the Buyraktar, but can also carry some of the same ordanance, like a tiny fighter-bomber. 4) Full-sized, possibly supersonic. Basically the same as the Loyal Wingman concept. Designed to counter other full-sized drones as well as manned aircraft.

Basically, the removal of a pilot, allows aircraft to span a wide range of sizes, and I would expect all sizes to be produced, as it is uneconomical to hunt quadcopters with full sized aircraft, while it is impossible to hunt supersonic aircraft with switchblade type aircraft.

In a way, this would be analogous to naval vessels, with different sizes serving different roles, from corvettes, MTB's and attacks subs, through frigates and cruisers, to WW2 style batlleships or carriers.

Now, for the AI researchers out there, imagine the careers that can be built around providing AI's for these vehicles. To see your own AI locked in a WW2 style dogfight with a similar drone from another country, in a real-time feed.....

Human drone pilots are cheap and reasonably easy to train. The "AI" that you need is the targeting and guidance system. If I can see any enemy drone on a camera, all I need to do is get a lock and fire.

The problem is that precision guided munitions are fairly expensive. Drones are cheap, but giving them any kind of offensive ability increases the cost significantly. A switchblade drone only costs about $6,000 and there still aren't swarms of thousands of them. Maybe once the economics enable massing hundreds of thousands of drones, autonomous control will become more reasonable.

> The problem is that precision guided munitions are fairly expensive. Drones are cheap, but giving them any kind of offensive ability increases the cost significantly. A switchblade drone only costs about $6,000 and there still aren't swarms of thousands of them. Maybe once the economics enable massing hundreds of thousands of drones, autonomous control will become more reasonable.

Exactly this. How many drones can you get for the price of 100 F35's?

If you get that many drones, you want to have drones that can attack both enemy land units AND enemy drones.

This is just silly and unrealistic. You have completely neglected detection and targeting issues. What types of sensors would be needed? What field of view do they have? How much power do they consume? How much does that sensor suite cost?

I answered a lot of this in another subthread.

Here I will just provide a couple of links to a couple of anti-air drone systems under development:

> But concepts that would fall under this class are under development. One relativly large example is Baykar Bayraktar Kızılelma:

https://en.wikipedia.org/wiki/Baykar_Bayraktar_K%C4%B1z%C4%B...

> This one is a hybrid drone, capable of both anti-air and anti-ground missions. It would fall under what I would consider "medium" size.

> Here is an early model of a Russian attempt: https://www.youtube.com/watch?v=UcAqF5BVLU4

You haven't actually answered anything. Sure it's theoretically possible to build a large, fast UCAV with advanced sensors. But it will be nearly as expensive as current manned fighters, and still not very useful at countering small, low-altitude drones.

Which is why I (a few posts up) argued that several size categories of drones are needed.

To counter the the small, low altitude drones, I argue for similarly small drones to be used. To counter a drone that costs $500-$5000 to make, one should build drones that cost $1000-$10000 (possibly a bit more expensive when adjusting for R&D and military inefficiency). These will not carry more advanced sensors than what you see on a modern car, primarily a few optical cameras.

Since one can built A LOT of such small drones (10s or 100s of thousand, if needed), they can be spread out along a frontline. They would be carried in a manner similar to switchblade drones or mounted on vehicles, and take off when a threat is detected in the vicinity. A drone (or a small swarm) would search for the enemy drone using primarily optical (including IR), and try to shoot it down WW1/WW2 style.

And, like manned fighters, they would not be restricted to purely defensive missions. These drones could also double as recon craft, do strafing against infantry or even carry small grenades or bombs.

Since these would be recoverable, it would be fine to use a $10000 drone to shoot down a $500 drone, instead of using a stinger missile.

Larger drones would be available in smaller numbers, but those would primarily be used vs larger drones and manned aircraft.

Optical and IR sensors are not adequate for the stated mission. The prices you stated are completely disconnected from reality. You're just making things up.

A Switchblade 300 costs $6000. Swap out the payload with a small gun as well as a better computer (to enable better AI).

Maybe make the wings a bit wider, to improve manouverability. The price should still be below $10000.

Provide one such unit per 20 infantrymen + a couple of units per artillery squad and other units that are particularily vulnerable, logistic vehicles, etc.

Right now they're shooting down small drones with $100k manpads. I'm not convinced drones are sensitive to ECM.

Russia said they're shooting down drones with laser weapons, which sounds kinda reasonable. Probably fairly quick to aim, and definitely very quick to deliver.

Colourless, instant laser is the Star Wars we deserve, not the Star Wars we wanted, though ;)

>Russia said they're shooting down drones with laser weapons

I tend not to believe this. In part because Russia is very comfortable with shameless lying, and in part because it's really hard to deliver enough energy with a laser to a moving target. That said, non-moving (hovering) drones would be very reasonable targets as you can illuminate for as long as it takes for an important part to melt.

BTW my favorite anti-drone tech are trained hunting hawks. Hawks are a weapon for a more...civilized time.

>it's really hard to deliver enough energy with a laser to a moving target.

Israel demonstrated that capability earlier this year. https://news.yahoo.com/video-shows-israel-testing-iron-13004...

Sounds like a fantastic way to rapidly maim a hawk.

I think the goal with this method could be to train the hawk to carry an object such as a net or dangling string that the hawk would drop into the propellers of the drone, thus disabling it.

One could also use a hawk as a platform to carry electronic means of disabling drones.

The problem with hawks and drones becomes one of teaching the hawk to recognize enemy drones since it is likely in future wars that all sides will use drones and that it may be difficult to identify who controls a particular drone without monitoring that drone to map its communication path.

It's cool to use a bird that instinctively hunts other birds but we need to understand that hawks are not hard to see and to eliminate so their utility on a battlefield is limited.

I upvoted you because you may be right. I don't really know if people are IRL training hawks for this purpose, but if they did maybe they could give them protection of some kind, like articulated metal "shoes". It's also possible that smaller drones' propellers just don't have the kinetic energy to damage a bird in any meaningful way. It's all speculation, but I still think it would be very cool.

Weaponised hawks have many interesting properties. They don't emit radio waves, and not much infrared either, so are hard to locate and shoot down. They can loiter forever (well, limited by hawk's lifetime, whatever it is), are self-refueling, and, to some extent, also self-repairing and self-replicating.

Hawks also can't be suborned.

The problem with laser weapons so far is that they have insufficient power and are fighting the inverse square law.

Heating to create sufficient damage takes multiple seconds in good conditions. They work great in field tests where the operator can keep the target aligned exactly for a while and the target operator isn't evading. Add mirror finishes, thermal shielding, automated detection and evasion, much of which is cheap, and that'll stay a demonstrator weapon for a long time.

I don't think anyone is worried about the claimed Russian lasers against drones (particularly since Russia tends to tout a lot of vaporware), but I'd be worried that since Russia is already happy to systematically commit and even advertise their war crimes, that they'd start using the lasers as blinding weapons despite the fact that it is also a war crime.

Mostly the use of laser blinders is uncommon because a laser which can blind a person will also happily do that to you from a reflective surface or down a military sighting optic.

Good point - tho if you are wearing protective optical gear and you can expect the enemy is not wearing it...

Lasers also suffer from easily deployed countermeasures. Anyone who has been passed by some truck rolling coal can attest to how trivial it is for anything running a combustion engine to be modified to foul the air around it to the point of zero visibility. In addition, flying projectiles like missiles, mortars, or ICBMs can simply spin such that the laser has to heat the whole circumference of the body rather than a single point in order to be effective, which they may be doing anyway for stabilization

Laser weapons of this sort aren't manually operated. They automatically track the target while they are firing, and they are normally pulsed to allow higher instantaneous powers, which make mirror finishes useless (the pulsed power is high enough that the amount of power that isn't reflected is enough to heat the surface of the mirror finish enough to make it non-reflective). The target also becomes very bright while it's being engaged, making it much easier for electro-optical systems to track.

The multiple seconds to kill is dependent on the laser power, the higher the power, the longer it takes to kill. The LaWS tested by the USN was 30kW or so, but they're looking at lasers with 10x the power for operational use. So the 2 seconds to down a UAV turns into 2/10s of a second, which is probably beyond reaction time, even for automated detection and evasion.

Yes, I understand that much or all of the tracking is automated. That does not negate all countermeasures. If the laser is powerful enough to punch through a mirrored surface, then one can use a sacrificial surface, a scattering surface, heat shield, emission of a scattering fog/smoke/chaff, etc., etc., etc.

I've watched the successful tests, and I'm not impressed. sure, it can shoot down a black-painted sitting duck with no countermeasures in several seconds. Merely starting to do twisting and spinning aerobatic maneuvers would have foiled it by constantly presenting new surfaces such that no tracking would work (can't make the laser point at the same spot from the other side).

The point is that until lasers get to insane powers not yet available in any portable form, it is not at all a slam-dunk. This stuff was funded decades ago for the "Star Wars" anti-missile initiative. It could use all power in fixed positions on the ground, so as big, heavy, and sprawling as you want, and the target missiles couldn't significantly evade, and it was an utter failure. Sure, we've since seen orders of magnitude increases in strength, but we're still orders of magnitude away.

Scattering surfaces aren't going to be any better against lasers, they'll heat up and become black bodies relatively quickly. Neither will emitting smoke or fog, because we're talking about airborne items, and they won't stay surrounding the aircraft very well. Sacrificial or heat shields will help, but that's extra weight that reduces your payload, and for quadcopter type drones, you can't reasonably add those to your props.

SDI wasn't successful, no, because they didn't have the capability to make high power electrical lasers worth a damn back then. But the semi-operational AN/SEQ-1 was 30kW a decade ago, and technology has improved since then. Again, we're looking at an order of magnitude more power. And I don't think we need many more orders of magnitude to get down below the threshold of : can't do anything in the time you've got.

Just shoot some stringy gel into the propellers. I'm sure a material exists that can both tangle the propellers or block the inlets of jets which would make it effective against a variety of types.

You have to get the stringy gel to the propellers, which is most of the cost of a MANPAD. The warhead doesn't cost all that much, whether it's explosive or silly string.

>Yossarian sidled up drunkenly to Colonel Korn at the officers' club one night to kid with him about the new Lepage gun that the Germans had moved in. "What Lepage gun?" Colonel Korn inquired with curiosity. "The new three hundred and forty four millimeter Lepage glue gun," Yossarian answered. "It glues a whole formation of planes together in mid-air."

-Catch-22

Thanks for this. Catch-22 is my all-time favorite war related book. There is so much in there to take in and enjoy. Absurdity, off-color humor, puns, social commentary. Heller was magnificent in this book.

Yeah you kinda expect a swarm of drones like a small ocean fleet. With smaller drones to protect the main cruiser.

Can manpads lock on 400$ DJI style drones ?

I purposely left that detail out because I dont know. They have been hitting those junk Orlans which is what I meant. I guess you can bring down the mini drones with a rifle? Like OP says there are a bunch of radio jamming guns around which probably confuse DJI.

This post says they downed 65 "targets" in one day so I'm guessing they're the cheap drones. https://t.me/LastBP/9398

> No piece of autonomous weapons delivery hardware in the modern world exists with a price tag less than 100x that amount. Drones are simply too cheap.

I recently found a video at https://twitter.com/FortemTech/status/1526925732122968067 (source: https://twitter.com/TomReiner4/status/1528214932461846528) which shows an anti-drone drone. Unfortunately, I couldn't find its price tag, but looking at the images in the manufacturer's site, it doesn't look like it would be much more expensive to manufacture than an equivalent-sized DJI drone.

for quadrocopters, you want a hunter-killer drone with some kind of a net or tape to jam the propeller.

for plane-like drones... guess something similar. gliders may be super hard to spot and will still be somewhat effective without a working propeller, so tough luck.

Net or tape? In a war? Why not suicide your drone into the target drone? If you want something reusable or longer range, a shotgun-like apparatus.

Gliders will be easier to spot because they won't be able to actively navigate through trees for cover the way the way a more nimble, powered drone could.

Ramming seems the most practical and cost effective solution.

Have you ever tried to catch a sparrow with a drine? Imagine how hard it is to hit a target that is actively avoiding you

A suicidal drone is a guided missile. There are many options on the market for these already. They’re universally way more expensive than necessary to combat a cheap swarm. You want something that won’t bankrupt you while you’re mounting a defense.

A good old fashioned flak cannon would do the trick, except now with modern targeting systems

Reply on news.ycombinator.com