Hacker News Reader

Comment by zer0x4d | Hacker News Reader

This is definitely interesting in the face of the news that Ukraine has been using very small FPV drones successfully on the front lines as opposed to more sophisticated systems like the Bayraktar drones. There seems to be a huge gap in air defense systems when it comes to small UAS and this should effectively address that.

View on HN

You have to get the costs balanced too- it can't cost 1 million dollars to kill a 250 dollar drone, for obvious "what if the adversary was China" reasons.

I don't understand why nobody has put auto-shotguns on an tripod with an aiming system, like a mini-CIWS.

I've seen Russian soldiers successfully shoot down drones using shotguns, the problem is that the smaller ones can be too fast and nimble for humans to reliably target. Even if you can shoot down 2 of 3, the third one will get you.

Combine the YOLO algorithm with a high-frame-rate camera, a two-axis servo, a heavy tripod, and a shotgun... and you're defended!

This could be mounted on the back of a tank or humvee also.

Something like you describe really does sound like an ideal terminal stage defense system, with 20 or 35mm autocannon covering longer ranges.

Purely theory crafting, I think the final piece would be small, standalone AESA radar units perhaps the size of a man-carryable generator that can be distributed across a given area and networked to likewise distributed shooters, spreading out the targets, adding redundancy to the system, and reducing the odds of your dugout-with-air-defense-asset getting hit by an anti-radiation munition that homed in on a known emitter.

The reason that you don't need radar for this is that the small suicide drones use cameras (vision) to attack. If you can't see them, they can't see you! Contrast this to general aviation where planes can be kilometers up, behind cloud-cover and can attack you with radar-guided missiles.

Radar probably won't have high enough resolution and won't be able to differentiate between drones and birds. Not to mention that most drones are carbon fibre or plastic and won't have much of a radar return to begin with.

Conversely, vision systems could be trained to ID drones reliably, and could use data such as the high-frequency "beat" of the propellers to detect drones from just a few pixels of sensor data.

Shotguns have a very limited effective range. Round pellets have a low ballistic coefficient and lose energy quickly. Using larger bore shotguns only delivers marginal range improvements. Automated shotguns might end up being one aspect of SHORAD, but based on recent developments it looks like they'll end up being less important than jammers plus autocannons (short term) and lasers (longer term).

I think many are trying something along these lines. The devil is in the details.

The shotgun needs to be reloaded. The system needs to be reasonably reliable and durable. Feels like the range of the shotgun is also a problem. The system needs to have some sort of friend vs. foe identification, possibly not shoot at birds or other random objects. The camera needs to be able to deal with looking at the sun and low light, day and night, likely needs to be able to scan. Pretty soon you get to: https://en.wikipedia.org/wiki/Phalanx_CIWS

> The shotgun needs to be reloaded.

Eventually, certainly, but you can put it off for a while without much trouble.

You can get 20 round drum magazines for 12 gauge shotguns, and keep hot-swapping 'em until your barrel overheats.

> Phalanx_CIWS

But that's 1980 technology. I bet a current GPU could outcompute it by orders of magnitude.

Range is definitely an issue with shotguns. On the other hand, you're dealing with low-speed drones, not enemy missiles.

A Phalanx_CWIS is overkill. Also, I suspected that the system received continuous upgrades throughout the decades.

Likewise, the Abrams tanks are cold war era tanks, but they are continuously upgraded that a modern Abrams tank isn't really a 1980 tank. Modernization is likely to remove some weight and make Abrams more survivable against drones, in response to the Russo-Ukrainian war;

My point was once you start adapting to all the requirements you inevitably end up with a more complex system. Another thing that happens in a war is that the enemy will adapt to what you're doing.

The principles are fine (well, not sure camera is the right only sensor but aside). The actual engineering of a working system you can use in real life is complicated and that's what many are working on.

I just came back from a walk and someone was flying a drone. Those things are fast, small, highly maneuverable.

Ukraine has been using improvised small FPV drones since Russia invaded pretty much. The Russians in contrast seem to have been using larger drones (like those supplied by Iran).

Hezbollah has been attacking Israel from Lebanon for many months now with small drones and the air defence systems like Iron Dome haven't been able to effectively deal with that. One recent example: https://www.timesofisrael.com/hezbollah-swarms-north-with-at...

There is a major effort right now in Israel to address that gap.

> The Russians in contrast seem to have been using larger drones (like those supplied by Iran).

You are confusing short-range and long-range campaigns. Both sides are using fpvs, mavics-style, short-range fixed wing and long-range winged (like shahed-136) drones

I thought the Russians relied more on military gas powered drones (like the https://en.wikipedia.org/wiki/Orlan-10 ) and were not using cheaper fpvs. Certainly in the earlier days there was a lot of footage of Ukranians dropping grenades and mortars via improvised drones and I think the Russians were more spotting with Orlans and then using artillery.

Most if not all the random videos I've seen of cheaper FPV drones were Ukraninian but that could totally be my social media bubble.

Reply on news.ycombinator.com