this post was submitted on 09 Mar 2024
390 points (100.0% liked)

Technology

70365 readers
3593 users here now

This is a most excellent place for technology news and articles.


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related news or articles.
  3. Be excellent to each other!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
  9. Check for duplicates before posting, duplicates may be removed
  10. Accounts 7 days and younger will have their posts automatically removed.

Approved Bots


founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 40 points 1 year ago (4 children)

Drones currently outpace their countermeasurs. This will definitely not be a thing forever. I think the effectiveness of cheap drones will go down as be countermeasures are invented.

We already see new very effective military drone jammers starting to come out

[–] [email protected] 18 points 1 year ago (3 children)

Onboard AI guidance is not difficult.

[–] [email protected] 9 points 1 year ago (2 children)
[–] [email protected] 7 points 1 year ago* (last edited 1 year ago) (1 children)

Both of you are right.

It's difficult, but how difficult depends on the task you set. If the task is "maintain manually initiated target lock on a clearly defined object on an empty field, despite the communications link breaking for 10 seconds" -> it is "give a team of coders half a year" difficult. It's been solved before, the solution just needs re-inventing and porting to a different platform.

If it's "identify whether an object is military, whether it is frienly or hostile, consider if it's worth attacking, and attack a camouflaged target in a dense forest", then it's currently not worth trying.

[–] [email protected] 2 points 1 year ago

That was a good guess but unfortunately it is just difficult even in the scenario you proposed

[–] [email protected] 2 points 1 year ago

Realtime person detection and following it with a drone? Difficult for me, certainly, but there are enough people out there who have done it.

[–] [email protected] 9 points 1 year ago (4 children)
[–] [email protected] 15 points 1 year ago* (last edited 1 year ago) (1 children)

Tracking a moving object in realtime with video is a standard task for a machine learning engineer. You can do it on an embedded platform with ML hardware support. I don't know what hardware newer Lancets use but they can already do it according from developer reports from Telegram channels like e.g Разработчик БПЛА.

[–] [email protected] 6 points 1 year ago* (last edited 1 year ago) (2 children)

Honestly, I was just objecting to the use of "AI". We've had both fire and forget and loitering munitions for decades now, neither of which use ML. Will it happen? Sure. For now, ML/AI is too unreliable to be trusted in a deployed direct attack platform, and we dont have computing hardware powerful enough to run ML models that we can jam in a missile.

(Though yeah we run tons of models against drone data feeds, none of those are done onboard...)

[–] [email protected] 3 points 1 year ago

The point of modern deep learning approaches is that they're extremely easy on the developer skill. Decades ago realtime machine vision needed a machine vision expert, these days you throw the hardware at the problem at learning stage, and embedded devices to run the results are stupidly powerful (doesn't even take a Jetson board), if you compare to what has been available even a decade ago.

[–] [email protected] 3 points 1 year ago* (last edited 1 year ago)

For now, ML/AI is too unreliable to be trusted in a deployed direct attack platform

And probably can't ever be trusted. That "hallucinations can't ever be ruled out" result is for language models but should probably apply to vision, too. In any case researchers made cars see things and AFAIU they didn't even have to attack the model they simply confused the radar. Militaries are probably way better at that than anything that's out in the open, they've been doing ECM for ages and of course never tell anyone how any of it works.

That doesn't mean that ML can't be used, though, you can have additional non-ML mission parameters such as the drone only acquiring targets over enemy territory. Or that the AI is merely the gunner, there's still a human commander.

[–] [email protected] 6 points 1 year ago* (last edited 1 year ago) (1 children)

A combination of GPS, or even inertial based guidance to get them to the target area and then some simple vehicle / object identification, I'd think those are possible.

[–] [email protected] 6 points 1 year ago (1 children)

GPS is usually the first thing to be jammed in the battlefield.

[–] [email protected] 7 points 1 year ago (1 children)

GPS is useful, but not required for operation. Inertial guidance, and ground tracking cameras can easily maintain a good position sense, while completely RF passive. This is also already normal on many toy drones.

You would also want to jam it over a large area. That jamming is akin to a "kick me" sign, in neon lights.

[–] [email protected] 3 points 1 year ago

Inertial guidance sucks balls for any meaningful amount of time. Combining it with ground tracking gets it a lot better, if you have good time of flight sensors to measure the distance from the ground. But this also falls flat on its face when the ground is too uniform (grassland, wetland, snow etc).

[–] [email protected] 4 points 1 year ago (1 children)

You can't be very knowledgeable on the topic.

[–] [email protected] 3 points 1 year ago* (last edited 1 year ago)

I am, in fact, fairly well versed in the topic. You're 30+ years away from being able to fit hardware powerful enough to run a ML model into a missle, though I cant see a single reason you'd ever want to. Look into the declassified, 40+ year old design paradigms for missiles or other self-guided munitions and it'll start to give you an idea of why the idea of "AI" guidance is so laughably stupid. There's so very many reasons we use FPGAs, none of which are compatible with AI.

[–] [email protected] 4 points 1 year ago* (last edited 1 year ago) (1 children)

Edit, apparently I'm an idiot and my ability to tell truth from fiction is a lot worse than I thought.

In my defence however, all the parts are completely viable. I also saw it mixed in with Boston dynamics videos.

I'll leave the original comment for context of my folly.

The US already has them.

There are single shot drones, designed to be deployed into a building, or cave system. They then use cameras etc to navigate, while running face recognition. When they find their target, they fly just in front of it. The shaped C4 charge is designed to reduce their head to red mist, while not risking those close by.

AI + cheap drones will completely change warfare. Probably on the same level as the tank, or machine gun.

[–] [email protected] 9 points 1 year ago (1 children)

The US already has them.

Do you have a source for this? These sound uncannily close to the Slaughterbots short film...

[–] [email protected] 1 points 1 year ago (1 children)

I'll try and remember to dig it out later. It was a sales demo from a weapon's company. I can't remember exactly which one it was, but the implications scared the shit out of me.

[–] [email protected] 2 points 1 year ago (2 children)

Are you certain you're not thinking of the Sci-Fi Short Film "Slaughterbots"? The plot is almost exactly what you describe.

[–] [email protected] 3 points 1 year ago

Apparently I'm an idiot. I saw a modified version, mixed in amongst other legit videos.

[–] [email protected] 2 points 1 year ago

Here is an alternative Piped link(s):

Sci-Fi Short Film "Slaughterbots"

Piped is a privacy-respecting open-source alternative frontend to YouTube.

I'm open-source; check me out at GitHub.

[–] [email protected] 4 points 1 year ago* (last edited 1 year ago) (2 children)

I will be very suprised if this isn't already happening.

[–] [email protected] 1 points 1 year ago (1 children)

It's one thing detecting a person with machine learning in a test and an actual soldier with camouflage in a very imperfect environment. Also good luck telling friend from foe from civilian.

This has all sorts of problems while making the whole system more complicated and prone to issues. Not the mention moral questions of autonomous weapons. I have no doubt it will happen but not yet, not here.

load more comments (1 replies)
[–] [email protected] 1 points 1 year ago

I know it already does, at least in newer Lancets. Expect this in fpv type devices soon.

[–] [email protected] 7 points 1 year ago (5 children)

The only reliable counter to a drone is likely another drone.

I suspect Peter F Hamilton got it close, in the Confederation series, with WASPs. They are space based weapon platforms. They carry a mix of offensive and defensive subsystems, and operate with swarm logic.

I could easily see a larger drone carrying a swarm of 1 shot micro drones. When close, some would be sacrificed to get better sensor data, others would go on the attack. Conversely, a defensive target would launch their own swarm. It's goal would be to stop the attackers getting a good shot on a high value target. It might also counterattack, either against the mother ship drone, or backtracking to find the launch site.

Jamming would also be part of this. A jammer could easily cut off the swarm from external data sources. Live satellite or remote surveillance systems would be cut. Point to point lasers are far harder, as are burst transmissions. Local sensor drones could easily punch short range data back, or paint targets, until they are destroyed by defensive systems.

[–] [email protected] 6 points 1 year ago (2 children)

There are in fact a huge number of reliable counters to drones, including but not limited to anti-aircraft gun systems, anti-aircraft lasers, RF jamming devices (especially effective against cheap/makeshift drones), and several more. Drones are currently an emergent threat without a robust countermeasure scheme, but given their massive role in the Ukraine war that is not going to go unaddressed for long. From a purely mechanical standpoint, small drone munitions are also physically very vulnerable, making them readily destroyed by anti-air autocannon fire or even laser weapons if you assume RF jamming will not solve the problem.

[–] [email protected] 5 points 1 year ago

Between GPS (jammable but likely gets you into the target area), dead reckoning, optical flow sensors and increasingly impressive onboard camera processing, RF jamming will soon be irrelevant.

Almost a decade ago I was flying agricultural mapping missions that were 99% autonomous, and the parts that weren't were problems a military drone doesn't have (soft launch and landing)

The clear counter is autocannons, likely fully automated themselves to manage large swarms. The other would be cheap anti-drone missiles that either are basically a drone themselves or a glorified model rocket. Possibly tiny, cheap and fast interceptors launched from fixed-wing drones. The weak point of drones is literally their physical weakness.

[–] [email protected] 3 points 1 year ago

It's worth noting we are at the start of an arms race. It will iterate all over the place.

For example, smoke and chaff deploying drones would make defensive fire harder. Anti air can be either baited (and so depleted) or rushed. Lasers can be shielded against, at least for a time. Jamming can be countered with line of site communications.

In turn, each of these can be countered.

A key thing of note is that your solutions are heavy. Fine for defending a static target, but problematic when dealing with defending a mobile unit etc infantry of transports. In those situations an extremely rapid, focused highly dynamic response would be required. The obvious way to deploy those fast enough is to have them automated and airborne, aka a drone swarm.

I might be completely wrong, current drone warfare is akin to the invention of the smoothbore musket. How it will develop remains to be seen (for better or for worse).

[–] [email protected] 6 points 1 year ago* (last edited 1 year ago) (2 children)

https://www.youtube.com/watch?v=rr7ym1zkda8

Anti-air guns are the countermeasure. RADAR good enough to detect drones + an aimbot and programmable air-burst round to "shotgun" your pellets to damage those soft plastic bits.

We're going back to WW2 tech. AA guns were considered obsolete because Helicopters + Missiles had more range. But now we need to build cheaper AA Guns for the anti-drone role.

AA Guns are also useful vs infantry, so in an infantry vs infantry fight, having an AA Gun platform will be useful even without any drones around. Airburst and rapid fire is always useful, and I expect the computers that make RADAR possible will be far cheaper today than decades past.

[–] [email protected] 1 points 1 year ago

Here is an alternative Piped link(s):

https://www.piped.video/watch?v=rr7ym1zkda8

Piped is a privacy-respecting open-source alternative frontend to YouTube.

I'm open-source; check me out at GitHub.

[–] [email protected] 2 points 1 year ago (1 children)

It's going up be interesting and scary when we see the first mega swarm of drones, a river of them just pouring through the sky and hammering from every direction at the defenses.

Constant evolution of drone and antidrone, a production race with frontlines being slowly shifting walls of drone combat, them pouring out of factories as fast as they can be made with the front moving based on who can make more per hour

[–] [email protected] 3 points 1 year ago

I suspect it will be more subtle even if it's only battery life limited. Huge swarms will also struggle against fixed defences. More likely it will be used in ambush. E.g. air deployed near an enemy convoy, or swarming from rooftops and windows onto an infantry unit. Counter deployment will have to be seconds to stop the lead elements. Potentially with heavier reinforcements flying in.

I've personally got visions of a Boston dynamics dogbot with a harness full of drones. 1 button press and a few dozen micro drones swarm out, with larger ones launching as needed.

I could also see facial recognition drones being deployed from a predator drone, like cluster bombs. A little akin to the bots used in the film minority report. They swarm a building or block, and try and identify all the faces they can find.

The key thing however will be battery life. Multicopters are power hogs. You need around 40% battery to get maybe 5-20 minutes flight times (depending on how the manoeuvre). Longer times can be achieved , but requires larger systems with higher costs. Is 1 system with a 2 hour flight time worth 20 smaller systems only good for 10 minutes?

[–] [email protected] 1 points 1 year ago (1 children)

You use the word easily so many times here where it becomes more and more apparent that you probably don't think it means what it means

[–] [email protected] 1 points 1 year ago (4 children)

I've worked with drones of various sizes. Bigger and more expensive ones are more capable, but hard to make bullet proof. If you can remote off their sensors and weapons into cheap, more disposable systems, it makes sense.

A big drone, like a predator, drops a package into an area. Mid sized multicopters provide local computing power and coordination. Small planes provide fast loiter surveillance. Small multicopters with cameras give more accurate coverage. For attack, you have what amounts to a hand grenade with props. Protection takes the form of similar disposables. A flying strobe light to mess up optical tracking. Chaff bombs to mess up radar tracking. Smoke to obscure the high value units.

A lot of these I could throw together myself, given a few weeks, and a few grand. What part wouldn't be easy, for a large and well funded military r&d team?

load more comments (4 replies)
[–] [email protected] 4 points 1 year ago

I saw a YouTube clip explaining that they use a secondary drone to boost the signal to get around some of the jamming.

The game of cat and mouse continues.

[–] [email protected] 1 points 1 year ago* (last edited 1 year ago) (1 children)

Jammers only work against remote controlled drones. Autonomous ones have no such issue. And jammers are never a problem against civilians, which tech like this will eventually be used on.

load more comments (1 replies)