The drone topic

[glenn239]

17 hours ago, Rick said:

Just use the birds...https://en.wikipedia.org/wiki/The_Birds_(film)

 

No, I was wondering about something more like this, (around the 1:08 period).

 

[sunday]

Interesting. First attempts to heavier-than-air aircraft were trying to imitate birds. Seems we are going to go full circle.

[Sardaukar]

18 hours ago, bojan said:

Orlan is about 5k a pop (price for Russian army), still quite cheap. It has comercial level sensors, but so does most of the drones that Ukrainians use.

IDK what problems with Sosna sight you mean, one local tanker that was at tank biatlon described it as having a better picture on TI channel than Lecrerc and FCS being quite good.

One thing is: According to open sources, Thales supplied Catherine FC thermal imaging cameras to Russia, which were used to manufacture the Essa, Plissa and Sosna-U thermal sighting systems. 

That makes it still sanctioned foreign technology. AFAIK, Sosna-U is made in Belarus, which might add another layer of complications. 

So, it might be difficult to repair or resupply those. Plus, Catherine is not exactly latest tech.

[bojan]

Open sources as usually know general idea, but lack some subtle (or not so) differences. Originals were assembled in Russia from French parts, later % of Russian made parts increased. ATM they are making most difficult part, TI matrixes, so I don't see them being unable to produce it further, just that they might have hiccups with the line.

[Junior FO]

22 hours ago, bojan said:

Orlan is about 5k a pop (price for Russian army), still quite cheap. It has comercial level sensors, but so does most of the drones that Ukrainians use.

Considering I'm hearing upwards of 50-60k for the military grade mini UAV's we've gotten that sounds like a steal. Apparantly breaking the rotor blades alone comes with a price tag of 1k.

[bojan]

Stated prices for Russian army are however w/o any manufacturer's profit, iow, just their expanses. But even with 40% markup, that is just 7k$. Most of the price of the whole system is in the ground control station.

Edited June 19, 2022 by bojan

[Sardaukar]

48 minutes ago, bojan said:

Stated prices for Russian army are however w/o any manufacturer's profit, iow, just their expanses. But even with 40% markup, that is just 7k$. Most of the price of the whole system is in the ground control station.

Did you count all the graft going on?  

Problem is, not all promised things get to front lines....just like infamous Chinese truck tire scandal. Wonder if they prosecuted anyone from that....maybe not. 

[Josh]

On 6/17/2022 at 2:57 AM, KV7 said:

I agree with Ssnake.

Technology has made destruction of enemy forces somewhat easier, and this has made attacking harder, especially when the most destructive offensive weapons are in short supply or have too many targets to deal with. Achieving a huge local overmatch in armour and other maneuver elements isn't enough to permit quick and low cost advances. You can perhaps  use this overmatch to force the objective, but it will be very costly.

Motivated infantry with plentiful ATGM and artillery support defending some position can be broken with concentrated use of drones and artillery, and/or infantry assault. For Russia, the former is limited by resources and takes time, and the latter is often costly and also takes time and well trained and motivated infantry.

In the most pessimistic case for the attack, armour starts to look more like a defensive rather than offensive weapon. I.e kept in close reserve to plug gaps and halt breakthroughs and too vulnerable to spearhead an attack.

 

+1

I would add that achieving operational or even tactical surprise is far more difficult with drones and satellites overhead and strategic surprise is basically unachievable against a high end competitor with EO satellites *or* anyone supplied intelligence from such. Trying a river crossing or opposed landing usually requires air supremacy and severe overmatch.

[KV7]

26 minutes ago, Josh said:

+1

I would add that achieving operational or even tactical surprise is far more difficult with drones and satellites overhead and strategic surprise is basically unachievable against a high end competitor with EO satellites *or* anyone supplied intelligence from such. Trying a river crossing or opposed landing usually requires air supremacy and severe overmatch.

There is a complication here in that drones also make attacking easier, by making concealment of the defenders more difficult. And so with sufficient concentration of drones and artillery attacking can perhaps even become a bit of a cakewalk. However as above there likely will not be enough drones and artillery to do this over a wide front.

What it might end up looking like is both sides being able to destroy parts of the enemy defense cheaply, but not quickly enough to collapse a front. Basically holes are going to get burned and plugged on both sides and victory will come about through attrition and/or salami slicing.

[DB]

You can cue optical trackers from a radar picture and use image recognition to differentiate between birds and drones at a pinch, but you might also be able to deduce rotors directly from the shape/phase shift of the radar return signal as well. The latter would be candidate for machine learning.

One thing that people are still doing, though - the cost benefit analysis isn't about the cost ratio between the drone and the asset used to shoot it down, it's about the potential cost of letting the drone live.

Of course, if you have a more cost effective method of shooting the drone down, then that's great, but in the end, removing a drone and preventing a battery of artillery from destruction is still a positive, even if you used a missile that cost 50 times the value of the drone.

[CaptLuke]

15 hours ago, DB said:

. . . Of course, if you have a more cost effective method of shooting the drone down, then that's great, but in the end, removing a drone and preventing a battery of artillery from destruction is still a positive, even if you used a missile that cost 50 times the value of the drone.

At a tactical level this is very true and you raise a good point about the options that armies have in the short run, but strategically this can be untenable.

• If the enemy only has a small number of drones, or just has much less money than you do, then you can afford to do these kinds of shoot downs, but . . . 
• If the enemy starts to have thousands of these drones, which means you now need hundreds of thousands of missiles and that cost starts to add up to the point where you have less tanks and artillery to protect in the first place . . .

At some point it might be cheaper to scrap the artillery battery, avoid the need for all the missiles to protect it, and invest in some other kind of fire support.

1. One option might be longer ranged weapons that fire from outside the "cheap drone" ranges at the front
2. Another might be using something like a network of unattended launchers, like NLOS or, IIRC, a simpler Israeli equivalent that just used laser guidance, so that your fire support is both very distributed and less sensitive to the loss of any specific, unmanned, firing point. 

[JWB]

.........or an army could deploy drones that hunt and kill enemy drones.

[DB]

7 hours ago, CaptLuke said:

At a tactical level this is very true and you raise a good point about the options that armies have in the short run, but strategically this can be untenable.

• If the enemy only has a small number of drones, or just has much less money than you do, then you can afford to do these kinds of shoot downs, but . . . 
• If the enemy starts to have thousands of these drones, which means you now need hundreds of thousands of missiles and that cost starts to add up to the point where you have less tanks and artillery to protect in the first place . . .

At some point it might be cheaper to scrap the artillery battery, avoid the need for all the missiles to protect it, and invest in some other kind of fire support.

1. One option might be longer ranged weapons that fire from outside the "cheap drone" ranges at the front
2. Another might be using something like a network of unattended launchers, like NLOS or, IIRC, a simpler Israeli equivalent that just used laser guidance, so that your fire support is both very distributed and less sensitive to the loss of any specific, unmanned, firing point. 

If wishes were horses then beggars would ride.

I would wager that the cost of a successful engagement using drones will not ultimately differ significantly from the cost of using other means. At least, it won't vary by orders of magnitude. Specifically, I'd bet that suicide drones that can operate in a hostile EM environment almost certainly cost roughly the same as missile based precision weapon systems and generally fly more slowly as well. There's no evidence that anyone has anything approaching a "drone swarm" capability, outside of choreographed ersatz fireworks displays. All of the systems seen so far fit more closely into the "loitering munition" category, and none of them seem to be autonomous. Without the latter capability, I can't see anyone deploying "thousands" in short timescales.

[Ssnake]

I think we will contnue to see, as long as no major (new) conflict is started, loitering munitions with man-in-the-loop control which, indeed, do not scale (or don't scale well). I think that at least some of these LMs will have a non-advertised autonomous capability. Whether the customers actively ask for it or some engineers simply activate the capability later "on urgent customer request" is probably a function of a. the customer's ethical frame of reference and military doctrine, and b. how well the conflict can be managed without it.

But a fair share of public drone footage and/or manufacturer capability claims suggest that there's already at least some form of image recognition at work to support the human operator. Well, that same capability can easily be reprogrammed for autonomous killbot functions. Whether we should do this is a different question of course, but as hinted, the answer depends on a. and b. above because I suspect that eventual technical difficulties with autonomous operations can and will eventually be solved, and today's operator assistance functions lay the groundwork for "liberating" future operators from the need for kill decisions. Given that there are many countried operating from a different ethical framework, my rather pessimistic outlook is that it's just a question of when, not if.

[Ssnake]

2 hours ago, DB said:

Specifically, I'd bet that suicide drones that can operate in a hostile EM environment almost certainly cost roughly the same as missile based precision weapon systems and generally fly more slowly as well.

I agree.

[Stuart Galbraith]

On a related note, the British mod have cancelled  Mosquito, the loyal wingman tech demonstrator.

Edited June 26, 2022 by Stuart Galbraith

[Ivanhoe]

It seems to me that there will be circumstances where offensive drones won't need good IFF capabilities, for example scenarios where there is a clear and known battle edge. Throw hundreds or thousands of cheap killer drones programmed to blow up every truck and tracked vehicle, make sure they explode reliably and violently on impact so they cannot be cannibalized and reused by OPFOR. Kind of an area denial thing. Yeah, various poor tractors and farm trucks will be pounded to nothing, but disrupting OPFOR's rear echelon will do good things.

 

[Ssnake]

Image recognition is already better than that. You can already be selective enough in your targeting to go for armored vehicles exclusively. Even with a demand of 99% confidence in the identification you'll eliminate maybe 60...80% of all discovered targets. And you can, of course dial the threshold up and down as you like, depending on how much collateral damge you are willing to accept (or, less than ethical, how much collateral damage you want).

[bojan]

You already had those weapons, kind off, in the late '80s - US "Assault Breaker"and Soviet/Russian equivalent fired from BM-30 MRL. Fire loads of EFP equipped submunitions over potential target, they orient toward armor and fire EFPs toward their roof.

[Ivanhoe]

7 hours ago, Ssnake said:

Image recognition is already better than that. You can already be selective enough in your targeting to go for armored vehicles exclusively. Even with a demand of 99% confidence in the identification you'll eliminate maybe 60...80% of all discovered targets. And you can, of course dial the threshold up and down as you like, depending on how much collateral damge you are willing to accept (or, less than ethical, how much collateral damage you want).

Not any different in practice than H&I fire by the US Army in the 20th C, and probably vastly better in terms of cost effectiveness and ratio of OPFOR/collateral damage.

My thoughts are aimed at blunting a massive mech/armor assault; not an all-situations/all-OPFORs solution, but a low-cost/high-volume tool.

[Ssnake]

All I'm saying, once that the neuronal network has been trained, you get both a robotic aerial reconnaissance platform as well as the ability to mass a swarm attack, provided you can find a suitable combination of carrier, warhead/munition, and flight endurance. Switchblade S-600 sounds like a pretty well balanced design, if you have to pick a specific example.

[KV7]

Even if useful offensive or recon capability is expensive (and I doubt this will be the case going forward) proliferation of UAV that are low capability and e.g. run some preprogrammed route (and then are resistant to jamming even without sophisticated electronics) over some position will be a huge problem, because one cannot determine they are a threat or not, and will then need to be hit, or if in sufficient numbers they will permit capable drones mixed within them to operate in an environment where the air defenses are overloaded with targets.

[Ssnake]

Precisely. That's why I think that doctrinally most western armies are well behind the curve, catching up to a threat picture from five years ago, rather than anticipating the next five to ten years.

[KV7]

2 hours ago, Ssnake said:

Precisely. That's why I think that doctrinally most western armies are well behind the curve, catching up to a threat picture from five years ago, rather than anticipating the next five to ten years.

The 'doctrine' will likely be to buy a lot of expensive SHORAD and other solutions and enrich the companies that can cobble together something to offer, even if it isn't a particularly efficient solution.

[lucklucky]

 

https://www.thedrive.com/the-war-zone/swarm-of-40-drones-over-fort-irwin-an-ominous-sign-of-whats-to-come