How Hot is Ukraine Gonna Get?

[zinz]

https://ludic.mataroa.blog/blog/i-will-****ing-piledrive-you-if-you-mention-ai-again/

On the topic what AI can and can't do.

 

 

[Mattias]

22 hours ago, Probus said:

Does anyone remember this automatic shotgun weapon system that never quite made it in the early 80s:

 https://en.m.wikipedia.org/wiki/Atchisson_AA-12

Wouldn't one of these loaded with "bird shot" be the perfect weapon for squads to have to help fend off drones in Ukraine?

It feels logical but I doubt it will cut it. Skip to minute nine in the video below and see why. Seeing that I felt the wave of sheer terror. And imagine swarms.

 

Edited July 6 by Mattias

[Battlefront.com]

1 hour ago, Mattias said:

It feels logical but I doubt it will cut it. Skip to minute nine in the video below and see why. Seeing that I felt the wave of sheer terror. And imagine swarms.

 

Oh bleep me.  I mean, intellectually I know all of this is possible, but to see it in action like that... ugh.  This is even superior to some of the Hollywood sci-fi stuff we've brought into the discussion to show how the battlefield has changed.

This right here is why I look at point defense systems, like APS, Phalanx, and anything missile based and conclude they are technological dead ends.  Picture something like this that has self selected a target and then goes after it using a randomized pattern of 12G movements.  I don't see a solution for this using any of the currently implemented or conceived point defense tech.

The only thing I can conceive of that can handle something like this is an area spread of energy.  Microwaves, lasers, etc.  However, none of these are practical for a variety of reasons.

I'll say this again... it's not just the tank that's dead.  It's pretty much anything an enemy wants to destroy that's dead.  Be it a guy on foot, an IFV, an oil refinery flash point... ugh.

Steve

[kimbosbread]

Interceptors it is!

[sburke]

11 minutes ago, Battlefront.com said:

I'll say this again... it's not just the tank that's dead.  It's pretty much anything an enemy wants to destroy that's dead.  Be it a guy on foot, an IFV, an oil refinery flash point... ugh.

Steve

troublemaker.  I think Steve is gonna give you a time out.. oh wait...

[dan/california]

6 hours ago, zinz said:

https://ludic.mataroa.blog/blog/i-will-****ing-piledrive-you-if-you-mention-ai-again/

On the topic what AI can and can't do.

 

 

This is a truly great article. these are the money quotes.

Quote

But most companies can't do this, because they are operationally and culturally crippled. The median stay for an engineer will be something between one to two years, so the organization suffers from institutional retrograde amnesia. Every so often, some dickhead says something like "Maybe we should revoke the engineering team's remote work privile - whoa, wait, why did all the best engineers leave?". Whenever there is a ransomware attack, it is revealed with clockwork precision that no one has tested the backups for six months and half the legacy systems cannot be resuscitated - something that I have personally seen twice in four bleeping years. Do you know how insane that is?

Quote

The only thing you should be doing is improving your operations and culture, and that will give you the ability to use AI if it ever becomes relevant. Everyone is talking about Retrieval Augmented Generation, but most companies don't actually have any internal documentation worth retrieving. Fix. Your. CRAP.

The article has lot to say about AI, too, but it is really about the dystopian hellscape of corporate tech culture. The poor guy has gone around the bend, over the edge of the cliff, and is now hanging on to his sanity by two fingernails.

Edited July 6 by dan/california

[poesel]

1 hour ago, Battlefront.com said:

This right here is why I look at point defense systems, like APS, Phalanx, and anything missile based and conclude they are technological dead ends.  Picture something like this that has self selected a target and then goes after it using a randomized pattern of 12G movements.  I don't see a solution for this using any of the currently implemented or conceived point defense tech.

The only thing I can conceive of that can handle something like this is an area spread of energy.  Microwaves, lasers, etc.  However, none of these are practical for a variety of reasons.

'up to 12g'... - that is in the axial direction of the 4 props and probably right at the start when the torques of the motor is highest. It does not mean the copter has 12g anytime in any direction.

I'm not saying the copter is not nimble. But building a two-axis gimbals that can follow this bird wrt to angular speed is no problem. Servo motors allow very, very dynamic movements. If your sensors are up to it, a laser pointer on that gimbals in the same room would paint the copter during its whole flight.
Put enough power into the laser or replace it with a gun, and the copter is down.

I've read this here several times, that a ground based gun could be overcome by attacking from two sides. But I think this is not true. You can turn a gun like an M2 (just as an example) by 180° literally in the blink of an eye if you want. It comes down to the factors: speed of the drone, range of the gun and accuracy how many drones you need to overcome one gun.

Let's assume your gun has a range of 1000 m and the drone flies with 50 m/s that is 20 s in the range of the gun (without time spent for evasion). How long does a gun need to take down one drone? I don't know, but I guess it is much less than 20 seconds.
Of course, there is the question of how accurate the sensors are. I don't know this either, but these are problems with known solutions.

Drones have come to stay, but they are not the end of everything else.

[ArmouredTopHat]

2 hours ago, kimbosbread said:

Interceptors it is!

This really does make me wonder if mass drone interception of ISR and loitering munitions will give vehicles a bit more breathing space again.

 

2 hours ago, Battlefront.com said:

This right here is why I look at point defense systems, like APS, Phalanx, and anything missile based and conclude they are technological dead ends.  Picture something like this that has self selected a target and then goes after it using a randomized pattern of 12G movements.  I don't see a solution for this using any of the currently implemented or conceived point defense tech.

 

1 hour ago, poesel said:

ve read this here several times, that a ground based gun could be overcome by attacking from two sides. But I think this is not true. You can turn a gun like an M2 (just as an example) by 180° literally in the blink of an eye if you want. It comes down to the factors: speed of the drone, range of the gun and accuracy how many drones you need to overcome one gun.

Comparing a bulky phalanx CIWS that's optimised for shooting down extremely fast targets like ASMs with a gatling system is not really an apt comparison to anything PD related that's going to come down the line for counter UAS, its just what happens to be on hand for immediate use. Any system designed for small drone and loitering munition destruction is going to be much lighter, if only because whatever it ends up firing does not need to be nearly as heavy or fired at the same rate of fire as something like CIWS, you dont need a heavy and sophisticated radar either, most likely a thermal detection system will be more than sufficient (perhaps combined with an acoustic system) With AI this could be a largely automated process of tracking, acquiring and confirming a drone target in the space of a second or two before destroying it.

This means an extremely fast tracking turret in all likelihood and something that's crucially, likely to be much cheaper too in both system and ammunition used for disposal. (At least compared to missiles or heavier PD) There is also the concept of using APS radars to perform the detection as well that was previously mentioned. The current thought is something in the 20-30mm cannon range with airburst munitions, though I wonder if something that is simple enough to fire shotgun or rifle calibre rounds might be sufficient and a cheaper system that you could mount on the top of most vehicles. It does not have to be pretty, just a sufficiently traversable small turret equipped with a shotgun type weapon. This is not exactly rocket science. 

If AI can be used to coordinate swarms, why cant it be used to coordinate counter measures? Defensive munitions (Ie other drones) being used to intercept a swarm while ensuring PD guns are efficiently used to clean up the rest seems just as plausible as a drone swarm on the offensive.  

Declaring PD to be a dead end just because of current generation PD performance against what is a very new threat is a little odd.

https://www.peoplesproject.com/en/9-types-of-anti-drone-weapons/

https://www.asianmilitaryreview.com/2023/05/naval-rws-more-than-just-uas-swatters/

There is a lot of potentially interesting counter systems in play, time will tell what works and what doesn't. 

Edited July 6 by ArmouredTopHat

[FancyCat]

Quote

NATO countries are set to donate another massive package of long-range air defense systems to Ukraine at next week’s NATO summit in Washington. The US, Germany, and Romania will send 3 PATRIOT SAM units. Italy and France will join to supply a SAMP/T system.

 

[chrisl]

2 hours ago, Battlefront.com said:

Oh bleep me.  I mean, intellectually I know all of this is possible, but to see it in action like that... ugh.  This is even superior to some of the Hollywood sci-fi stuff we've brought into the discussion to show how the battlefield has changed.

This right here is why I look at point defense systems, like APS, Phalanx, and anything missile based and conclude they are technological dead ends.  Picture something like this that has self selected a target and then goes after it using a randomized pattern of 12G movements.  I don't see a solution for this using any of the currently implemented or conceived point defense tech.

The only thing I can conceive of that can handle something like this is an area spread of energy.  Microwaves, lasers, etc.  However, none of these are practical for a variety of reasons.

I'll say this again... it's not just the tank that's dead.  It's pretty much anything an enemy wants to destroy that's dead.  Be it a guy on foot, an IFV, an oil refinery flash point... ugh.

Steve

It's not one of them doing that - a point defense system might be able to deal with that, even at crazy high speeds and accelerations.  It's a half dozen of them, communicating with each other.  On every single target.  The only way to be sure you hit them is if you're a frag grenade with a high density and uniform distribution of metal.  

eta: it's a good thing you design historical games and don't have to stick to modern games.  Because CM:Drone Wars will be a one-turn slaughterfest.

Edited July 6 by chrisl

[ArmouredTopHat]

6 minutes ago, chrisl said:

It's not one of them doing that - a point defense system might be able to deal with that, even at crazy high speeds and accelerations.  It's a half dozen of them, communicating with each other.  On every single target.  The only way to be sure you hit them is if you're a frag grenade with a high density and uniform distribution of metal.  

I imagine any defence against this is going to involve layers, medium range airburst munitions from the stuff we see operating in Ukraine against Shaeds being the middle aspect of that combined with interceptor drones replacing traditional missile systems that are just simply not economical against smaller drones. 
 

We already see some pretty small anti drone systems, and even this one shown is something I think can be sized down further and simplified. 

Edited July 6 by ArmouredTopHat

[dan/california]

3 minutes ago, chrisl said:

It's not one of them doing that - a point defense system might be able to deal with that, even at crazy high speeds and accelerations.  It's a half dozen of them, communicating with each other.  On every single target.  The only way to be sure you hit them is if you're a frag grenade with a high density and uniform distribution of metal.  

And then the NEW combined arms is artillery delivered chaff and glorified fireworks just to overload the defense systems while the drones do the killing, or, or.... I can lay out a five generation improvement plan for the drones that is not crazy expensive, will be exponentially more effective than what is out there now, and is essentially just picking up and integrating off the shelf bits and pieces. The plan for drone defense is harder than that, and at a minimum involves unproven engineering.

[chrisl]

Just now, ArmouredTopHat said:

I imagine any defence against this is going to involve layers, medium range airburst munitions from the stuff we see operating in Ukraine against Shaeds being the middle aspect of that combined with interceptor drones replacing traditional missile systems that are just simply not economical against smaller drones. 
 

We already see some pretty small anti drone systems, and even this one shown is something I think can be sized down further.

For sure.

But the sensor problem isn't easy, and we're just at the start of what you can do with small autonomous things.  @The_Capt's smart land mines, if you will, but moving in 3D at anything from zero to 50 m/s.  Instead of laying a bunch of land mines, you ornament a bunch of trees with a similar number of drones that just sit quietly until their victim is surrounded and within 100 m.  That's when you slap the big red "EMP" button if you have one.

[chrisl]

3 minutes ago, dan/california said:

And then the NEW combined arms is artillery delivered chaff and glorified fireworks just to overload the defense systems while the drones do the killing, or, or.... I can lay out a five generation improvement plan for the drones that is not crazy expensive, will be exponentially more effective than what is out there now, and is essentially just picking up and integrating off the shelf bits and pieces. The plan for drone defense is harder than that, and at a minimum involves unproven engineering.

Not just artillery delivered chaff - why aren't you mixing drones in?  It's a way to get more range out of drones with small batteries.

[Battlefront.com]

2 minutes ago, ArmouredTopHat said:

We already see some pretty small anti drone systems, and even this one shown is something I think can be sized down further and simplified. 

No, it can't.

Please... this time follow through with the physics of this instead of waiving your magic wand.

What happens to ANY point defense system when a half dozen drones are coming in at from all angles, zig-zagging about, all at 2m above the ground level.  Tell me, where do all of those outgoing defense rounds go when they are fired? 

Put another way, it might be able to get a couple of the drones before being hit, but I wouldn't want to be 8000m anywhere near this thing when it does so.

Steve

[dan/california]

Speaking of doing a good job of combining pieces we already have. This system for chucking cruise missiles out the back of a C-130 in Quantity seems brilliant. Rather like putting the SM-6 on an F-18, somebody needs a medal and a promotion for that one. Both of these things also highlight that the missile is rapidly becoming much more important than the airplane, or at least the specifics of the air plane.

Edited July 6 by dan/california

[ArmouredTopHat]

7 minutes ago, Battlefront.com said:

No, it can't.

Please... this time follow through with the physics of this instead of waiving your magic wand.

What happens to ANY point defense system when a half dozen drones are coming in at from all angles, zig-zagging about, all at 2m above the ground level.  Tell me, where do all of those outgoing defense rounds go when they are fired? 

Put another way, it might be able to get a couple of the drones before being hit, but I wouldn't want to be 8000m anywhere near this thing when it does so.

Steve

The system I showed above was a Gatling type system, that alone adds a fair bit of weight and complexity even though the system is like a quarter of the size of Phalanx. 

Take the sensor systems from that unit and put it on a regular 50 cal or some sort of shotgun housing that's half the size and less complex. Now you can easily stick it onto a vehicle. 

Now network a defence system around that that includes heavier airburst and drone interceptors that also uses AI to coordinate a response that starts miles out from target. The idea is not to focus on PD as the be all end all solution but the end system that's employed as a final resort. The more things that disrupt a kill chain the better. 

The future is not as clear cut I think. We dont even know if AI drone swarms will work in practicality. 

Edited July 6 by ArmouredTopHat

[Battlefront.com]

6 minutes ago, chrisl said:

But the sensor problem isn't easy,

Absolutely.  The ability to track targets often requires significant time to lock on.  The sort of drone attacks we're talking about should be presumed to not give that sort of benefit.  Also, whatever sensors can do there's going to be things they can't do.  It is probable that it is easier to reprogram a drone to exploit those limitations than it is to reprogram (if even possible) a sensor array to respond to new drone behavior.

Steve

[dan/california]

11 minutes ago, chrisl said:

Not just artillery delivered chaff - why aren't you mixing drones in?  It's a way to get more range out of drones with small batteries.

And, and, and...although it is easier to pack strips of foil into an artillery round. One the massive advantages of drones economically is that they don't have to survive the Gs of being gun launched. Aluminum foil and glorified flares are pretty simple in that regard. Although I think sticking drones into a GMLRS round is an absolute no brainer. I assume someone is already working on that, surely.

Edited July 6 by dan/california

[Battlefront.com]

3 minutes ago, ArmouredTopHat said:

The system I showed above was a Gatling type system, that alone adds a fair bit of weight and complexity even though the system is like a quarter of the size of Phalanx. 

Take the sensor systems from that unit and put it on a regular 50 cal or some sort of shotgun housing that's half the size and less complex. Now you can easily stick it onto a vehicle. 

Now network a defence system around that that includes heavier airburst and drone interceptors that also uses AI to coordinate a response that starts miles out from target. 

The future is not as clear cut I think. We dont even know if AI drone swarms will work in practicality. 

I asked you specifically about the physics of the outgoing rounds.  You did not address that at all.

Steve

[chrisl]

2 minutes ago, dan/california said:

And, and, and...although it is easier to pack strips of foil into an artillery round. One the massive advantages of drones economically is that they don't have to survive the Gs of being gun launched. Aluminum foil and glorified flares are pretty simple in that regard. Although I think sticking drones into a GMLRS round is an absolute no brainer. I assume someone is already working on that, surely.

The gun launch wouldn't be pleasant, but it's a speedy delivery system and not that terrible.  I have plans to do worse to more fragile things.

[Battlefront.com]

5 minutes ago, ArmouredTopHat said:

The future is not as clear cut I think. We dont even know if AI drone swarms will work in practicality. 

This is a huge problem for me.  You are confident in systems that don't exist, even on paper, and yet are skeptical of ones that already exist being adapted for military use.  That's a huge red flag for me.

The weight of evidence shows that there should be NO doubt about swarms. None.  Zero.   At least that should be the premise to adopt when planning out billions of Dollars in spending. 

At worst, swarms are way ahead of the defenses against them.  Even if it takes a couple of years to make swarms practical, it is going to take way longer than that to get a defense system in place that can even handle single drones.  Going forward with a defense system that ASSUMES swarms won't ever be a tangible threat is just about the worst idea I can think of.

Steve

[chrisl]

3 minutes ago, Battlefront.com said:

I asked you specifically about the physics of the outgoing rounds.  You did not address that at all.

Steve

I bet there are at least 3 movie scenes that use that concept for a good guy to get the bad guy to wipe out his own side.  Without searching, I'd bet at least one from Marvel and at least one involving Jackie Chan, probably shot in Hong Kong and dubbed.

[Probus]

3 hours ago, Battlefront.com said:

This right here is why I look at point defense systems, like APS, Phalanx, and anything missile based and conclude they are technological dead ends.  Picture something like this that has self selected a target and then goes after it using a randomized pattern of 12G movements.  I don't see a solution for this using any of the currently implemented or conceived point defense tech.

Yeah.  I wonder if this is how our ancestors felt after the machine gun of WWI or the atomic bomb/Air Power/Armored Power of WWII.  What the heck comes next?  It's the wild, wild west again and defense engineers are gonna have to start thinking outside the box again.  Surely there is a simple defense against this new threat. 

We need a wasp 'spray' for a drone.  Or some new efficient way to store energy for a laser/microwave defense system.  Just as importantly we need a way to either blind our enemies surveillance drones or camouflage our units to the drone's sensors.  This is really a terrifying new threat on the battlefield.

[Battlefront.com]

1 hour ago, poesel said:

'up to 12g'... - that is in the axial direction of the 4 props and probably right at the start when the torques of the motor is highest. It does not mean the copter has 12g anytime in any direction.

I'm not saying the copter is not nimble. But building a two-axis gimbals that can follow this bird wrt to angular speed is no problem. Servo motors allow very, very dynamic movements. If your sensors are up to it, a laser pointer on that gimbals in the same room would paint the copter during its whole flight.
Put enough power into the laser or replace it with a gun, and the copter is down.

And what if the drone is coming in 2m "off the deck".  What are you going to paint besides the drone?

The other thing we've already discussed is the impracticality of lasers.  They require way too much power generation to do the job they are being conceived of doing.

I also have the same doubts as chrisl that solving the sensor/tracking problems is practical in the near term.

1 hour ago, poesel said:

I've read this here several times, that a ground based gun could be overcome by attacking from two sides. But I think this is not true. You can turn a gun like an M2 (just as an example) by 180° literally in the blink of an eye if you want. It comes down to the factors: speed of the drone, range of the gun and accuracy how many drones you need to overcome one gun.

Let's assume your gun has a range of 1000 m and the drone flies with 50 m/s that is 20 s in the range of the gun (without time spent for evasion). How long does a gun need to take down one drone? I don't know, but I guess it is much less than 20 seconds.
Of course, there is the question of how accurate the sensors are. I don't know this either, but these are problems with known solutions.

Drones have come to stay, but they are not the end of everything else.

And yet you are adding to the argument that there's nothing practical to defend against them.  Sooooo...

Steve