chrisl

I am guessing we will see large scale use by fall. This horse has left the barn, the farm, and the county.
 

Sadly there are lot of people that ought to see the inside of a prison cell forever, b ut probably won't
In the short to medium term I think smart fires are going to move away from GPS. Either laser designation from drones. Or smart submunitions with infrared sensors like the Smart/Bonus rounds.
This type of submunition is going to get put on ATACMS, HIMARS, and similar as well. There is also no reason you couldn't do a fragmentation version for infantry and light vehicles.

You could use the Drone based system I described above to lay charges on mines and only set them off at the moment the assault begins ... which would provide a degree of surprise. A considerable degree of surprise.
Given the general sluggishness of Russian responses to ... anything ... coupled with the degradation of their artillery systems (heck, how many trained artillerists do they even have left?) plus, I would presume, a lack of artillery launched mines ... such a surprise might work. Especially if you separate the attack lanes beyond the effect range of artillery delivered minelets and didn't trigger all of the lane clearing charges at once ... so if the Russians did manage to get their act together and hit one set, simply blow another series of lanes and hit there.
Of course, Friction would be the problem to beat, per Clausewitz.

Which is why I indicated 'not for assaults' ... I just don't think engineering vehicles are the answer, or not ones as currently envisioned.
ISTR that someone posted here (or I read elsewhere) that there was at least one ongoing attempt to have drone based mine detection and elimination ... detecting them with onboard sensors and marking their position for either another drone to drop a charge on it or for sappers to neutralise it. Presumably you could have a drone swarm with mixed mine detector drones and charge carrying drones work in tandem.
Do it at night and the drones would be virtually undetectable ... 
Would require a lot of drones, of course, but you could run them so they simply cleared lanes through a larger field while mapping all the routes fior downloading into the GPS of the attacking vehicles.

These are features of tank as a weapon. But what is more important are the effects and functions that the weapons gives you. If you can replace the effects and functions in your toolbox, you do not need to recreate the exact features mix.
IMO These features allowed the tank to fulfill 3 functions: 1) it could break through; 2) it could exploit; 3) it could fight other tanks. Function 3) is easily replaceable by NLOS ATGM and drones, in fact tank is not longer the best AT weapon. Function 1) is also replaceable, by application of sufficient amounts of artillery you can level the trench and bury the people in it. Same principles as Bruchmuller used, just substitute PGMs for gas shells. 
Where I see a problem is function 2). The tank allowed forward movement at vehicle speed and relative immunity from indirect HE and machine gun fire, i.e. the killers of extended advance on foot/horse. Tanks limited the threats to AT assets which, when they were AT guns and early ATGMS, could be overloaded and outshot by massing tanks against them, whereupon the advance would resume at tank speed. The enemy had to countermass vs your mass by creating a Pakfront, and a Pakfront usually could not be everywhere. Not anymore. The NLOS AT assets, drones and artillery PGMs can instantly be concentrated from a wide area,  without having to locate them physically close to one another. Having eliminated the tank from the equation how to transition from a breakthrough to an extended advance at vehicle speed? I don't know. In other words, the only advance possible looks like a constant series of breakthroughs with the enemy always able to retreat and recreate a new defence line in front of you, and you always frontally assaulting. Which seems a fair description of the Zaporozhe offensive 2023. In yet other words, over a 100 years of warfare has just been erased, and conceptually we are back at the Kaiserschlacht exactly. We are unable to recreate Amiens or any later battle of movement.

In 2021 the Pentagon drafted a proposal for a replacement for the 50+ year old TOW. They're hoping for 10+km range, non-LOS launch and 'cooperative engagement' (whatever that means). Their specs seemed futuristic at the time, now it just sounds like a bog standard suicide drone. But fired out of a TOW launch tube (their mandated specs.)

You know, I could think of a lot lighter, smaller weighs to a bunch of computers in the field than an AFV. Same with power generation. You are schlepping around so much armor and fuel that it just becomes a snowball of logistic nightmares.
You don’t want to couple sensors to this big heavy noisy thing, because you lose it you lose your eyes. Sensors should be on all the small platforms. And then there’s the whole rub of this- why do you need a big node on the ground? You’ll have LOS issues with radio signals, so you’ll need a repeater drone or an antenna on a big *** pole.
What about distributing the nodes as cheapo disposable things you can just drop on the battlefield? Like mines, but for information. Why not just have the compute on a drone that has no LOS issues?  And if you go that far, why not just beam it up to a satellite where you can have more or less infinite compute on tap?
Wrong. You can’t eliminate small stuff smart mines or drones or sensors. You can’t take out flying stuff. It’s not even that good for infantry unless you have HE or flechette shells.
EDIT: To be clear, I think the future battlefield will have many smaller, distributed systems that are better served with autocannon fire.

So when talking about capabilities tanks provide and nothing else does - what does a tank have that something like Bradley don't? It probably can't fight tanks, but it can fight other IFVs and provide support - and some claim that autocannon does that better than a tank gun, because it can just keep firing and other than heavy fortifications it's powerful enough. It can also carry troops.
Looking at it from the other angle, if the consensus is that future military needs to be tankless but infantry and fires- and drones-heavy, how does the infantry get around? Are IFVs as dead as tanks and soldiers use Toyota Hilux now because you can fit many drones into the truck bed?

Star Wars never really went away.  The pieces got split up on paper, but a large fraction of them continued and have turned into things that are part of the space ISR and anti-missile systems we have today.  Some of the wackier ones (coughEdTellercough) went away, but a lot of stuff got built, and fair bit of it is actually useful.

Is there any real evidence that the Armata is anything more than a Rose Parade float that they haven't gotten around to putting the flowers on yet?  For as much as actually gets shown, it could be plywood over a T-34 chassis stolen from an old monument.

An interesting discussion about tanks being big and hot:
https://x.com/JonHawkes275/status/1807758981831495819
 

Oh yes. I was only thinking about the Orlan equivalents. You cannot cover the frontline with training planes, so some sort of reusable yet not too extensive interceptor drone might come in handy. Does not risk the lives of the pilots but allows to push the Russian observation bubble back a bit.

That's probably much more effective for hitting the slow fixed wings like Shaheds and Orlans that are more like slow cruise missiles on a "set and forget" path than for dealing with electric quadcopters.  The hard part with all the drones is detection.  If you know where something is, hitting it with something is relatively straightforward.  
The Orlans and Shaheds have what are basically lawnmower engines that are loud and relatively hot, not as maneuverable as a quadcopter, and at long ranges from their launch are generally on a programmed flight path.  They're also relatively large, so they're easier to see visually, though they're still mostly made of things like wood and fiberglass, so they won't have big radar signatures.
Quadcopters (and their 6 and 8 motored relatives) are a lot harder - they're smaller, to start, even the ones big enough to carry an RPG-7 warhead.  They're much quieter - almost inaudible at 50 or 100 feet, made of mostly plastic, and not generating a lot of heat.  They're also typically controlled by an operator and capable of very abrupt maneuvers.  So hitting them with an MG from a moving fixed wing aircraft is probably a lot harder unless you have some way to detect them reliably and use an auto-aim system.

There's a very obvious reason for the difference in your prediction vs. the predictions of the so-called experts: you've been doing quantitative modeling of actual combat capability for 20 years, including models of most of the forces and equipment involved in this war.  The framework for that modeling has been validated on older equipment by comparing outcomes of scenarios to real battles that are well documented.  There's less to validate against for the more modern equipment, but there's still some info, and lots of details of equipment became available after the end of the USSR.
I've been reading Kahneman's "Thinking Fast and Slow" on and off and he points out why experts like Kofman are often wrong - for the most part they don't get feedback on their predictions because they're either dealing with things that are very fuzzy and they can claim "I was close" or they're very long term and they just get lost in time.  By doing quantitative modeling you've had a ton of practice in making predictions of how things should happen in a peer conflict like this and then seeing if your model plays out as expected. 

Did someone utter "Peng?"
Peng. It's what's for dinner.

Ukraine is on track to build 1M drones in 2024 - that's at least 80,000/month, which is why we see them chasing individual russian soldiers around.  They aren't in any kind of short supply and their combat efficiency is high (fewer than 10 drones per casualty, possibly fewer than 5).  Most of the parts come from (drum roll).....China.  It's basically all consumer stuff and the attitude from China is that it goes into a container off to foreign distributors and it's not China's problem as long as their manufacturers get paid.  
Part of the reason the US pays a lot for drones is that they try to not be dependent on parts from China and there's a lot smaller and less efficient supply chain for those.  There's also never been commitment by the US military to buy enormous numbers of them, so there aren't factories to crank out the drone parts in huge quantities at low unit costs.  The chinese drone parts are all intended for civilian/business/hobbyist use and are produced in large quantities with good scale efficiencies.  
As recently as 2019, the US was struggling to keep the cost of a comparable drone at 10x the cost of a Chinese drone because the private commercial base in the US couldn't compete with DJI. https://www.c4isrnet.com/unmanned/2019/09/17/can-the-army-secure-an-american-made-quadcopter/ 
I went and looked up the production dates, costs, capabilities, and quantities produced of all the listed drones to the extent that the information is available.
Aeryon can do DJI knockoffs for what sounds like a few tens of $K.
All of them except the Drone40 are oriented towards local ISR in terms of design and production quantities.  A few of the less expensive (less than ~$50K/unit) have been produced in quantities of several thousand
The older ones (before about 2010) are basically RC airplanes that cost a few $M each for ISR and are basically obsolete now.  If the airframes were still in production you could redo their entire inside for cheap, but I suspect that they're not.  Of those, it looks like only the Puma was produced at scale of more than a few hundred (more than 1000 produced). The Wasp is archaic and has been replaced by the SkyDio X2D, which costs about $10K each without batteries or accessories.  But it's getting there in cost and capability.  Some of them don't even have any more range than an ATGM (Dragoneye, Black Hornet).  The Black Hornet really looks like it's most useful as a fancy mirror for looking around corners or into second story windows - it's only got a 2 km range, so it's about looking around corners in short visual range environments.
The Drone 40 is the only one under $5K.  It's about $1000/each for a 40 mm grenade with rotors and a 60 minute dwell time.  It has a kind of silly design feature - it's made to be launched from a 40 mm grenade launcher.  But it probably is more cost, mass, and capability effective to not include the features that enable that and spend the mass and volume on either more battery or more bang.
So yes, there's been some experimentation with drones by the US, but given what it's been, it's entirely local ISR oriented, and likely limited to special units, given the quantities.  There's been almost no development of drone munitions (other than the Drone 40 and Switchblade), and switchblade was procured in experimental quantities and even "mass" production is ~500/month, as compared to 80,000+/month that Ukraine can use.  There's only just starting to be development of precision munition delivery with drones, and there doesn't seem to be any doctrine developed around it - that will more likely be copied from Ukraine than the other way around.

Ukraine is on track to build 1M drones in 2024 - that's at least 80,000/month, which is why we see them chasing individual russian soldiers around.  They aren't in any kind of short supply and their combat efficiency is high (fewer than 10 drones per casualty, possibly fewer than 5).  Most of the parts come from (drum roll).....China.  It's basically all consumer stuff and the attitude from China is that it goes into a container off to foreign distributors and it's not China's problem as long as their manufacturers get paid.  
Part of the reason the US pays a lot for drones is that they try to not be dependent on parts from China and there's a lot smaller and less efficient supply chain for those.  There's also never been commitment by the US military to buy enormous numbers of them, so there aren't factories to crank out the drone parts in huge quantities at low unit costs.  The chinese drone parts are all intended for civilian/business/hobbyist use and are produced in large quantities with good scale efficiencies.  
As recently as 2019, the US was struggling to keep the cost of a comparable drone at 10x the cost of a Chinese drone because the private commercial base in the US couldn't compete with DJI. https://www.c4isrnet.com/unmanned/2019/09/17/can-the-army-secure-an-american-made-quadcopter/ 
I went and looked up the production dates, costs, capabilities, and quantities produced of all the listed drones to the extent that the information is available.
Aeryon can do DJI knockoffs for what sounds like a few tens of $K.
All of them except the Drone40 are oriented towards local ISR in terms of design and production quantities.  A few of the less expensive (less than ~$50K/unit) have been produced in quantities of several thousand
The older ones (before about 2010) are basically RC airplanes that cost a few $M each for ISR and are basically obsolete now.  If the airframes were still in production you could redo their entire inside for cheap, but I suspect that they're not.  Of those, it looks like only the Puma was produced at scale of more than a few hundred (more than 1000 produced). The Wasp is archaic and has been replaced by the SkyDio X2D, which costs about $10K each without batteries or accessories.  But it's getting there in cost and capability.  Some of them don't even have any more range than an ATGM (Dragoneye, Black Hornet).  The Black Hornet really looks like it's most useful as a fancy mirror for looking around corners or into second story windows - it's only got a 2 km range, so it's about looking around corners in short visual range environments.
The Drone 40 is the only one under $5K.  It's about $1000/each for a 40 mm grenade with rotors and a 60 minute dwell time.  It has a kind of silly design feature - it's made to be launched from a 40 mm grenade launcher.  But it probably is more cost, mass, and capability effective to not include the features that enable that and spend the mass and volume on either more battery or more bang.
So yes, there's been some experimentation with drones by the US, but given what it's been, it's entirely local ISR oriented, and likely limited to special units, given the quantities.  There's been almost no development of drone munitions (other than the Drone 40 and Switchblade), and switchblade was procured in experimental quantities and even "mass" production is ~500/month, as compared to 80,000+/month that Ukraine can use.  There's only just starting to be development of precision munition delivery with drones, and there doesn't seem to be any doctrine developed around it - that will more likely be copied from Ukraine than the other way around.

There are other factors involved in military R&D and other decisions. There still massive and cut-throat “turf wars” that the U.S. Military services engage in and shape the R&D and operational decisions by the Politicians. For example, after WW II, the newly created USAF worked very hard with their supporters in Congress to disband the U.S. Navy because the USAF had the Atomic bomb and there was no use for the Navy, and Congress could save all the Navy’s funding (or give it to the new USAF instead). A portion of the USAF, know as the “Fighter Mafia” has tried to get rid of the A-10 Warthog because it was designed to support the U.S. Army. The U.S. Army has been trying to gain control of the USMC or to absorb it since WW II ended. We’ve all seen how those “turf wars” ended. And the “turf wars” continue to this day!
One further note for those of you who think they aren’t still being fought behind closed doors. In the late 1980s I worked with USAF, USA, and USN reps to create an aircraft engine  Joint Engine Specification because each service had their own unique engine specification for the same basic engine. We met for about two weeks, but failed because the AirForce and Army reps couldn’t stop arguing about how the Army violated the “1948 Keywest Accords,” that basically divided air responsibilities between the USAF and the USA, when the Army put weapons on Helicopters!
This reality continues to determine what is researched and funded.
 

Is there any real evidence that the Armata is anything more than a Rose Parade float that they haven't gotten around to putting the flowers on yet?  For as much as actually gets shown, it could be plywood over a T-34 chassis stolen from an old monument.

That list is almost entirely small-quantity, high-cost uncrewed ISR aircraft that function mostly as crewed aircraft substitutes. The principle behind most of them is to spot something so that you can then send some precision strike later (probably from a crewed aircraft, sometimes fired from the drone itself or artillery)  The lower half of the list lean toward smaller things, but still very much ISR rather than munition.  Several of them can carry and fire standard aircraft munitions (in small quantities), and two, maybe three of them are actual munitions unto themselves.  The cheapest thing on the list is about $100K/unit and has a production rate of ~500/month (switchblade).  
Ukraine is making (and using) more drones every month (80K to 100K) than the combined lifetime production of everything on that list.  They're using them in a fundamentally different way than everything on the list - they're taking advantage of what is apparently an infinite supply of of RPG-7 warheads and are replacing the launcher tubes with drones.  It extends their range to well beyond visual and drastically reduces the risk to the operator.  In running out of artillery shells, they've figured out a way to take advantage of something that they can get in large quantity (drone parts) and turned them into personal delivery systems for RPG-7 warheads.  There are other small munitions that they put on them, but it's the same idea - low cost, disposable system that has long-ish range and direct visual feedback, each carrying a munition that can destroy one of anything on the battlefield.  And up until the moment that each drone blows up on something, it's also providing ISR (usually redundant to other ISR drones that are on overwatch to guide and evaluate).  
 
 

There's a very obvious reason for the difference in your prediction vs. the predictions of the so-called experts: you've been doing quantitative modeling of actual combat capability for 20 years, including models of most of the forces and equipment involved in this war.  The framework for that modeling has been validated on older equipment by comparing outcomes of scenarios to real battles that are well documented.  There's less to validate against for the more modern equipment, but there's still some info, and lots of details of equipment became available after the end of the USSR.
I've been reading Kahneman's "Thinking Fast and Slow" on and off and he points out why experts like Kofman are often wrong - for the most part they don't get feedback on their predictions because they're either dealing with things that are very fuzzy and they can claim "I was close" or they're very long term and they just get lost in time.  By doing quantitative modeling you've had a ton of practice in making predictions of how things should happen in a peer conflict like this and then seeing if your model plays out as expected. 

Ukraine is on track to build 1M drones in 2024 - that's at least 80,000/month, which is why we see them chasing individual russian soldiers around.  They aren't in any kind of short supply and their combat efficiency is high (fewer than 10 drones per casualty, possibly fewer than 5).  Most of the parts come from (drum roll).....China.  It's basically all consumer stuff and the attitude from China is that it goes into a container off to foreign distributors and it's not China's problem as long as their manufacturers get paid.  
Part of the reason the US pays a lot for drones is that they try to not be dependent on parts from China and there's a lot smaller and less efficient supply chain for those.  There's also never been commitment by the US military to buy enormous numbers of them, so there aren't factories to crank out the drone parts in huge quantities at low unit costs.  The chinese drone parts are all intended for civilian/business/hobbyist use and are produced in large quantities with good scale efficiencies.  
As recently as 2019, the US was struggling to keep the cost of a comparable drone at 10x the cost of a Chinese drone because the private commercial base in the US couldn't compete with DJI. https://www.c4isrnet.com/unmanned/2019/09/17/can-the-army-secure-an-american-made-quadcopter/ 
I went and looked up the production dates, costs, capabilities, and quantities produced of all the listed drones to the extent that the information is available.
Aeryon can do DJI knockoffs for what sounds like a few tens of $K.
All of them except the Drone40 are oriented towards local ISR in terms of design and production quantities.  A few of the less expensive (less than ~$50K/unit) have been produced in quantities of several thousand
The older ones (before about 2010) are basically RC airplanes that cost a few $M each for ISR and are basically obsolete now.  If the airframes were still in production you could redo their entire inside for cheap, but I suspect that they're not.  Of those, it looks like only the Puma was produced at scale of more than a few hundred (more than 1000 produced). The Wasp is archaic and has been replaced by the SkyDio X2D, which costs about $10K each without batteries or accessories.  But it's getting there in cost and capability.  Some of them don't even have any more range than an ATGM (Dragoneye, Black Hornet).  The Black Hornet really looks like it's most useful as a fancy mirror for looking around corners or into second story windows - it's only got a 2 km range, so it's about looking around corners in short visual range environments.
The Drone 40 is the only one under $5K.  It's about $1000/each for a 40 mm grenade with rotors and a 60 minute dwell time.  It has a kind of silly design feature - it's made to be launched from a 40 mm grenade launcher.  But it probably is more cost, mass, and capability effective to not include the features that enable that and spend the mass and volume on either more battery or more bang.
So yes, there's been some experimentation with drones by the US, but given what it's been, it's entirely local ISR oriented, and likely limited to special units, given the quantities.  There's been almost no development of drone munitions (other than the Drone 40 and Switchblade), and switchblade was procured in experimental quantities and even "mass" production is ~500/month, as compared to 80,000+/month that Ukraine can use.  There's only just starting to be development of precision munition delivery with drones, and there doesn't seem to be any doctrine developed around it - that will more likely be copied from Ukraine than the other way around.

That list is almost entirely small-quantity, high-cost uncrewed ISR aircraft that function mostly as crewed aircraft substitutes. The principle behind most of them is to spot something so that you can then send some precision strike later (probably from a crewed aircraft, sometimes fired from the drone itself or artillery)  The lower half of the list lean toward smaller things, but still very much ISR rather than munition.  Several of them can carry and fire standard aircraft munitions (in small quantities), and two, maybe three of them are actual munitions unto themselves.  The cheapest thing on the list is about $100K/unit and has a production rate of ~500/month (switchblade).  
Ukraine is making (and using) more drones every month (80K to 100K) than the combined lifetime production of everything on that list.  They're using them in a fundamentally different way than everything on the list - they're taking advantage of what is apparently an infinite supply of of RPG-7 warheads and are replacing the launcher tubes with drones.  It extends their range to well beyond visual and drastically reduces the risk to the operator.  In running out of artillery shells, they've figured out a way to take advantage of something that they can get in large quantity (drone parts) and turned them into personal delivery systems for RPG-7 warheads.  There are other small munitions that they put on them, but it's the same idea - low cost, disposable system that has long-ish range and direct visual feedback, each carrying a munition that can destroy one of anything on the battlefield.  And up until the moment that each drone blows up on something, it's also providing ISR (usually redundant to other ISR drones that are on overwatch to guide and evaluate).  
 
 

Phones have to be tied to the phone network to function as phones.  Drone's don't.  Autonomous drones don't have to be tied to anything.  At what level of chip do you put it in?  The CPU?  I'll tie the "shutdown" pin to an "everything is just fine" input.  The comm chip?  I can burn my own FPGA and know that it doesn't have the shutdown.
And how do you get China to include the feature in their drone chips?
Email spam will end before anybody can make this one work.

A carrier group against commercial drones.
Carrier: $10B
Ten escort vessels at $1B/each
Total Personnel: ~8,000
Daily operational cost: ~$8M (admittedly, they're all getting paid to be on a boat somewhere in the world, anyway)
vs:
Cost of a Shahed drone: ~$50K
Amount Houthis are spending/drone: ~$2K (even if they paid US labor costs, it would be less than $10K).