As an addendum, the future of infantry combat is probably the man-portable flak cannon. The US Army developed, a few years back, an airburst grenade launcher that sighted range and told the grenade to explode at that range. The reason for its abandonment was the intelligence of the average G.I. More elite soldiers of higher quality found the weapon extremely useful- it’s still I believe in limited deployment.
A similar weapon with a slightly smarter computer would allow infantry to combat low and slow flying drones. A rangefinder tells the grenade when to explode, distance and speed of target, and the sight tells the soldier where to shoot to lead a moving drone, let along destroying other infantry behind cover, which was the original purpose. The software required seems trivial to write compared to air defense software.
Great article! I was unaware of the existence of "volley sights" and the expectation that infantry warfare would be indirect fire at long range - what a good lesson on the foolishness of predicting the future based on the past.
Probably the best summary of the evolution of the modern warfare meta I've ever read.
In case you wish to discuss more military theory in the future:
Did you notice the extent to which modern military strategy is still dominated by the Kriegsspiel mentality where every faction is a state with clear borders and the objective is to conquer and hold land to extract taxes and feed the army?
However, what we've come to realize in the 20th and especially the 21st century is the power of non-state factions.
A guerrilla organization can coordinate a vast network of operatives that do not appear on maps, be financed by completely legitimate businesses located anywhere (including the target state) and a single terrorist with little to no military training can attack using a car and/or a gun and/or cheap explosives in a crowded street of a capital city's business district and inflict casualty ratios of 1:100 (aside from the political and economic damage).
Of course, the weakness of guerrillas is that they cannot really conquer without losing their advantage - as soon as they do that, they become a state with an army, that can be easily defeated by stronger states with armies - but what if someone managed to come up with a self-perpetuating rebellious organization?
This organization would have to 1) be able to move its members across countries faster than said countries' police and border patrols can get them, 2) be able to sell its assets and buy new ones faster than governments can seize them, and 3) manage to replenish its numbers internally without the need of recruiting.
Discussing this topic could be interesting regarding the theoretical organizational structure of New Lacedaemonia.
One way to reduce costs is to have a key drone with expensive optics, a spotter, then pair it with large batches of cheap drones with inferior optics. So you keep the spotter in reserve to guide the garbage tier machines.
Indeed- but for the spotter to be useful against camouflaged infantry, it needs to follow the sneaky infantry around until the cheaper drones can get there, which makes it vulnerable. Oftentimes the viability of a tactic will succeed or fail on small technical details, as the minute timing of a creeping barrage could make an assault succeed or fail in WWI. The only way to know for sure is to test in combat, so a range of possibilities based on basic principles open up
I've flown drones before so I think it would be pretty easy to have a spotter drone remain at a distance and then send in waves of cheap garbage drones.
The main problems are:
1.) Fuel/Battery power for Spotter drone, needing about 3x flight time of the cheap expendable drones.
2.) Trained operator who is accustomed to intercutting perspective between multiple cameras of variant quality and different heights, angles, velocities of each drone in the formation. (Much harder than it seems at first, organizations will strongly oppose this step because they don't want to empower local operators at the expense of centralized authority).
By “follow around”, I mean remain in line-of-sight, which, if it remains relatively low, opens it to small arms fire, and if relatively high, becomes a target for air-to-air by other drones. Long operational time increases size as well.
I heard the statistic that a small handful of drone operators on either side are responsible for 95% of drone kills. It is similar to the Pareto distribution of effective rifleman. The side that allows authority to devolve to the command of its most effective troops gains a massive advantage. The elite drone operator is probably the best man to give the orders on where to direct artillery, the best man to coordinate combined-arms tactics with other branches, but I suspect it will take a long time for the socialist military to catch up to that fact.
The spotter drone gets you back into the Optics Question. To carry what is needed becomes expensive and vulnerable. Master/Slave configuration requires stability, or at least a reasonably predictable punctuated equilibrium.
Early hacking was one very smart guy using one very powerful computer looking for one specific, valuable target. That evolved into one guy, many computers, many targets. Once the computer component got cheap enough, dumb guy with many computers become the most common "warrior" on that "battlefield." This is what we see today with jeet "hackers" using cheap computers going after many small targets, the aggregate is valuable because the economies of scale come into play. One important evolution cycle in that chain was when command and control servers were replaced by distributed C&C software riding on board the slave device, thereby negating the need for a C&C server, making bot networks almost invisible and mostly untraceable.
Second Nagorno-Karabakh war saw extensive use of drones in the opening days. First example I can think of using the drone bait tactic on air defense.
US army had shotguns as a standard item for CQB, they'll probably just start using them as a short run anti-drone asset until they develop some new over-weight, under-performing, and over-priced super weapon that gets everybody paid back in Washington
Somehow I doubt that good enough AI is here to allow these things to be cost-effective in terms of P(hit). They are all the rage in Ukraine, but I think the version that it straight-up remote piloted is, not the autonomous version. I would have thought EW would be enough to defend against these, but obviously am mistaken.
The remote piloted ones are more expensive than Shaheds. If a smartphone can do facial recognition, a similar chip can pick something that looks like a field gun or tank out of a fairly low res video feed. That this isn’t done is concern over civilians. Shaheds are primarily used to hit buildings now, because buildings have precise GPS coordinates, but fully autonomous weapons, even on a very rudimentary level, make command a bit squeamish. I don’t think anyone has specifically written a program for telling a drone to crash itself into anything that looks like a vehicle or artillery piece, but it is eminently possible.
EW jamming broadcasts the location of the jammer, and an autonomous drone does not get radio signals from anywhere. It’s all on board. The potential for EMP exists, but we get back into cost and the sensor visibility of shooting out EMP pulses.
The cheapo remote piloted ones allegedly doing the carnage in the Ukraine cost a few hundred a pop, but have little punch.
Doing image recognition as your propose is indeed feasible...if you have the time, a stable pic and computing power. Doing that with what hardware you can fit on a drone flying at 100kmh is an entirely different proposition. It's like the self-driving car all over.
A self driving car needs to not crash. A lot easier to tell a self-driving car to crash than to get it to be safe. I have gotten a chance to play with facial recognition apps. They are almost instantaneous.
I didn’t realize you were talking about the “Amazon” cheap drones dropping grenades. For now, they are very useful in an infantry engagement. Over the range at which one can shoot another with a rifle, flying a small slow grenade over the enemy trench and dropping it is quite effective. In terms of discovering and destroying enemy material at a long range, not so much.
“commanders develop reams of weird tactical implementations, which, through experience, they slowly narrow down to a set of game-theory optimal implementations.“
Yes, which is why I paint with a broad brush on how refinement of drone warfare might change the metagame. Number of possibilities based on small tactical questions.
“the vulnerability of radio and GPS communication signals to jamming.”
I forgot to include a section on jamming- currently, jamming is defeated by a similar method to TOTP, where the frequency constantly changes, and your equipment knows what frequency to listen on at any given microsecond by a shared encrypted secret. Or at least that is currently in the works. Don’t know if deployed yet. Flooding the full spectrum with RF jamming prevents one’s own stuff from working, which is why the Russian trucks only do it in a limited area. Direct laser communication between satellite and drone is also a possible, but unexplored, remedy.
“why the US Military puts so much emphasis on land navigation training is because the top brass know that near-peer adversaries such as Russia or China could easily jam GPS signals or destroy GPS satellites.“
The destruction of satellites is an unexplored space- it is likely that they will all end up destroyed in a total war, and then a nations lift to orbit capacity becomes very important. If one can put up more satellites than the adversary has missiles to destroy, and those missiles are themselves scarce and hard to build. The US, and specifically Elon Musk, has a massive advantage in that field.
As an addendum, the future of infantry combat is probably the man-portable flak cannon. The US Army developed, a few years back, an airburst grenade launcher that sighted range and told the grenade to explode at that range. The reason for its abandonment was the intelligence of the average G.I. More elite soldiers of higher quality found the weapon extremely useful- it’s still I believe in limited deployment.
A similar weapon with a slightly smarter computer would allow infantry to combat low and slow flying drones. A rangefinder tells the grenade when to explode, distance and speed of target, and the sight tells the soldier where to shoot to lead a moving drone, let along destroying other infantry behind cover, which was the original purpose. The software required seems trivial to write compared to air defense software.
Yes very smart.
Great article! I was unaware of the existence of "volley sights" and the expectation that infantry warfare would be indirect fire at long range - what a good lesson on the foolishness of predicting the future based on the past.
Probably the best summary of the evolution of the modern warfare meta I've ever read.
In case you wish to discuss more military theory in the future:
Did you notice the extent to which modern military strategy is still dominated by the Kriegsspiel mentality where every faction is a state with clear borders and the objective is to conquer and hold land to extract taxes and feed the army?
However, what we've come to realize in the 20th and especially the 21st century is the power of non-state factions.
A guerrilla organization can coordinate a vast network of operatives that do not appear on maps, be financed by completely legitimate businesses located anywhere (including the target state) and a single terrorist with little to no military training can attack using a car and/or a gun and/or cheap explosives in a crowded street of a capital city's business district and inflict casualty ratios of 1:100 (aside from the political and economic damage).
Of course, the weakness of guerrillas is that they cannot really conquer without losing their advantage - as soon as they do that, they become a state with an army, that can be easily defeated by stronger states with armies - but what if someone managed to come up with a self-perpetuating rebellious organization?
This organization would have to 1) be able to move its members across countries faster than said countries' police and border patrols can get them, 2) be able to sell its assets and buy new ones faster than governments can seize them, and 3) manage to replenish its numbers internally without the need of recruiting.
Discussing this topic could be interesting regarding the theoretical organizational structure of New Lacedaemonia.
One way to reduce costs is to have a key drone with expensive optics, a spotter, then pair it with large batches of cheap drones with inferior optics. So you keep the spotter in reserve to guide the garbage tier machines.
Indeed- but for the spotter to be useful against camouflaged infantry, it needs to follow the sneaky infantry around until the cheaper drones can get there, which makes it vulnerable. Oftentimes the viability of a tactic will succeed or fail on small technical details, as the minute timing of a creeping barrage could make an assault succeed or fail in WWI. The only way to know for sure is to test in combat, so a range of possibilities based on basic principles open up
I've flown drones before so I think it would be pretty easy to have a spotter drone remain at a distance and then send in waves of cheap garbage drones.
The main problems are:
1.) Fuel/Battery power for Spotter drone, needing about 3x flight time of the cheap expendable drones.
2.) Trained operator who is accustomed to intercutting perspective between multiple cameras of variant quality and different heights, angles, velocities of each drone in the formation. (Much harder than it seems at first, organizations will strongly oppose this step because they don't want to empower local operators at the expense of centralized authority).
By “follow around”, I mean remain in line-of-sight, which, if it remains relatively low, opens it to small arms fire, and if relatively high, becomes a target for air-to-air by other drones. Long operational time increases size as well.
I heard the statistic that a small handful of drone operators on either side are responsible for 95% of drone kills. It is similar to the Pareto distribution of effective rifleman. The side that allows authority to devolve to the command of its most effective troops gains a massive advantage. The elite drone operator is probably the best man to give the orders on where to direct artillery, the best man to coordinate combined-arms tactics with other branches, but I suspect it will take a long time for the socialist military to catch up to that fact.
Yes very illuminating perspective.
I was guessing a spotter drone could stay far away while tracking infantry but maybe you are right, I wouldn't know
The spotter drone gets you back into the Optics Question. To carry what is needed becomes expensive and vulnerable. Master/Slave configuration requires stability, or at least a reasonably predictable punctuated equilibrium.
Early hacking was one very smart guy using one very powerful computer looking for one specific, valuable target. That evolved into one guy, many computers, many targets. Once the computer component got cheap enough, dumb guy with many computers become the most common "warrior" on that "battlefield." This is what we see today with jeet "hackers" using cheap computers going after many small targets, the aggregate is valuable because the economies of scale come into play. One important evolution cycle in that chain was when command and control servers were replaced by distributed C&C software riding on board the slave device, thereby negating the need for a C&C server, making bot networks almost invisible and mostly untraceable.
Thank you, good points Outgoing Misanthrope.
Should have put it here first, but motorcycle warfar in action: https://t.me/intelslava/57216
Second Nagorno-Karabakh war saw extensive use of drones in the opening days. First example I can think of using the drone bait tactic on air defense.
US army had shotguns as a standard item for CQB, they'll probably just start using them as a short run anti-drone asset until they develop some new over-weight, under-performing, and over-priced super weapon that gets everybody paid back in Washington
Somehow I doubt that good enough AI is here to allow these things to be cost-effective in terms of P(hit). They are all the rage in Ukraine, but I think the version that it straight-up remote piloted is, not the autonomous version. I would have thought EW would be enough to defend against these, but obviously am mistaken.
The remote piloted ones are more expensive than Shaheds. If a smartphone can do facial recognition, a similar chip can pick something that looks like a field gun or tank out of a fairly low res video feed. That this isn’t done is concern over civilians. Shaheds are primarily used to hit buildings now, because buildings have precise GPS coordinates, but fully autonomous weapons, even on a very rudimentary level, make command a bit squeamish. I don’t think anyone has specifically written a program for telling a drone to crash itself into anything that looks like a vehicle or artillery piece, but it is eminently possible.
EW jamming broadcasts the location of the jammer, and an autonomous drone does not get radio signals from anywhere. It’s all on board. The potential for EMP exists, but we get back into cost and the sensor visibility of shooting out EMP pulses.
The cheapo remote piloted ones allegedly doing the carnage in the Ukraine cost a few hundred a pop, but have little punch.
Doing image recognition as your propose is indeed feasible...if you have the time, a stable pic and computing power. Doing that with what hardware you can fit on a drone flying at 100kmh is an entirely different proposition. It's like the self-driving car all over.
Good point on EW.
A self driving car needs to not crash. A lot easier to tell a self-driving car to crash than to get it to be safe. I have gotten a chance to play with facial recognition apps. They are almost instantaneous.
I didn’t realize you were talking about the “Amazon” cheap drones dropping grenades. For now, they are very useful in an infantry engagement. Over the range at which one can shoot another with a rifle, flying a small slow grenade over the enemy trench and dropping it is quite effective. In terms of discovering and destroying enemy material at a long range, not so much.
“commanders develop reams of weird tactical implementations, which, through experience, they slowly narrow down to a set of game-theory optimal implementations.“
Yes, which is why I paint with a broad brush on how refinement of drone warfare might change the metagame. Number of possibilities based on small tactical questions.
“the vulnerability of radio and GPS communication signals to jamming.”
I forgot to include a section on jamming- currently, jamming is defeated by a similar method to TOTP, where the frequency constantly changes, and your equipment knows what frequency to listen on at any given microsecond by a shared encrypted secret. Or at least that is currently in the works. Don’t know if deployed yet. Flooding the full spectrum with RF jamming prevents one’s own stuff from working, which is why the Russian trucks only do it in a limited area. Direct laser communication between satellite and drone is also a possible, but unexplored, remedy.
“why the US Military puts so much emphasis on land navigation training is because the top brass know that near-peer adversaries such as Russia or China could easily jam GPS signals or destroy GPS satellites.“
The destruction of satellites is an unexplored space- it is likely that they will all end up destroyed in a total war, and then a nations lift to orbit capacity becomes very important. If one can put up more satellites than the adversary has missiles to destroy, and those missiles are themselves scarce and hard to build. The US, and specifically Elon Musk, has a massive advantage in that field.