I think most manned war vehicles are dead now.
We are definitely entering into the era of drone-vs-human warfare.
Drones are cheap and deadly and can be remotely operated, and soon probably operated by AI.
We are definitely entering into the era of drone-vs-human warfare.
Drones are cheap and deadly and can be remotely operated, and soon probably operated by AI.
As seen in the case of Russia’s super drone being downed by their own this week after presumably loosing control, only relying on unmanned vehicles en mass leads to a single point of failure.
Disrupt or corrupt signals, disrupt the entire force.
The US is the force it is because it can do combined arms like nobody else.
Unique capabilities don’t matter if your opponent saturates them with low tech or combined arms.
Look at last weeks Iran ICBM barrage vs. iron dome and David’s sling.
Sure, unmanned vehicles bring unique capabilities to any field of battle, but combined arms is as much the future as it was the past.
I believe the issue was not overwhelming the defences but rather the defences (Iron Dome and David's sling) were designed for short range and slow rockets from Lebanon or Gaza and not long range fast/high rockets from Iran.
It may be that the cost per Arrow missile (for long range intercepts) is $3M as well, so they didn't fire that many of them? https://en.wikipedia.org/wiki/Arrow_(missile_family)
https://x.com/catebrown12/status/1842167731547721788
If there was 180 missiles were fired and 24 got through to targets, that is a penetration rate of 13%.
We don't know if they were let through because their predicted targets or because there weren't enough interceptors or some other reason.
Air bases are pretty hardened targets. Most important stuff is kept in bunkers. They're also huge. The design assumption of an air base is that it's the prime target for rocket/missile/air raid attacks (and Israel airbases have been attacked by rockets in past wars, e.g. in the Yom Kippur war).
I've seen someone review satellite photos from Nevatim that seemed to indicate one or two hits to some structure. As far as we know nothing of value was hit and no one was hurt.
All that said, I wouldn't want to be around when these volleys of ICBMs are fired at you. Definitely not fun. But not a war winner either as long as they're conventional warheads.
This. Israel knew the 3 targets in advance, according to reporting. Any jets or other valuable equipment from the two airfields would have been removed before the attack. It would have been a waste to use Arrow interceptors to protect empty hangars. Also, according to a satellite image of impact craters, many of the missiles hit dirt around the airbase, so wouldn't have been intercepted even if they needed to protect jets.
Defense against ICBMs attack definitely matters, even if they can be saturated it's not that easy, you need launchers and you need to keep them capable of launching. The ICBM defense isn't just about knocking some of those missiles (and you can selectively do that by priority) it's also about early warning. The combination of these capabilities makes a huge difference.
Germany launched 1400 V2 1-ton warhead ICBMs on London and that didn't stop the British fighting.
Both in Israel and in Ukraine those missiles and UAVs are not really a huge factor.
Definitely combined arms is where it's at today, and today attack helicopters have a huge role to play within that scenario, what the future looks like with more sophisticated autonomous machines is hard to predict. Maybe an autonomous bi-pedal robot combined with a UAV combined with an autonomous tank combined with an autonomous jet...
They're launched, they fly, but at some point close to reaching terminal phase, all they do is fall along a pre-calculated trajectory.
Case in point with regards to Iran, a combination of factors played a role in Israel's response:
They knew the launch locations, times and most likely targets.
Hence were able to move out any valuable assets of targeted bases well ahead of time and, had they wanted to, could have even pre-emptively striked the launchers.
The latter again, requiring combined arms and from what I've seen about military ops, there are so many factors to pay attention to with sometimes such little time windows calculated by mission planners, that I don't see all that going full on autonomously for still some time to come.
Check out this video on the 2017/18 US/ coalition response to the Syrian regime perpetrating gas attacks against civilians: https://www.youtube.com/watch?v=Vnbkmi3Iieo
The nominal number of assets involved, the number of types of naval and aerial assets involved, the highly detailed operation plan with a short window for payload launch resulting in a 2 minute window of said payloads reaching their targets simultaneously: That is such a dynamic and complex situation that the reliability required by each component in such a calculation is high.
That's why all of the humans operating these assets receive years of training.
Autonomous vehicles will have support roles or be optimized for high risk or standard, low-complexity scenarios, but I don't see them making huge strides to replace people in a big way in high-complexity, high-impact ops.
But drones are changing warfare. Drone hunting other drones, drone with self guidance - they are being used in battle right now.
500$ or even less is the cost of FPV drone with night camera capable to carry 2kg bomb. Russia is buying soldiers for up to $25000 to be killed by such cheap drone later.
And one regular 155mm shell is 5x more expensive than FPV drone.
For the sake of humanity, I wish that was the case, but it won't be. Defensive drones will exist for sure, but generally the targets will be human or infrastructure, not other drones.
That theoretical highly maneuverable drone with a highly advanced sensor suite isn't going to be cheap either. At which point, what's the advantage? You wouldn't be bound by the number of pilots, but if the drones are too expensive it doesn't matter.
You gave part of the answer right there. Not only would you not be limited by the number of - hard to replace and time-consuming and expensive to train - pilots but you don't run the risk of losing them either.
> but if the drones are too expensive it doesn't matter.
The large expense for fighter aircraft tends not to lie in the actual production costs but in the development costs which are spread over a limited production run. Build more drones and they get less expensive per item. Build enough of them to overwhelm the enemy and you win drone superiority.
Modern fourth generation fighters would be more than sufficient up close. Except that fifth generation fighters hunt in packs and none of them need to be pointed at their target to hit them.
If you're communicating, you are broadcasting your position to the enemy. That means you can be shot down.
If you're in passive stealth mode, you can't detect the enemy and they can't detect you until both sides are extremely close and that means a dogfight.
All of this also ignores the massive advancements in IRST and AI which can likely pick out subtle patterns in radar and other data that normal hard-coded algorithms cannot.
Your side can have another plane paint the enemy plane, so now you can see him without your emitting anything.
If the illuminating plane is a stealth plane, then you've just traded on stealth plane for another. Even worse, there's a decent chance that your broadcasting plane gets shot down, the illumination fails, and the enemy stealth plane still gets away leaving you one plane less for no gains at all.
IMO, submarines - especially nuclear subs probably have decades of life left at a minimum.
But yeah - everything else is starting to look very vulnerable.
Torpedoes are forced to use physical wires for remote guidance because nothing else works.
It would be pretty easy for someone to develop a quiet, long-range torpedo that can be launched undetected outside active sonar range. (Assuming this capability doesn’t already exist.)
Also, SSNs already carry long-range drones called cruise missiles.
Manned subs will take a long time to become obsolete.
To be fair, I’m (partly) playing devil’s advocate, but I don’t think the tech to create a fully or nearly fully autonomous attack sub is that far out. Key point is that you would have to be willing to give it more autonomy than the current crop of UAVs.
As such, only nuclear AI subs may be viable, since that gives the strategic advantage of a submarine that doesn't need to resurface for like years. But putting nuclear reactors in an autonomous AI system sounds like a terrible idea.
Human societies have spent millenia developing the infrastructure to produce reliable/trustworthy soldiers. AI is not nearly at that level in terms of trustworthiness/fault tolerance, even if they will quickly be as smart as soliders.
There are a lot of problems.
Technically, you need the resources of a nation-state to make stealthy ones that can operate at depth. There are no off the shelf systems you can strap together. Salt water and pressure destroys things very quickly. It wouldn’t take long for non-moving propellers to get fouled with debris. If you want a mine that launches a torpedo, I think that already exists. If you don’t need to operate at depth, a surface drone would be sufficient. If you’re attacking ships, aerial drones or missiles are better.
If you’re not a nation-state, diesel-electric submarines have been around for more than a century. They are easier to build and will be for a long time.
Even once you have drones, what are they attacking? The US has less than a hundred submarines.
In my opinion, this is an advanced technology weapon as part of a combined arms strategy. Basically a more sophisticated torpedo.
Not much radio range under 500m of water.
Another way of saying this is "may well not be".