116 comments

[ 3.2 ms ] story [ 207 ms ] thread
Not the most accurate headline, but it's about time that we started implementing this at the level of individual tanks and APC's, not just warships.
It isn't US Army tech, it is mostly Israeli army tech that was developed to manage the Palestinian population in Gaza and the West Bank.
There are two Israeli offers, one American and one German. That's just 50% subjugation of Palestinian territories. Get your facts straight.
The article actually mentions this (and an Israeli defense contractor is a one of the few that's lobbying for it.) I think the key difference, from my understanding, is that it has never been deployed onto a convoy. Also the "Iron Curtain" relies on the projectiles being lobbed instead of shot directly, since the people shooting those projectiles into Israel don't care who they hit military or civilian. It has been interesting watching defense move along, from the Patriot missile program, to the Iron Curtain, and now something that is mobile and can take on smaller projectiles.
I think you are confusing Iron Curtain with Iron Dome. Iron Dome is the area anti rocket / mortar (lobbed) system. Iron Curtain is a system that protects individual vehicles from "shot directly" projectiles like RPGs.
>to manage the Palestinian population in Gaza and the West Bank

>manage the Palestinian population

Not sure if that euphemism is beautiful or really horrifying...

Being under constant threat helps to develop defensive technology. Israeli developers have very tight feedback loops. And they see first hand what happens when they fail.
There's 3 vendors working on this.

A (older) video of one of them in action, blocking an incoming RPG headed towards a Humvee: https://www.youtube.com/watch?v=n_yz_ONZltA

Russians have had it for a while. It's called Arena:

https://www.youtube.com/watch?v=YpmcmKwWzYo&t=1m36s

Yup, and Israel has had Rafael's Trophy system deployed since 09 - https://www.youtube.com/watch?v=OtMgnRMIspQ
Iron Curtain and Quick Kill (not mentioned in the article) have been in development since 2006.

The idea and the capability is nowhere near as new as the article implies. The slow adoption my the US and Israel is motivated more by the collateral damage concerns that caused the Russians to scrap the Drozd system in 1980. As far as I know, Arena has similar issues, just not nearly as badly as Drozd (which basically shot a 107mm fragmentation grenade at the incoming round).

This is just the extreme miniaturization of counter battery fire. CIWS for tanks.

My mind immediately thinks of the poor dismounted infantry near this thing when its engaged.

We'll quickly be approaching the debate similar to the self driving car. Does it crash and kill the driver to protect pedestrians. or vice versa. Protect the men or the tank?

The military is better at resolving these sorts of issues. Tanks are a force multiplier, Keeping them in the fight is important (worth the collateral?). But cheap options work aswell. Plenty of men alive and LAVs still rolling along thanks to bar armour.

It's not exactly safe standing next to a tank hit by a rocket anyway. So, active defenses are kind of a wash in terms of infantry.
But you really want to avoid discouraging your infantry from being in the vicinity of your tanks, they are a key part of a combined arms team, critical in the field to keeping the tanks alive with their better viewpoints. In cities, where the infantry play the key role and use tanks as mobile pillboxes, you certainly don't want to discourage them from calling up tanks as needed....
It's not safe standing anywhere near a 60ton vehicle with a 120mm cannon that has huge muzzle blast. Throw in shitty visibility for the driver, so dismounts end up being a bit crunchy. The ironic thing is dismounts are necessary for protecting tanks from hunter/killer teams.
It's not m safe standing by a tank anyway. Integrated light/heavy tactics have changed since Full Metal Jacket.
Except they haven't, really. One of the modifications added to the M1s after experiences in Iraq were infantry phones. The Marines bolted infantry phones on to their M1s almost immediately after the war started.
Grunt phones were a thing long before GWOT, it isn't even a case of reinvention - an old timer noticed the missing phones and corrected the problem. I don't know if the capability was intentionally dropped and then reversed, but if it was then this exact same thing happened in Vietnam with cannonless F-4s - because dogfighting was dead...
It was an F-4 cannon deal with the grunt phones. When the M1 was designed the phone was seen as superfluous, so it wasn't added. The Army thought it was obsolete.

I'm actually surprised the Marines didn't fit the phones as soon as they transitioned to the M1. I'm guessing they didn't have budget, since they were getting post-Cold-War Army hand-me-downs.

The problem could have been headed off if one of two things happened on the USMC end: consideration given to the fact that the Army doesn't use the M1 the way the Corps does, or if armor played a bigger part in combined arms exercises. The first case is a glaring oversight, the second is understandable - tanks are expensive and dangerous, the more time you spend around them the higher your peace time training death toll will be. I don't remember ever training with any armor outside of LAVs (or the humvee that would notionally be a tank before transforming into a notional CH-46).
Ukraine uses active protection on tanks in war against Russia. It works well: http://wartime.org.ua/uploads/posts/2015-04/1427962059_7.jpg .
Are you being sarcastic, or is that actually a good end result for one of these systems? Genuinely curious.
That's a good result. The APS system is usually paired with reactive panels that explode in the face of the ATGM. This looks bad on the turret, but the turret is intact, and the tank still operational. I'd hate to be a dismount nearby when that goes off, but if your goal is protecting a multimillion dollar tank, that's the dues you have to pay.
That's a very good result.

This is what a bad result looks like: https://www.youtube.com/watch?v=MJwkGy6EOj8 [NSFW]

When the ammo and fuel of the tank burns up, that's called cook-off. You can't repair that, and unfortunately the crew will not survive.

Thanks for the vid. Is there a specific reason why the protection system doesn't detonate the projectile further out to avoid any damage whatsoever?
The picture above looks like it's reactive armor, which is slightly different. There are multiple versions around, but the core idea is to sandwich explosives between metal plates in such a way that a round striking the assembly triggers the explosive, which drives the plates tangent to the round in a way that can disable it.
Ukrainian Nizh tries to fragment incoming missile or shell wit cumulative charge. It allows to be effective against broader range of threats.
(comment deleted)
Magazine fires do not go over well with the crew. So yes.
They do use Kontakt reactive armor but what's the APS?
No, Ukrainians developed their own active protection because they had no money to license Kontakt and it is quite different. Kontakt fires metal plate against shell to stop and detonate it, Nizh is cumulative and tries to fragment incoming shell.
These are not active protection systems though. Also, I'm pretty sure that they have at least older generation Kontakt ERA that dates back from the Soviet Union in use.
' “Dozens of threats were launched at these platforms, many of which would have been lethal to these vehicles. Trophy engaged those threats and defeated them in all cases with no collateral injury and no danger to the dismounts and no false engagement,” the DRS official said.'

The RPG-7 travels at a maximum velocity of 295 m/s. Fired from 200 meters and assuming linear acceleration, that means you'd have roughly 1-2 seconds to identify it, calculate trajectory and fire the interceptor.

Compare this to intercepting an ICBM, which travels at a maximum (re-entry) velocity of 7,000 m/s. Fired from several thousand miles, you'd have about 30 minutes to identify it, calculate trajectory and fire the interceptor.

The two systems are definitely not in the same league (agreed!), if that's what you're pointing out.

That said, the time required to calculate an intercept point is not the hard problem, so 1-2 seconds versus 30 minutes is not really super relevant.

Arguably more relevant factors are detection, precision, accuracy and ultimate intercept effectiveness.

In the case of ICBM you're essentially trying to hit one bullet with another bullet... hundreds of miles away. It doesn't matter that you have 30 minutes to think about it. It's still ridiculously hard to actually know there's an icbm, where it is, and then be able to hit it with another missile (and have that make a difference).

RPGs around a armored vehicle are way more straightforward by comparison. The projectile is easier to detect, it's a much smaller volume of space (easier to scan continuously) and the projectiles are traveling slower!

Still I imagine it's quite a trick detecting incoming then swiveling, aiming and firing (accurately) your defense auto-cannon all in 1-2 seconds. It would be very impressive to see that in action.

Yeah, for me, it's the intercepting projectile. One is a bullet fired from auto-aiming system, the other a rocket with a guidance system. Bullets vs. rockets definitely not the same thing.

Still, the performance of Trohpy is impressive. And, it has actually been proven to work in combat, something we pray to never see with ICBMs.

Heh. The missile defense systems are super controversial even now exactly because we have been unable to effectively demonstrate that they work[1].

This stuff sounds like it already works in the field. Nice.

[1]: Tests are run and performance is usually quite horrible, until they make them way way easier than real world conditions. https://www.google.com/search?q=missile+defense+effectivenes...

You don't have 1 cannon. Better option is to have round battery with mini rockets on top of the vehicle. Your reaction time will be milliseconds for the first target. And you could intercept two targets if they are offset enough.
The Trophy and other APSs don't really have to deal with countermeasures either. Anti-ballistic missiles that try to intercept the warheads as they're on their way back down have to deal with loads of decoys that can be packed into an ICBM vs a single rocket with very limited movement and a much more well defined final target (ie a rocket that isn't going to hit the tank isn't much of a concern vs a warhead that could be aimed anywhere in the country).
There's also the element of unit cost. An anti ballistic missile system can easily be worth spending billions on even if effective against low numbers of weapons thanks to North Korea. An anti RPG system is going to be protecting against hardware valued in the hundreds of thousands of dollars.
The other issue with intercepting an ICBM is that the ballistics are much more difficult. You have to factor in the curvature of the earth, atmospheric conditions (is there a storm front between the interceptor pad and the missile?), and the changing nature of the projectile (is it a MIRV? some missiles can jink now, how do you handle that?).
One of my friends is a rocket scientist (aerospace engineer). I asked him about kill vehicles once, and he said that intercepting Russian ICBMs is a losing battle even if you can identify the ICBM and intercept it reliably. Russian ICBMs have inflatable decoy payloads in them that are balloons for all intents and purposes. Since ICBMs deploy their payload while in space, where there is no atmosphere, the decoy balloons behave just like the actual nuclear warheads, despite the huge difference in mass. Once they hit the atmosphere the decoy payloads slow down quickly, however you should target the ICBM and its payload much earlier than this if you want to reliably defeat it. The decoys are much less expensive than the intercept vehicles, and the balloon decoys are light, so Russia can afford to put a whole mess of them in each ICBM.

*edited for clarity.

It's still a fight worth fighting.

Multi-layer defense works. Yes, yes, it sucks when a city gets nuked, but things go way worse if you just give up on the problem and thus allow 1000 cities to get nuked. Think of it like computer network defense. You don't just give up because there is a chance of failure.

You start with treaties, export control, diplomacy, and a threat to respond in kind. Next it's sabotage, both physical and cyber. Next it's launch-phase attack... and if they move their launch sites to a less-desirable location to avoid your attack, it's still a partial win. Next it's an attack right before engine cut-out, possibly from orbit. Next it's an attack on the warheads in space; those balloons sure are vulnerable to lasers. Next it's an attack right after reentry. Next is building codes, bomb shelters, and emergency services.

Every layer has value, even if imperfect. It's negligent to leave out even one layer.

New Russian (and Chinese by extend, the tech was sold about 10 years ago) warheads are also maneuverable on re-entry.

    > * It's still ridiculously hard to actually know there's
    >     an icbm,
    > * where it is
    > * be able to hit it with another missile
    > * (and have that make a difference)
Are these tasks all equally difficult, or is one part of that substantially harder than the others? I had (naively?) assumed the first two pieces would be reasonably easy.
Something inbound to a ship that doesn't support aircraft can be assumed hostile meriting a violent response with reasonable safety.

The same assumption cannot be made of objects inbound to a tank passing through a village with playful children.

The technology does not destroy the attacker but the projectile.

Also I assume objects travelling below a certain speed or heat threshold are not "processed" so most children or children-thrown objects would not be targeted.

>so most children or children-thrown objects would not be targeted.

>most

Great/s

This seems like a willful misinterpretation of the comment.
Do not mind the naysayers, they may be vocal now but Tsahal will also deal with them with equal retribution once the constant threat of palestinians is eradicated.
...so next step would be strapping anti-tank mine to heavy duty quadcopter?

Another possibility would be adding stealthy nose-cones to RPGs. Then you have to increase radar power of the defending system. ...which would help enemy artillery to target the tanks.

I'm not saying this system is nullified. Just that there will be ways to bypass it.

I strongly believe that the next step in asymmetric warfare will be the underdogs using swarms of drones where most are cheap, harmless "toys" to turn it into a numbers game to figure out which 1%-10% are full of explosives. As the cost of small drones drop, you'll potentially be able to drive up the cost of counter-measures dramatically with lots of dummies.
I'm not sure decoys are worth it really. Conventional explosive are cheap. Very cheap. You might as well stuff them into every "thing" you deploy.

It's different with ICBM's.

But yes, I think you have a point in that with COTS robot technology, it should be very cheap to produce swarms of simple missiles. If the adversary answers with "military-grade" countermeasures, he'd go bankrupt..

As someone pointed out elsewhere, though, drones that can carry enough explosives, on the other hand, are not cheap.

But drones that can be made up to look like they are drones that can carry enough explosives when seen from a bit of a distance is another matter.

The payload is cheap compared to delivery and control.

If you have a type of drone that can carry a useful payload (shaped charge powerful enough to disable a tank, in the ballpark of 5-10 kg, is not that feasible for a small drone - but a something like a frag grenade may be feasible), then you might as make them all real, the dummies would cost almost the same.

There was recent news about drone that could carry full sized movie camera. Price was few thousands for the drone.

If you go handgrenade size, the drone costs just bit over 100$.

A frag grenade costs something like $30 (and that's with military procurement overhead that would also raise the drone price to $300; mass produced third world grenades cost $5-$10 to their armies); a detonator, explosives and ballbearings for a DIY modification to that drone would be much cheaper.

Anti-tank shaped explosive charge would be hundreds of dollars but again, a fraction of the drone cost.

The dummies don't need to be a type of drone that can carry a useful payload, that is the point. They just need to sufficiently resemble one for long enough that you have to shoot them down just in case.
The radar signature of a small projectile coming in at, say >30m/s is indeed a lot different from a running, tiny human. What about something like a soccer ball, though? It's a small-ish projectile, coming in quite slowly. What about a hand grenade?

Hopefully they test for this stuff and have a firm idea of where they want their thresholds but, then again, sometimes bugs happen. https://www.wired.com/2007/10/robot-cannon-ki/

I get the feeling that destroying a kids soccer ball thrown at a tank is the least of the kid's worries in that situation. I would assume it would only go active with the permission of the people in the tank.
These systems are set to ignore anything moving slower than 50 m/s or so. Even professionals can't kick a ball that fast. The lower limit is specifically set to avoid triggering the system on rocks, falling debris, birds, people, and anything else benign that might be moving in the area.
The problem is that a thrown hand grenade is detectable by the system, but well under the limit.

One publicized case of a soldier in a vehicle "protected" by this system but killed by a tossed hand grenade or hit by a grenade launcher from long range (ordinary 40mm grenades are only fired at about 75 m/s at the muzzle), and there will be calls for this value to be adjusted.

The system is also capable of detecting firearm projectiles, but is specifically tuned not to react. Regardless of what calls may come, there won't be a change because the designers and the customers have already considered that possibility and specifically excluded it as not being worth the risk. This is not a case of being too cheap or short-sighted to do something.
The system rightly should not react to tossed hand grenades or ordinary 40mm grenades, just as it shouldn't react to rifle rounds or thrown rocks - those grenades have no chance of penetrating the tank's armor anyway.

If soldiers are exposed, then such a system reacting is a greater threat to their lives than a hand grenade; triggering the system is quite likely to harm nearby troops.

These systems don't try to deal with grenades which aren't really a threat to anything with armor so the code can ignore anything that people could throw/kick really. What the APSs are designed to catch are RPGs and other anti-tank rockets which all travel >100 m/s which is well above what a playful kid can make.
rockets which all travel >100 m/s which is well above what a playful kid can make

Speak for yourself. I loved rockets[0] as a playful kid!

[0] https://www.estesrockets.com/

Even without this new system, a person setting a toy rocket off near a deployed military force is really asking for trouble.
I disagree. It is the deployed military force that is asking for and finding trouble. It is the innocent who bear the brunt of the danger for this. This is the moral hazard[0] of war.

[0] https://en.m.wikipedia.org/wiki/Moral_hazard

It's like playing airsoft in the middle of a police parade. You're going to have a bad time, and it's your fault for being a dumbass.
Possibly, though I bet the return from the cardboard/cardstock that most of those are made from are very different from the radar return from a metal missile.

I played with Estes rockets too. How many people are likely to in war zones like these APS devices will be deployed in are likely to though? The rocket motors aren't super cheap and aren't really reusable or easy to make your own.

The G in RPG is Grenade.
I think RPG actually stands for ruchnoy protivotankovy granatomyot

'rocket propelled grenade' is a backronym.

Uh, what do you think "granatomyot" means?
What do you think the rest of it means? According to Wikipedia[1], the phrase "ruchnoy protivotankovy granatomyot" means "hand-held anti-tank grenade launcher".

[1]: https://en.wikipedia.org/wiki/Rocket-propelled_grenade

There are hand-held grenade launchers, crew served grenade launchers, etc. None of which means it's a "backronym."
Sure it does. The acronym RPG literally comes from the phrase "ruchnoy protivotankovy granatomyot", which was the name given to a series of Russian weapons in the category of portable anti-tank rocket launchers. And this name definitely pre-dates "rocket-propelled grenade", and in fact predates the weapons being rocket-propelled at all. For example, the RPG-40 is simply an anti-tank hand grenade (https://en.wikipedia.org/wiki/RPG-40), or the RPG-43 (https://en.wikipedia.org/wiki/RPG-43) and RPG-6 (https://en.wikipedia.org/wiki/RPG-6).

Actually, looking at these sources, the non-propelled versions were actually called "ruchnaya protivotankovaya granata" ("hand-held anti-tank grenade"). Still the same acronym of course, it just doesn't include the launcher. So when the launcher part was added, the phrase changed slightly but the acronym remained the same. And it was only at this point that the phrase "rocket-propelled grenade" could be used to refer to this weapon, and the phrase was presumably created to match the acronym RPG.

The RP in RPG is Rocket Propelled. /pedantry

There's a huge difference between RPGs and hand thrown anti personnel grenades. The first uses a shaped charge to create a very directional blast of fast hot metal that can punch through armor and the latter is an unfocused omnidirectional blast that can't get through light armor.

"The Defense Review report also says that Iron Curtain’s sensors can target destroy approaching RPG fire to within one-meter of accuracy."

Can anyone translate that into English?

"The Defense Review report also says that Iron Curtain’s sensors can target destroy approaching RPG fire to within 3.28 feet of accuracy"
I feel like reality is finally catching up with David Drake. Now we are just waiting on 400kg fusion plants and hover tanks.
I had the same thought. I seem to recall that they could also link the entire brigade together into one giant air defense network. I wonder if they'll do something like that for the anti-missile systems, maybe at a platoon level. Or maybe just have the next tank in line automatically kill the attacker.
Indeed, all the "tribarrels" (equivalent to the M2 Browning used in Vietnam (or, rather, back from 1933 to today, it's a keeper)) could be/were slaved to the anti-artillery and air network, and could be automatically taken away from their operators to stop an incoming barrage.

They had proximity fuzed Claymore style anti-"buzz bomb" defenses, but arming those was problematical for a zillion reasons, the best defenses were as they are today, sharp eyes and trigger fingers. Later killing a buzz bomb team that killed a super expensive tank and its crew was not considered an acceptable exchange.

I read elsewhere that a principle challenge of this tech is not to kill your own people who are near the protected vehicle.
I don't see the problem. Self-LART's have a long and colorful history.
I'm very skeptical of this technology. The Russian equivalent is currently deployed on T90s in Syria, but has been shown useless against rebel ATGMs. Ex - https://www.youtube.com/watch?v=2rfyeR-YaJw [War footage, crew survives but maybe NSFW depending on your sensitivity level.]

The tank was disabled but not fatally (it was later repaired and put back in service), but that's just because the T-90 is a great tank and the shot was unlucky. You can see the damaged reactive defense system on the right from after this hit: https://pbs.twimg.com/media/Cd_gdCnWoAA5tyN.jpg

They said it was because the hatch was open, but I think it's actually because this is pure snake oil. It's just too hard.

Success is a combination of idea and right implementation. Maybe T-90 suffers from problems with second.
Yeah, but I think that's going to be the case for the foreseeable future for all vendors. They can all make compelling pitches and demo videos, but as far as I know there's no evidence of any of these systems actually working on the battlefield.
Israel's shot a bunch of projectiles out of the air after detecting and calculating their trajectory in flight.

What makes smaller projectiles much harder?

Occlusion. I watch a lot of ATGM videos, and most of the shots are when tanks are parked next to buildings, driving through an urban environment, or perched on a hill, rather than in open terrain. There is very little time to react no close range sensors.
Israel much more advanced version worked pretty well in the recent wars in Gaza.

Such active protection systems is the future of combat. Beside tanks use case, a helmet mounted radar today is able to detect the incoming bullets, so similar system can be built for human body protection. At another scale - a military plane today do see incoming missiles, so similar active protection can very effective there too.

The development of those systems is electronic tech dependent - one need fast precise radar and fast precise calculation for the 2-3 sequential stages of the defense projectile boost as the incoming projectiles of interest are starting from the Mach 2 and all the way to about Mach 6 (RPG with its less than 1 Mach is not that important as new systems like the Israel's one do handle them easily, it is more of a matter of whether your military bought it or not). With modern electronics and computation power the interception of projectiles with up to 6Mach (Russians declare up to 10Mach on Armata tank, though one yet to see it :) becomes real.

The SHTORA on that tank was not turned on. You can tell by the two shtora lights not glowing bright red.

As to why it wasn't on, we have only speculation.

SHTORA is a an old shrapnel system and it causes a lot of collateral damage to infantry around as it was found in Afghanistan war and soldiers (not tankists :) disliked it.
Off topic but I found it disturbing the way they kept saying "Allahu Akbar". It reminded me of Frisky Dingo https://www.youtube.com/watch?v=LmnaeU0k80I.
It is very commonly used to express astonishment or surprise, the same way we say "Oh my god!" in English.
That makes a lot more sense than the literal translation.
Yep, this is commonly misunderstood by non-Muslims.
Things like this, along with the new suite of drone modules capable of jamming/destroying radars, seem to be creating an atmosphere where our military can operate with relative impunity. That is not, in itself, an ethical problem.

If we can kill 1,000 people with 0 deaths, however, the ethical bar for how we apply that force should be held to a stricter standard, as the slope to fascism is being made steeper. Unfortunately, the recent trend of extrajudicial, extralegal executive action 'war' declaration post-9/11 appears to be lowering that bar.

Even if you think there is a moral imperative to every one of our military deployments: having no skin in the game also makes it easier to demonize us, to the benefit of anti-US propaganda.

I think with a situation such as this it may make the prioritizing of only damaging enemy combatants, of course this is suggesting the term enemy combatants being based around people engaging directly at firing weapons towards those that the us military or those who are employed by the military are in charge of protecting or themselves.
> If we can kill 1,000 people with 0 deaths, however, the ethical bar for how we apply that force should be held to a stricter standard...

What do you mean "if"? We've already killed thousands of people with drone strikes. And sadly the ethical bar on that is not held to a stricter standard, indeed it's held to a lower standard. Some people are targeted based on social networking algorithms that are hugely flawed. The ratio of "collateral damage" deaths to "good kills" is perhaps as low as 50/50, maybe even worse, since there's very little investigative followup.

Mostly this seems to advocate the US adopting the Trophy system from Rafael although it does mention there are other vendors in the picture.

The fundamental point, which is also made in Dunnigan's "How to Make War" is that the cost of fortification goes up faster than the cost of attack capability, so you can only match scale if you use two attack systems. Or to put it in terms of this article, if you added 50 tons of armor to a tank it might be impervious to RPG fire but also unable to move effectively.

The cost of building a guided munition has gone down with the miniaturization of sensors and compute power, and a number of field deployed systems (like Iron Dome and Trophy) have demonstrated that calculating intercept solutions has given the designers of such systems the ability to defeat or at least mitigate very near threats.

But given that I wonder just how many "army on army" sorts of wars will be waged. Looking at the tactics of modern day insurgencies they hit civilians rather than military forces.

Back in the day there was a front line of multiple opposing armies engaged in control over a battlefield while their aircraft went behind the lines and bombed the cities to degrade infrastructure. Today there is no "front line" just bombs going off in cities degrading infrastructure. And no visible army (well its fine when its attacking but between attacks it blends in).

As a result, the militaries of the world are developing strategies for moving around killing people without themselves being at risk. This tank strategy is already deployed in Israel. The ultimate answer to that seems to be remotely operated robotics. No threat to the operator, and equipped with a variety of offensive weapons to dealing with threats to others.

I am not looking forward to the day where you have a self driving tank which is impossible to kill. But when I look at the road we're on, I see that we are headed to that place.

There is a saying: "you don't put a $100 lock on a $50 bike". I think the same applies here.

If to block a $300 RPG (hypothetical cost, i don't know what the cost is), costs $10000, then they still win.

Not if blocking a $300 RPG for $10000 prevents $100000 of damage to the intended target of said RPG.
Yes, from the perspective of the defender it's better. I am only saying, that for the attacker, it only means winning less, as opposed to not winning.
An experienced tank crew is worth more than a tank.

One of the main reasons that Israel has developed so many defensive technologies is that it understood that keeping your experience and trains personnel alive is the only way to win a war of attrition no matter how much man power you can muster to replace casualties.