Monday, May 20, 2013

AAW - Hard or Soft Kill?

The Navy has been firmly committed to the hard kill option of the Aegis/Standard Anti-Air Warfare (AAW) system as the main line of defense against attacking planes and missiles.  Additional hard kill elements include a variety of shorter range missiles and guns such as RAM, SeaRAM, CIWS, and various rapid firing guns in the 3”-5” range.  A secondary, soft kill system of decoys and electronic countermeasures (ECM) is also used but has not been developed and upgraded with the same attention and priority as Aegis and the various hard kill components.  For example, the Navy’s main ECM system, the SLQ-32, is well behind the times, bordering on obsolete, and the Navy is only just now beginning to look at upgrading it.

SLQ-32 - Not Enough Love?

We see, then, that the Navy uses both hard and soft kill systems for AAW with an overriding emphasis on hard kill.  Is this wise?  What is the success rate of each?  Which is more likely to be successful against future threats?  Is the Navy pursuing the best course in AAW development?

We’ve already discussed the historical data regarding anti-ship missile attacks on passively defended (soft kill) vessels.  The data shows defensive success rates of around 80%.

Historically, there have been very few combat AAW missile launches so there is not much of a database to draw conclusions from.  The best data set that I’m aware of is from the Falklands War between Britain and Argentina.  During that conflict, at least 26 Sea Dart missiles were launched.  Wikipedia reports that of 5 missiles launched against helicopters or high flying, relatively slow aircraft, 4 hits were achieved.  By contrast, there were only 2 hits out of 19 launches against low flying aircraft.  The Falklands totals for the Sea Dart, then, were 6 hits out of 26 launches (23% success).  Wikipedia further notes that an unspecified number of launches were made without guidance in an attempt to break up low level attacks as a result of limitations of the missile system.  In addition, the Sea Wolf missile, designed for use against low level targets, achieved 2 kills in 8 launches for a 25% success rate, according to Wikipedia.

Note that the Falklands AAW actions were against planes, mainly, rather than missiles which are a much more difficult target.

To the best of my knowledge, Aegis/Standard has never been fired in combat in the AAW role so there is no direct data to examine.

We’ve also discussed from time to time that Aegis is such a complex system that it has fallen into a fleetwide state of reduced performance and readiness.  The Navy has had to implement special programs in an attempt to bring Aegis back up to standard but the system remains degraded across the fleet.  The complexity of the system largely precludes on-board repairs by the crew and, in fact, makes it virtually impossible for the crew to even spot degraded performance.

What do all of these bits and pieces tell us?

One obvious conclusion is that soft kill methods have a far better performance record than hard kill.  That’s probably not all that surprising given the difficulty of trying to guide an AAW missile on to an incoming, high speed, maneuvering target.  To be fair, that conclusion is drawn with no data input from the Aegis/Standard system.  Will Aegis perform markedly better than Sea Dart?  My guess, based on nothing, is that Aegis will perform better but nowhere near the 80% success rate demonstrated by soft kill systems.

Another point to consider is that any AAW system works best in a fully automatic mode.  Unfortunately, commanders are reluctant to operate that way for fear of unintended mishaps.  Indeed, there have been numerous such incidents.  For example, Sea Sparrow has fired on friendly ships during exercises and CIWS has fired on friendly chaff and helos with each example causing damage and casualties.  Contrast that to soft kill systems operating in automatic mode.  There is no danger.  Soft kill systems can be left on continuously, EMCON considerations not withstanding.

Further, consider the cost of upgrading the capabilities of hard and soft kill systems.  Hard kill systems require software upgrades, which are relatively easy to implement, and hardware (the missiles) upgrades.  The Navy has spent a great deal of money upgrading the Standard missile from Block 1 through the various versions to Block 6.  Soft kill system upgrades, by comparison, are almost exclusively software based, again relatively easy, and the hardware upgrades are far less expensive.  Thus, soft kill systems can be more easily kept up to date and responsive to the ever-changing threats.

Next, consider the real estate required for hard kill and soft kill systems.  Missiles and their launchers require enormous amounts of deck space and internal ship’s volume.   Soft kill hardware requires very little space.  The result is that the smaller the ship, the less hard kill defense capability it has.  To a large extent, soft kill systems are independent of ship size.

Finally, consider the danger that hard kill systems pose to themselves.  The missiles represent a source of internal explosions in the event of a hit on the defending ship.  Soft kill systems present little self-threat.

It’s clear, then, that soft kill systems offer historically superior performance, easier upgrade paths, can be operated more freely, require little space, and present no threat to friendly forces or the defender.  The Navy should be much more focused on soft kill systems than they currently are.

Why, then, is the Navy so focused on hard kills?  Well, for one thing, hard kills are dramatic, definite, and occur at some distance away from the defender.  Soft kills tend not to show a definite result until the missile has approached quite closely and passed the defender.  It’s a lot more reassuring to see missiles disappear at a distance!

What’s the takeaway from this discussion?  I’m not suggesting that the Navy abandon hard kill systems, so half of you can take your itchy fingers off the keyboard where they were preparing to lambaste me.  What I’m suggesting is that soft kill systems are part of a layered AAW defense, along with hard kill systems, but that they should be given much higher priority than they have historically received from the Navy.  As stated, soft kill systems are more effective, cheaper, easier to maintain and upgrade, and safer to operate.  You would think those characteristics would garner far more attention than they do.  C’mon Navy, pay attention to what works even if it isn’t the “sexy” approach.  Soft kill provides more bang for the buck – stop ignoring it!


  1. This comment has been removed by the author.

    1. B. Smitty, a good point. Like any system, ECM and decoys must be used intelligently! Sometimes, the job description of the escort calls for standing and taking one rather than decoying a missile onto the very ship you're protecting.

    2. This comment has been removed by the author.

    3. B. Smitty, great point about the area defense aspect of missiles. Again, though, you recognize that I'm not arguing against hard kill systems; I'm arguing for greater emphasis on soft kill systems than they've historically received. Thanks!

  2. I understand that Sea Dart missile was designed as a medium/high altitude anti-aircraft missile to take out the large Soviet aircraft, eg Tu-95 Bear,launching anti-ship missiles over the North Atlantic, it is a wonder that it hit any low flying aircraft, that was the job of the Sea Wolf.

    1. Anon, thanks for that heads up. I've modified the post to include the Sea Wolf data from the Falklands which was 2 kills out of 8 launches. Thanks!

    2. The argument is to kill a bullet with another bullet! That is hard kill.Having gotten involved in softkill at NAvSEA prior to my retirement I was involved in a debrief about a FFG that was almost sunk. It was said that the ansps49 should have detected
      the target. Subsequent analysis showed the an/sps49 could not have seen the target. Another report was CIWS was down waiting a part. Chaff could not be launched since the tao had not unlocked the chaff launcher. At this point I left the meeting.

    3. I assume you're talking about the Stark. That was a ship that was in an unfortunate position with an improper level of awareness and readiness. No amount of technology can overcome inept operations!

      Did you have a point to make?

      Do you have any soft kill insights to offer from your NavSea experience?

  3. Might I recommend:

    Schulte 1994 An Analysis of the Historical Effectiveness of Anti-Ship Cruise Missiles in Littoral Warfare

    1. smsemden, that report is the source data that Hughes summarizes in his book and which I reference frequently! Schulte's conclusions are identical to mine - not surprising since my conclusions are based on data and logic!

    2. Indeed. It further supports your argument as historically soft-kill countermeasures have either been very effective or nearly useless depending which side of the counter measure versus counter-counter measure update cycle you end up on when the shooting starts. The older your soft-kill countermeasures are the less likely they'll work when needed.

      On the flip side of the same argument, hard-kill systems are less susceptible to the rapid cycle of counter-counter measures. Building an ASCM that flies faster takes more time and money than improving the software.

  4. While I completely agree that soft-kill is often under underestimated, I think maybe you were a bit too hard on missiles by using the Falkland's data.

    As has been pointed out by other people, the SeaDart was never designed for low level aircraft or missiles. However the one type 42 destroyer that had a modern (for the time) radar, HMS Exeter is credited with shooting down two low flying Skyhawks (flying at 33-49 feet). The other ships fitted with SeaDart had a older radar that didn't handle clutter very well at low heights. Exeter also is credited with a couple of other kills and it is generally considered that the newer radar made a big difference.

    Another thing about missiles, is that they are often fired in pairs at the same target. So with ~26 SeaDart launches during the Falklands war, it is very unlikely that they were fired at 26 different targets. In the friendly fire incident where a British Army helicopter was destroyed by HMS Cardiff, it is reported that HMS Cardiff fired two SeaDarts at the helicopter, and I would think that two missiles were often fired at a single target. I understand it is still current standard practice to fire two SeaWolf missiles at each target.

    So should that 23% success rate from the 26 launches really be closer to 46% of targets that were targeted being destroyed?

    So again while I totally agree that soft-kill methods are just as important as hard kill, I wouldn't be at all surprised if modern systems that Aegis achieved ~80+ success.

    1. I meant to add that with the Sea Wolf launches, that although only two aircraft are credited with being hit for sure by them, during the engagement that led to those two Skyhawks being destroyed, another Skyhawk crashed into the sea. Some reports say that it was trying to evade another Sea Wolf missile, while other reports say it hit the debris field from the first two destroyed aircraft. Either way in my opinion it was still a success, just the same as a soft kill diverting a missile would be.

    2. Anon, am I being too hard on missiles? I'm reporting the data that exists. It is what it is. It's a fact. Now, one can explain away or rationalize all the misses and assume that Aegis/Standard will achieve 80+% success, as you suggest, and be quite happy and secure in the knowledge that nothing can penetrate the defense. Of course, when a real shooting war occurs those people will be surprised by the real world performance that will be well below what they anticipated.

      Not a single weapon system in history has performed as advertised but ________ (fill in blank for whatever your favorite weapon system is) will. That's blind faith with the emphasis on blind.

      The reality is that all those missiles I mentioned performed the way they performed. Also, as I pointed out, the majority of the engagements were against slow aircraft. Had those aircraft been missiles from that time period, the performance would have been far worse.

      Aegis/Standard will be matched against high speed, modern, maneuvering missiles with active countermeasures in a confused, ECM filled environment. That does not sound like a recipe for 80+% success. Heck, the Navy's contrived, ridiculously easy test firings are hard pressed to achieve that kind of success rate.

      If you can take all that and project 80+% success for Aegis/Standard then there's nothing more I can say. I guess I'm too hard on missiles!

    3. Maybe I didn't explain myself very well.

      I wasn't saying that I would be quite happy to believe that hard kill will be all that is needed. I certainly think soft-kill is way underestimated. I also think that in most cases if I was given a choice between only hard kill or only soft kill being fitted to a ship I was in, then I would pick soft kill (depending on the expected threat of course).

      What I was trying to say is that I just thought you were taking the data from the Falklands a bit too much on face value and not looking into the details of those engagements enough by just saying things like the Sea Dart only had a success rate of 23% and Sea Wolf only had a success rate of 25%. Without looking into how many aircraft were actually targeted, it isn't possible to really say a success rate. As I said in at least some of the engagements with Sea Dart, more than one missile was fired at the same target. So you can't just say X number of targets destroyed from Y number of missiles fired equals Y/X success rate. If two missiles are fired on one target, then if we just went with targets destroyed versus missiles launched then the best success rate possible would be 50%. That is all I was trying to point out. Now I'm not saying every engagement in the Falklands had two missiles fired at each target, but the data clearly records that at least some of them did have. I believe that in any modern engagement against anti ship missiles, that if possible two missiles would be fired at each approaching anti ship missile.

      With the Sea Wolf thing about the extra aircraft crashing, again I was trying to say that I think without knowing how many missiles were fired at each target, then if you counted that extra aircraft as a mission kill, then Sea Wolf's would have destroyed 3 aircraft from 8 launches, which could be a success rate of anything from 37% to 75%.

      So I wasn't trying to say that I would be secure in the knowledge that nothing would penetrate a missile based defence, I would just saying that I thought the data from the Falklands needed looking at a bit deeper.

      Also by pointing out that HMS Exeter which had a newer radar, had a better success that the other Sea Dart ships, I was trying to point out that the systems on the Royal Navy ships in the Falklands were nothing like as advanced as modern systems like Aegis or Sea Viper. The last type 42 destroyer has just retired from the Royal Navy and Sea Dart was upgraded a number of times since the Falklands. Also while Sea Wolf is still used by the Royal Navy (replacement starts entering service in 2016), it has gone through at least 2 - 3 generations since the early versions used during the Falklands. So I was trying to point out that I wasn't sure how useful it was to look at the Falklands data when considering how successful modern systems would be.

    4. However even if Aegis could achieve 80% success, I would still want good soft kill systems. One of the main problems with all the missile systems in the Falklands is that they weren't working or for some other reason were unable to fire when needed. There was a number of times when the Sea Wolf system just wasn't working when it was needed. There was also a time when one ship locked onto approaching aircraft and then another royal navy ship moved in front of it and broke its lock, so it was unable to fire. So even if a missile system could achieve 80% success rate, I would never want to rely on just that.

      There is also has you said the various ROE etc, in why a soft kill system could be used in situations where hard kill couldn't.

      Another soft kill system that I don't think gets the recognition that they deserve are the corner reflector based decoys. I would feel more secure if a pilot had been tricked into firing his missiles at a false target than I would at having a missile being fired at a ship even if that ship had both good soft kill and hard kill systems.

      I also agree with you that all the engagements in the falklands were against slow flying aircraft, and it would have been different against missiles. As far as I'm aware the only successful combat engagement of a anti-ship missile by another missile was the Sea Dart fired by HMS Gloucester in 1991, that destoyed an Iraqi Silkworm missile that had been fired at USS Missouri. However Silkworms are anything but like a modern sea skimming anti ship missile.

      So I guess I made a unfair mistake by saying you were being too hard on missiles, but did just mean that I thought you were taking the data a bit too much at face value without considering how many aircraft had actually been targeted by those missile launches.

    5. Anon, I completely understand your points about multiple launches at a single target and so forth. Remember, though, that all the factors you're citing that cause the numbers to look worse than they "should" be will still be present in future engagements. Consider an engagement right this moment. Many (most) of the Aegis ships have significantly degraded Aegis performance. So, the "average" Navy ship is going to have "poor" radar performance. Navy doctrine is shoot-shoot-look so at least two shots will be fired at each target. Ships will still cross each other's paths. Some missiles, even in exercises, fail to guide. Confusion will reign (think USS Vincennes). An ECM enviroment will play havoc on guidance. When the engagement is over, the hit percentages are going to be 20%-40%, I'm guessing. On the lower side, I suspect, against missiles.

      The problem with rationalizing and projecting 80+% results is that it leads to a false confidence in the weapon system that may result in failure to put developmental and procurement resources into point defense or soft kill systems which is exactly what I think has happened to the Navy. The Aegis "mafia" has suppressed point defense and soft kill system development and deployment with artificially false claims about Aegis performance. Navy ships have very few point defense weapons on a relative basis and are going to be shocked at the number of leakers in a real engagement.

      Also, don't misunderstand me. I don't particularly care what the success rate of Aegis/Standard is. If it requires four missiles instead of two to ensure a target kill, that's fine. Hit probabilities matter only to the extent that they determine the likelihood of a target passing through the engagement window and, to a lesser extent, they dictate missile inventory levels. As long as we know the realistic probabilities we can plan our layered defense accordingly. I strongly suspect that a realistic assessment will show we need far more point defenses and much more robust soft kill systems. Hence, the point of the post.

      Does all that make sense?

    6. I completely understand what you mean, so yes it does make sense, and I don't disagree with it.

      Of course all the factors that make the number of hits look worse than they really are will still be there in a current engagement. As you know, the main thing I was trying to point out is that some of the misses that you were classing as misses weren't really misses. As if two missiles are fired at one target then they both can't hit it and destroy it. Well most of the times they can't, I guess in some cases they could hit it nearly together, but even they would both be counted as a hit? So in most cases even if they both worked perfectly, one missile would hit and destroy it's target while the other would be classed as a miss.

      And of course the same thing would still happen today. So I guess it comes down to if you are talking about hit probabilities or number of engaged targets destroyed. I was talking about percentage of engaged targets destroyed when I said I think it could be possible for Aegis to achieve 80% success, and not that 80% of the missiles fired would hit their target. Of course 80% of missiles hitting their target is different and better than 80% of engaged targets being destroyed. But even if a missile could achieve a 80% hit probability, due to the multiple launches per target, we still wouldn't ever achieve the 80% of missiles fired destroying their target.

      So I think really we were just both kind of talking about different things, but now understand where each other is coming from.

      However as both of us have pointed out, even if a missile system can destroy 80% of targets engaged then there are still a lot of problems left.

      As I said in the Falklands, there was most likely more targets that weren't engaged that should have been, than there were targets that were engaged but weren't destroyed. None of the exocet missiles that hit a Royal navy ship were engaged. It's impossible to know, but I do think some of those Exocets would have been destroyed if they were fired at.

      Of course a missile is different to the aircraft that were destroyed, but I think there is some room for debate on how much harder a block 1 Exocet would be to hit than a low flying SkyHawk was. Both have a maximum sub-sonic speed. The Argentinian's certainly flew their SkyHawks at low level. So maybe not much difference between the height of some of those destroyed SkyHawks and a Exocet. While as far as I know the block 1 Exocets didn't perform any sort of invasive maneuvers. So while I think there is some room for debate on what the difference in outcome would be, I'm not going to try to make any sort of prediction.

      Even if we could say that x% of exocets would have been destroyed if they had been fired on, I don't think it tells us much about how Aegis would perform against modern anti-ship missiles, as both Aegis and modern anti-ship missiles are a lot different to the missiles used in the falklands

      The same as the 100% success rate of HMS Gloucester destroying a Silkworm doesn't tell us very much.

      I don't think what percentage of missiles that hit or engaged targets destroyed is the most important thing. More important are all those times when for various reasons a target couldn't be engaged. Hopefully we have learned from some of the problems that happened during the falklands, like a ship not being able to guide a missile while it was using it's satellite communication equipment, but there will no doubt be a lot of other problems that stop engagements.

      So I completely agree with you that hard kill systems on their own aren't enough. A layered defense is certainly required. Even then a bit of luck and good training is required. I think it was the report on the sinking of HMS Sheffield that found that although the ship didn't have time to fire any missile, it should have fired chaff but made no attempt to do so.

      Thanks for the interesting debate, I think we are both in agreement on most issues, and really it was just a matter of semantics.

    7. Btw, I think the whole thing in the news today about Russia agreeing to deliver the S-300 to Syria is somewhat related to this.

      In that all the media etc are going on about how the S-300 is undefeatable and would stop any Western/Nato action.

      As we both agree no system is perfect. Without thinking about any weaknesses of the S-300 or the fact that it reported that it is only a single battery that Russia is selling. It still won't work as it is meant to all the time.

      If it came to it, with a single battery there are going to be a limited number of missiles sold. So say Russia sells at most 100 missiles, then its not impossible to just fly ~50-100 UAV's over the top of it and let the battery expend all it's missiles shooting them down. Or rather use the miniature air launched decoys. Quite expensive but certainly not impossible. Also each decoy should cost less than each S-300 missile (~$1 million each)

      Not that I think or hope we will get involved in Syria.

    8. Anon, fair enough. We've worked our way around to substantial agreement. A good day's work!

      Regarding the S-300 effectiveness, all I can say is that no SAM system in history has come close to working as claimed and I see no reason why this one will be different. Is it a threat? Yes. Is it unbeatable? Not even close. Look at the history of Soviet made SAMs. Everywhere they've been used they've met with underwhelming success. The success rate in Viet Nam, for example, was on the order of 1%-5%. I think Desert Storm had an even worse success rate. No reason to think these will be different. Sure, these missiles are newer, bigger, faster, badder, etc. Well, the planes they would be used against are faster, stealthier, and have better on-board ECM along with better escort jammers and potent anti-radiation missiles. In the end, all the improvements balance out on both sides and we're left with basically the same performance. Soviet weapon systems, in general, have demonstrated very poor performance in combat.

    9. During the introduction of SSDS on LSD's 2000 we could not convince ship CO's to operate in Auto engage. Even after demonstrating be in auto or die. Unlike Aegis the CO's did not trust SSDS or their crew given their lack of AAW training.
      Another point to consider is that any AAW system works best in a fully automatic mode. Unfortunately, commanders are reluctant to operate that way for fear of unintended mishaps. Indeed, there have been numerous such incidents. For example, Sea Sparrow has fired on friendly ships during exercises and CIWS has fired on friendly chaff and helos with each example causing damage and casualties. Contrast that to soft kill systems operating in automatic mode. There is no danger. Soft kill systems can be left on continuously, EMCON considerations not withstanding.

    10. SSDS had, and continues to have, significant problems including poor sensor location, blindspots, sensor limitations, ESSM issues, tracking, target identification, etc. I'm not surprised CO's wouldn't trust it. I don't know the status of the system today. Of course, this is only based information in the public domain. There are, apparently, other problems that are classified and classified reports have been issued to Congress.


Comments will be moderated for posts older than 7 days in order to reduce spam.