Thursday, December 22, 2016

Sensor Attrition

We’ve previously discussed that sensors are more important than weapons (see, “Weapons Don’t Matter”).  It doesn’t matter what kind of around-the-world range your weapon has if you can’t find a target for it.  The challenge, of course, is to get the sensor to a useful location which, almost by definition, means in enemy controlled air/land/water space since that’s where the targets of interest will be.  Actually, that’s not the real challenge, is it?  Getting the sensors to a useful location is doable.  The real challenge is getting them to survive long enough to conduct useful surveillance and transmit the targeting data back to attack units. 

Before we go any further, let’s think about what kinds of sensors we’re talking about.  The US military’s main means of generating long distance targeting data is through the use of aviation platforms, both manned (P-8, E-2, EP-3, AWACS, etc.) and unmanned.  The problem with manned platforms is that they are hideously expensive, defenseless, non-stealthy, and slow.  That combination of characteristics means that they won’t be risked penetrating deeply into enemy territory to find the kind of targets we want to find.  A notable exception might be the F-35.  It has the stealth and range to perform moderately deep penetration targeting.  What it lacks is the kind of wide area sensor coverage that other manned surveillance platforms have.  It’s just not possible to put that kind of sensor on a fighter sized aircraft.  Still, in sufficient numbers, it might prove useful in that role.  None of us know exactly what kind of wide area search capability the F-35 really has, if any.  The down side to using the F-35 as a penetrating sensor platform is that every F-35 dedicated to that role is one less for the crucial air superiority battles that will be going on concurrently. 

Increasingly, therefore, the military’s emphasis is on unmanned platforms (UAVs).  There are two problems with using UAVs for deep penetration surveillance and targeting.

  1. UAVs are quite small in terms of volumetric and weight capacities and just can’t mount the kind of long range, wide area sensors that manned aircraft can.  This can be partially offset by the long endurance of UAVs which allows for greater coverage even if the sensor field of view is somewhat limited.  This leads us to the second problem.

  1. UAVs are slow, not particularly stealthy, and not very maneuverable.  In short, they’re not survivable in enemy airspace.  A sensor that can’t survive long enough to accomplish its surveillance is useless.

So, how can we conduct successful deep penetration surveillance and targeting?

In order to answer that, let’s briefly consider what we can’t do. 

  • We can’t use P-8’s.  They’re large, slow, and non-stealthy.  Combine those characteristics with the beacon like nature of their radars and they’re, literally, flying targets waiting to be destroyed long before they can find a target.

  • We can’t use large, expensive UAVs because deep penetration surveillance, by definition, will be dangerous and have a high attrition rate for the surveilling aircraft.  We simply won’t be able to afford to routinely lose $100M+ UAVs.  Heck, we balked at $1M LRLAPs so we certainly aren’t going to buy enough $100M+ UAVs to fill the deep penetration role!

So, what does that leave us?

That leaves smaller, cheap UAVs.

Wait, didn’t we say that small UAVs just can’t carry large enough and powerful enough sensors to effectively conduct wide area surveillance?  Yes, we did say that.  However, if we use enough low effectiveness UAVs we can cover the desired area even if each individual UAV is only marginally effective.   Coverage is provided by numbers rather than individual capability.

Wait, didn’t we also say that UAVs are not survivable?  Yes, we did say that.  However, if we use enough non-survivable UAVs, a sufficient number will survive to do the job.  The enemy has only a limited supply of SAMs, fighter aircraft, and air to air missiles at any given location and point in time.  To make the point with a ridiculous example, if we sent 1000 UAVs into an area, the enemy just wouldn’t have enough weapons and aircraft to respond to all of them and even if they did, they wouldn’t have enough time to hunt them all down individually before the survivors accomplished their mission. 

Of course, the key to both of the preceding points is that the UAVs must be cheap – cheap enough to flood the area and attain coverage and cheap enough to absorb the expected losses. 

Can we build UAVs that are effective enough and cheap enough?  Well, that’s the question, isn’t it?  I believe we can if we ruthlessly focus on what the actual requirements are rather than start loading up the UAV with “wishes” that would turn it into a strike/fighter/ISR/refueler aircraft, all in one.

Currently, we have small, cheap UAVs.  They are many different “brands” in use.  A typical example is the Boeing Insitu Scan Eagle.  Scan Eagle is 4.5 ft long with a 10 ft wingspan and weighs 44 lbs.  It has a max speed of 82 kts and an endurance of 22 hours.  A cruising speed of around 50 kts gives a theoretical range of 1100 nm (radius of 550 nm).  The practical range is currently limited by the communications package which is good for only 62 miles.  The UAV is launched from a miniature catapult and recovered by a shyhook.  Launches can be from vessels as small as a Mk V Special Ops boat.  Payloads include EO/IR or a mini-Synthetic Aperture Radar.

According to Wiki, the Royal Australian Navy tested a Scan Eagle with a Sentient Vision Kestrel Maritime ViDAR high resolution digital video camera that is claimed to be able to cover 13,000 square nautical miles over a 12-hour mission (1).

So, with some improvements in communications range, there is no reason why such a UAV could not fill the need for a small, cheap, effective surveillance UAV.

The cost of the Scan Eagle is listed by Wiki as $3.2 million (2006) for a system consisting of four UAVs, a ground control station, remote video terminal, the SuperWedge launch system and Skyhook recovery system (1).  Of course, once the control station and other equipment is purchased, additional UAVs can be bought for a fraction of the complete system cost.  Aviation Today website lists the cost for the UAV alone as $72,000 (2).  Mass production would lower that further.

Scan Eagle and Catapult

The next question is how to deploy these small, cheap UAVs.  Ideally, they would be carried by a dedicated UAV “carrier”.  Such a ship would carry hundreds of UAVs and would accompany every surface group to provide the long range coverage that is needed to effectively utilize the long range anti-ship missiles that are [hopefully] coming to the fleet.  A UAV carrier would be a commercial cargo ship suitably modified to conduct high tempo UAV operations.  Basically, this just means a small “flight deck” (a row of catapults and some space to move the UAVs) to launch UAVs from and several recovery nets/hooks.  Nothing fancy.  We can build supertankers for $100M so this should be a $200M effort.

Alternatively, small UAVs can be operated by the dozens from any ship or land bases where geography permits.

The biggest challenge to this approach is reversing the military mindset of “bigger and more complex”.  Everything the military builds is bulked up with added gadgets rather than stripped down to minimum essentials.

Our current plan to use a few large and expensive aircraft, both manned and unmanned, for deep penetration surveillance and targeting is unworkable in high end combat.  We need a different approach and small, cheap UAVs offers a viable alternative.


(1)Wikipedia, retrieved 20-Dec-2016,

(2)Aviation Today website, 2004


  1. A bit of a strawman. I don't think anyone is planning to us e a P-8A for deep penetrating ISR.

    1. The P-8's mission is broad area maritime surveillance in concert with the Triton UAV. Unless the Navy is only planning to monitor friendly waters then I've got to assume that they'll be trying to look at enemy waters.

      Your comment is a bit of a strawman in that the P-8 was not the focus of the post and it doesn't matter what platform is used if it isn't survivable and cheap. The P-8 was a likely candidate as would be Reapers and helos and AWACS and F-35s and any other potential surveillance platform I can think of.

      Alternatively, if the Navy has no intention of using the P-8/Triton for deep penetration targeting then one has to ask how the Navy thinks they'll obtain targeting data for the LRASM or any other OTH weapon? Developing the LRASM without a viable targeting platform/process would be idiotic and akin to building a new ship class without any ammo for its main weapon (oh wait, we did that) or building a modular ship class with no operational modules (oh wait, we did that).

    2. The P-8A is intended for Theater and Strike Group ASW. And when it comes to looking for enemy submarines - the term 'friendly' and 'enemy' waters is a bit irrelevant.

      F-35 seems to be the best approach for ISR&T in a denied environment. Cheap UAVs like Scan Eagle simply don't have the range or speed.

    3. I sense a change in strategy and costs direction under the new trump administration

      I believe we're seeing him opt for cheaper more effective weapons that can actually kill. Wait until he starts making comments on the ship building programs. Here's already seized on the F-35 program. Expect changes

    4. "The P-8A is intended for Theater and Strike Group ASW."

      Only partly. From the website,

      "the P-8A is designed to be combat-capable, and to improve an operator’s ability to efficiently conduct anti-submarine warfare; anti-surface warfare; and intelligence, surveillance, and reconnaissance missions."

      Thus, ISR is one of its missions.

    5. In case we thought China was not thinking long range attacks at sensor aircraft and tankers they are developing a new missile with a250 Mike range apparently.

    6. I should have added that this is a sound strategy. They attack or EW or tankers assets making our deep strike aircraft toothless. How about that 3rd offset strategy

  2. What contributions might satellites make to narrow the search field and how might that impact the need for air based surveillance? How realistic is it that all of America's Satellite capability would be rendered inert in high end combat?

    1. Satellites will be destroyed to some degree, probably significantly. The remaining satellites will be tasked with much higher priority work (like monitoring enemy nuclear capabilities) than looking for targets like surface ships. Further, satellites cannot provide real time targeting data.

    2. Why can't modern satelites do real-time targeting? I know in the old days they had to drop film or had bandwidth limitations, but these days?

      Take the Gaofen-4. It has a 50m resolution and is parked in geosynch orbit over Singapore. It's probably better at tracking than ID'ing, so lower high-resolution sats will initially identify the targets. Once that's done, I don't see any reason other than prioritization why it couldn't track ships and provide firing solutions.

    3. You almost answered your own question. Satellites are still image based. They may be digital but it's still individual images that have to be analyzed. Computer image analysis can help but there's still a lot of interpretation involved. By the time that process is done and the data has been passed through various levels of command and out to a platform that can shoot, the target has moved.

      Plus, satellites are a precious commodity. A ship is a relatively low priority target.

      Satellites can't just take a picture of the entire ocean and instantly find a ship. Well, they can take a picture of the entire ocean but it won't have any useful resolution. If you zoom in enough to get useful resolution then lose coverage area. It's almost a no-win situation.

      Satellites are not the all-seeing eye that many people believe them to be. Consider the missing Malaysian airliner from a while ago. No satellite saw that.

    4. I agree that the initial target acquisition will not be in real time. Just suspect that once that's done, modern systems might be able to use that to develop a real-time track.

      As always, I'm thinking of what China might do to oppose a carrier crossing the Pacific. Hard to believe that they're not working on ways to target ASBMs by satellite.

    5. You understand that most satellites are in orbit, meaning that they move. They can't track an object even if one could figure out a way to get that data instantaneously to a shooting platform.

      Geosynch satellites are reserved for fixed location, high value observations.

      Do you think we'll allow Chinese satellites to target us if we think they have that capability? We've demonstrated anti-satellite capability.

    6. TomR, you might want to read up on some of the Cold War era carrier tracking exercises and incidents. There are some very enlightening accounts scattered about the Internet if you dig a bit. The short summary is that the Soviets were rarely able to track carrier groups. Even in our own exercises, carrier groups have proven very hard to find.

    7. Sorry to get back to you so late on this.

      I've read the classic "How to Hide a Task Force" and it's a great read. Are there others like it? The thing is, the PLAN's probably read it too. Complacency on adversary’s capabilities is dangerous. There have been a lot of advances since 1982.

      I think you might not be seeing this operationally. Take a look at the assessment portion at the end of the link I posted above. The Chinese naval surveillance satellites cover every spot in the Pacific about once every two hours.

      Let's say it takes them two days to do the image analysis. Thing is, they didn't throw away the photos from 46 hours ago, so once you have the two day old fix, you can quickly search a much smaller area of the 46-hour old ones. Repeat for 44, 42, etc. hours old, and you pretty quickly have a few hour old location.
      That's when the Gaofen-4 comes in. My understanding is that it can take 400km x 400km images, with 50m resolution (400m infrared), once per minute, anywhere in a 3500km radius circle roughly centered on Singapore. With 50m resolution, it might not be able to do initial identification, but it can certainly distinguish and track a carrier once it knows what it’s looking at.

      This gets to a picture where, weather cooperating, 3-4 days after the effort starts, the Chinese have an ongoing fix on the carrier at one minute intervals. That’s what I mean by track, obviously the satellites aren’t physically following the carrier.

      Can they relay that information to their shooters fast enough to be useful? Given how much they’ve invested in ASBMs, it would be foolish to think they aren’t practicing.

      This does get us to about the same conclusion: US anti-sat capability is vital. I suspect and hope that we’ve got more geosynch capability than we let on.

      Read into this what you will:
      “RPO allows for the space vehicle to maneuver near a resident space object of interest, enabling characterization for anomaly resolution and enhanced surveillance,…” Yeah, we’re just looking, don’t worry about us.

    8. This might be something I shouldn't speculate on, but I wonder if any of the money on the Ford went into temperature control for the flight deck.

      It'd be really nice to be able to maintain the deck at the same temperature as the surrounding ocean.

  3. Wouldn't relaying the data from these UAVs back to the "weapons" rely on the same networks you have emphasized will be degraded or completely ineffective in combat with a peer adversary.

    1. That's correct. The same weak link exists and is a problem that needs to be solved. It may be that the entire concept is invalid due to comm issues. On the other hand, the Navy believes that they can maintain comms in high end combat. For the purpose of this post, I've assumed that we can communicate to a sufficient extent. We need to begin exercising against the best ECM we can generate and find out what really works and what doesn't.

  4. Exactly. In a serious nation state war, the enemy will shoot down larger UAVs.

    This means that only small, expendable UAVs will be able to get in and their loss rate will be high.

    This is one of those "bad habits" and "bad assumptions" that one gets from spending too little time thinking about conventional war, but too much time facing how to fight Islamic Fundamentalists. The enemy gets a vote in strategy too - especially with a well equipped enemy with access to the latest in conventional arms.

  5. Make the UAV mission a part of the AFSB mission set. Each has a large area that can serve as a launch and recovery area and on at least one of the models, it would not interrupt the mission of LCAC lUnching or vehicle staging area.

    1. That might be a partial solution for the case of an amphibious landing, however, even then, the AFSB flight deck is supposed to be dedicated to supply movement, I would assume. Having to throttle back supply in order to operate UAVs does not make sense. Further, the assault may be nowhere near the OTH areas of interest which may be hundreds of miles away.

      Now, operating some UAVs for naval gunfire support might be worthwhile but even then I would suggest operating the UAVs from the firing platform(s) for better coordination.

      There's also the issue of mission bulge. One of the characteristics of an overworked force is the piling on of additional missions, far beyond what a given platform/crew can realistically handle. A crew can only train for, and be good at, a very few missions - probably, one is the limit.

      Consider the Burkes. They're tasked with AAW, BMD, VBSS, ASW, strike, naval gun support, helo ops, signals analysis, and the list goes on. The feedback from the active duty personnel is that the Burkes do not do ASW, at least not well, because they simply don't have the time to train for it.

      So, adding another mission to the AFSB is just a way of trying to avoid the fact that we need a dedicated UAV carrier!

      What do you think?

  6. Just a few thoughts . . .

    1) The $72,000 unit cost is from a 12 year-old article. Inflation alone would put the cost at $117K in today's dollar. Considering how the military works, the unit cost could easily be 2 to 3 times that number. But, I see your point on mass production.

    2) The Scan Eagle carries a limited payload, which according to Wiki is a "high resolution, day/night camera and thermal imagery." It cannot carry high-end radars found on larger UAVs. Though Wiki mentions the RAN tested Sentient Vision’s Kestrel Maritime ViDAR in 2015 giving it a "broad area maritime surveillance (BAMS) asset capable of covering up to 80 times more area in a single sortie than is possible with standard cameras."

    3) Scan Eagle cannot carry any weapons found on larger UAVs to strike targets of opportunity. Though Scan Eagle itself can be used as a weapon, but that would mean sacrificing a surveillance asset. Perhaps in this scenario, eavh UAV carries a nominal explosive charge to create more damage. With a number of these out there at any given time, you could also use swarm tactics to strike larger targets.

    4) Communication range is about 100 km. For this idea to work, you would have to find some way to increase communication range to 250+ km or so. Otherwise, your UAV is out ranged by many anti-ship missiles.

    5) I like the idea of a UAV carrier and it wouldn't have to be very big. Maybe the Independence-Class LCS is a starting point.

    6) You would also want these UAVs to be as hack proof as possible so they are not used against your own ships.

    1. Who mentioned weapons on a Scan Eagle???? The post was about sensing!

    2. The post was about using dozens, perhaps hundreds, of Scan Eagles as opposed to fewer larger UAVs costing $100+ million each. But, while you're out there sensing, what do you do if you come across a target that can't be serviced other means? For example, a mobile missile launcher or a radar truck. Or, a SCUD-like missile ready to be launched. In those cases, I think you would want the ability to hit those types of targets, either with a missile or the UAV itself.

  7. CNO; maybe this is too off topic... but could you (or anyone else) offer some suggested reading on what the Navies uses to target now?

    Some stuff seems straight forward. Aegis and the SM series of missiles seem (relatively) closely matched as Aegis is looking with air search radar, which is less constrained by horizon limits. So the SM's can go beyond Aegis range, but not drastically so; and the extra fuel might help to have around for alot of reasons.

    Harpoon/LRASM/TASM if they bring it back/NSM/Brahmos/etc...

    all have ranges (well) in excess of their launching platforms sensors. This makes some sense in air carried weapons (LRASM on the BONE, Harpoon on the F-18) but for others not so much.

    I.E. we have cruisers with Harpoons operating alone; or the entire idea of distributed lethality.... just doesn't work out. You have ships with BB ranged sensors and weapons with aircraft ranges. It seems a flaw in everyone's thinking. The Sovremenny class destroyers were long listed as 'carrier killers' with their Sunburn missiles, yet I don't see them being able to get within sensor range of a carrier. SSN's might, and that's a real pucker factor, but they are about it. I was told once that a big knock against the Pegasus class was that its radar mast was so low it couldn't easily target weapons for its Harpoons at any range.

    What am I missing? Are they just assuming that Ship X will be able to transmit coordinates to Ship Y while Ship X is under fire? Or that plane A can loiter and give coordinates to Ships B, C, and BONE D?

    Were the Russians planning on using SSN's to trail and give taret information to other platforms?

    Because if the targeting doesn't work it seems like we've spent alot of time and money on missiles we can't shoot. It would be better to have a short ranged (30NM) missile with a 2500lbs warhead and less fuel than a long ranged missile you might never be able to shoot. Or, better yet, go back to old fashioned guns with (much!) cheaper unguided shells and big magazines.

    Sorry if this is overly ignorant; Or just a summation of other posts, I just don't get it, and haven't been able to find anything on my own.

    1. Long range targeting has always been the challenge and it's no different for Russia or China or us.

      The Soviets "solved" the problem by using a combination of subs for general localization and lots of Tu-95 Bear recon aircraft. They planned to use them in large numbers and accept a significant degree of attrition. They deemed it worthwhile for the chance to find a US carrier group.

    2. Okay, so we've been building weapons with super long range, but no real great method of targeting them at long range other than having other units do the targeting. I just wonder at the efficiency of it. Might it not make more sense to not spend the money on backfires and Bears and just use existing RORSATS and buy more SSN's that can get closer.

      The Chinese method sounds a bit better. If the DF21 really costs a a couple million a pop, you can shoot *alot* at a general area before you equal the cost of a deployed CVBG.

    3. I'm not sure you fully realize just how big the ocean is and how small the field of view of a missile is.

    4. The Soviet RORSATS were never that effective at finding and tracking carrier groups. Add to that the fact that every major enemy now has anti-satellite weapons and satellites are just not going to be available nor effective - plus, they can't provide targeting quality data!

    5. "I'm not sure you fully realize just how big the ocean is and how small the field of view of a missile is."

      Very true.

  8. Walter, forgive me, but with the comments below aren't we risking the type of scope creep that has killed small weapons before?

    "2) The Scan Eagle carries a limited payload, which according to Wiki is a "high resolution, day/night camera and thermal imagery." It cannot carry high-end radars found on larger UAVs. Though Wiki mentions the RAN tested Sentient Vision’s Kestrel Maritime ViDAR in 2015 giving it a "broad area maritime surveillance (BAMS) asset capable of covering up to 80 times more area in a single sortie than is possible with standard cameras."

    3) Scan Eagle cannot carry any weapons found on larger UAVs to strike targets of opportunity. Though Scan Eagle itself can be used as a weapon, but that would mean sacrificing a surveillance asset. Perhaps in this scenario, eavh UAV carries a nominal explosive charge to create more damage. With a number of these out there at any given time, you could also use swarm tactics to strike larger targets.

    1. I agree with you on scope creep, but you're limiting yourself to Scan Eagle's small electro-optical payload. And, if other weapons aren't available, how do you service time sensitve targets or targets of opportunity?

    2. Walter, I've answered this in a comment reply above. Check it out.

      Also, I was NOT suggesting that Scan Eagle was the final, perfect product. I offered Scan Eagle as an example of a UAV that was in the ballpark of what is needed for this role. What specific sensor(s) we would place on such a UAV remain to be seen. I don't know all the sensors that are available. Besides, as I said in the post, the answer to the limited field of view is numbers - hence, small and cheap.

    3. If the answer is small and cheap UAVs, you're limited to the kind of payload they can carry. Which, for UAVs like Scan Eagle, is an E/O camera with thermal imaging capability. Which is great for adjusting artillery fire or spotting targets out in the open. What about finding more difficult targets, like the periscope of a submarine? Or, a camoflauged missile launcher?

      Then there is the communication issue, small UAVs usually have a limited communication range. If the intent is to fly deep into enemy airspace, they have to have some way to relay what they find back to base. And, a larger comm system will will cost you more in terms of power and weight, as well as driving cost up.

    4. "limited to the kind of payload they can carry."

      Are you grasping the basic concept here? There are many available payloads that will fit the size airframe we're talking about. I mentioned EO/IR, high resolution digital cameras, and mini-synthetic aperture radars. Fighter aircraft are carrying small IRST (infrared search and track) units that might be applicable.

      Did you understand the basic concept that we don't need an E-2 Hawkeye radar mounted on a UAV - we can make up for the individually small fields of view through numbers.

      I also don't think you're understanding the basic scenario. We're not talking about conducting surveillance at a microscopic level so that we can count the number of rifle rounds each soldier is carrying. We're talking about finding ships for the Navy's anti-ship missiles. We're also not talking about surveillance over land as a routine application. That's an Army issue. Of course, we may want to conduct some surveillance of larger land targets for Tomahawk targets so land surveillance certainly is possible.

      The main purpose is scouting for Navy purposes. If we want to expand the concept to general battlefield surveillance using the small, cheap UAV approach then we simply need to adjust the sensor package. No big deal. I get the feeling that you're trying to make this into something it isn't and not remembering the main point which is CHEAP so that we can afford large numbers for coverage and attrition.

    5. I understand the concept, but if your comm package has a range of 100 km, you're out ranged by most ant-ship missiles. Small and cheap is not neccessarily better if you can't find the enemy before they find you.

    6. You'll note that I stated in the post that the concept was dependent on developing a longer ranged comm package.

      Scan Eagle, as an example UAV, was developed specifically for short ranged applications. They had no need for, and made no attempt to find, longer ranged comm packages. There are many long range comm packages available. I assume it wouldn't be too difficult to find a suitable unit.

    7. The RN demonstrated in last years "Unmanned Warrior" Exercise a comms drone bouncing control for multiple drones from a shore based establishment out over the horizon.

      Relay Unmanned Aerial Vehicle

      ( I think they set some kind of world record with that particular stunt )

      Looks like the kinds of tech that could be fitted to a ScanEagle type platform.

      My understanding of LRASM is that the missile procecutes the target autonomously with just a approximate location and a target profile.

      The kind of ID needed is therefore not as specific as is traditional.

      I would also like to note that although not traditionally stealthy. Just due to size and build materials a ScanEagle has a very small radar cross section head on.

  9. Big, small they are all problematic and take away resources from real hardware for platforms and weapons.

    Case in point:

    Can anyone with a brain really think this payload restricted drone platform can do what an EP-3E with a "tube" filled with smart operators and the ability to carry about any sensor that's 1/2 way aerodynamic?

    Of course not but our leaders and acquisition "experts" and "geniuses" keep selling us stupidity..


    1. "Big, small they are all problematic and take away resources from real hardware for platforms and weapons."

      Okay, if you're not in favor of UAVs for long range targeting, what do you suggest?

    2. Depends what you mean by long range. You know what a single ships ESM and radar LOS is. You also know what a carrier air wings LOS is with aircraft like the E-2 and H-60. You also know shore based MPA aircraft, if they show up, can extend that quite a bit.
      Small drones are excellent for small boys but do nothing for carriers.
      Now it seems the MQ-25 is going to be the savior...Me thinks not however because everything we build seems to lack long range performance and persistence. Can't trust what we will come up with...
      We actually need an S-3 or a new type vehicle with 1200 nm out /in range ICW more P-8s. Anything less leaves us the same or worse. Tritons limited payload and shore basing while good for choke points is only a soda straw and vulnerable.
      Whenever I consider drones I think of them as fire and forget first day of the war devices, not robust 24/7 all weather reusable and defendable war platforms despite all the hype.

  10. Great post, and such an important problem.

    The bottleneck for the small UAVs seems to be their range/speed limits. e.g., the Scan Eagle has a decent range, but at 50kts, it takes 11 hours to get there, and tracking 30kt ships with a 50kt drone seems like trouble.

    What's the current technology about delivering drones from larger aircraft or missiles? Extreme example, a B-2 could probably carry 200 drones, if you can fold them to fit in an SDB case. There are probably situations where that would be more useful then a direct attack payload.

    Sub-launched drones? Would probably work, but it's always going to be a hard sell to replace other ordnance. Pre-positioned sea-floor drones? A bit science fiction, but looks good on paper.

    1. "decent range, but at 50kts, it takes 11 hours to get there"

      You're not thinking this through from an operational point of view. It only takes 11 hours at the very start of an operation. Thereafter, a new UAV relieves the old one whenever needed and arrives on station prior to the old one leaving. Thus, you ensure continuous surveillance and it's immaterial how long it takes to "get there".

      More to the point, think of these UAVs like the WWII carrier scout planes that would launch each morning. These UAVs will be sent out to see what's in reach of the group's weapons. How long it takes to reach the max range doesn't really matter. Whenever they find something, they'll report back and the group can attack. If we want to keep continuous relays of UAVs then we'll always have some at max range so travel time doesn't matter other than as it impacts time on station.

    2. "a B-2 could probably carry 200 drones,"

      Setting aside how many it could carry, what you're proposing is a one way, throwaway use since the UAVs could not be recovered. Even at $20K-$50K each, that's throwing away a lot of money! In your example, using the low cost figure of $20K, that's $4M thrown away with no guarantee that we'll find a target. We can't afford to do that too many times!

    3. Ah, I see what you're saying. In that case, take it one further: UAV-to-UAV in air refueling is not too far off. The sensor UAVs can stay on or near station indefinitely, only the tanker UAVs shuttle back and forth.

      $4 million--that's like what, five shots from the Zumwalt? (heh, that never gets old). But yes, it isn't something you'd like to do often, but I can think of times it'd be worth while, say as part of the SEAD on day 1 of a high intensity conflict or when you've got a carrier group on its way, and you'd really really like to know where the mobile ASBMs are before it gets in range.

      Curious, do you play Command: Modern Air / Naval Operations? I wish I had the time to game out some of these ideas.

    4. "The sensor UAVs can stay on or near station indefinitely"

      That's an incorrect belief that many people have. A UAV is just an airframe like any manned aircraft. It can't operate indefinitely. Stuff breaks. Fluids need to be changed. Stuff has to be lubricated. Thermal wear occurs. Etc. The practical time limit for a UAV seems to be around 24 hours before stuff starts failing.

    5. A better choice for a drone carrier than a B-2 would be a small blimp, which itself might be unmanned, and could also act as a comm hub.

      Randall Rapp

  11. Would be interesting to create, since I don't think I ever saw one in public domain, a graph of cost of UAV vs performance or sensor/targeting? There has to be a sweet spot where you could get a decent priced UAV for the right sensor/targeting price.

    Just throwing out price tags seems wrong, there has to be a better way to figure out what USN should be 2 cents.

  12. On the same thinking is there a write up of the principles behind the modelling models, algorithms, maths that are the basis of AoA, though do wonder if incorrect assumptions fed in GIGO.

    In a perfect world if such a respected model existed it could be used to kill off on conception programs like the F-35, Ford, LCS, Zumwalt etc. and so respected that proponents of such programs, Admirals and companies, would be made to look like halfwits.

    Until that utopia arrives have to rely on CNO :)


    1. You could not be more right when you mention GIGO. The main problem with military programs today is that no one is doing concepts of operation prior to initiating a program. Thus, we have no firm grasp of what the asset is supposed to accomplish and what characteristics it needs to accomplish its function. Lacking a firm grasp of what's needed, we just load up an asset with every capability we can think of in the hope that it will prove useful. Of course, we then moan when the cost skyrockets! To your point, the lack of CONOPS is the "garbage in" to our development and acquisition process that is producing the "garbage out" products that we're seeing.

      Great comment!

  13. The UAV system you cited has the performance characteristics for screening littoral spaces and supporting counter-piracy operations. I could see a ship similar to the one you described deploying UAV's to examine surface contacts in an area like the Persian Gulf. The UAV's would also possess small explosive charges capable of disabling swarm boats. This ship might also be able to provide reconnaissance for an amphibious assault group and could sacrifice its UAV's to destroy light ground targets.

    A different UAV is needed for operations to target surface ships at long range on the high seas. The new UAV would have to be more expensive to pay for the increased range and improved communications. It is unclear whether such a UAV would be sufficiently cheap to conduct a search in numbers.

  14. The RAN has actually been trying to develop or convert small UAVs into the long range sensor role filled normally by airframes like the P-8 or the Orion but for coastal waters. The idea is that by utilising smaller UAVs like the scaneagle to cover littoral waters, it frees up the longer range P-8s and Triton UAVs for blue water surveillance.
    There's some more info on the scan eagle experiment with the sentiant ViDAR here :

    The thing is that it's restricted to coastal waters due to the need to maintain comms with the UAV.

  15. I was thinking in the exploration / scouting efforts of the American and Japanese fleets at the beginning of the Midway Battle.

    This exploration / scouting had to be long ranged, timely, broad and reliable, i.e., accurate and survivable.

    And able to communicate back its findings.

    All this requirements are necessary today.

    Japanese used floatplanes launched by their cruisers, American used scout aircraft including dive bombers launched from the aircraft carriers.

    The modern scout substitute has to be fast, long ranged and able to carry a decent payload, capable of navigating by gps, ins or other. It has to be simple for getting an affordable platform and reusable for the same purpose.

    It's much better if this scout doesn't need to be based an aircraft carrier.

    I think the US Military has such a system for more than 40 years. It's name is BQM-34 Firebee.

    You don't have to change much on it for getting what we are looking for.

    Of course it is nothing transformational on an old aerial target, so it's getting no attention from nobody relevant.

    You could launch multiple Firebees from the deck of a converted container carrier. Many of them could even be launched simultaneously.

    After fulfilling their mission they could land with their parachutes close to the launching ship and be recovered by a couple of helicopters or work boats operating from the same ship.

    Under deck, the said converted container ship could have workshops for restoring the recovered Firebees to fly condition. There will be also lot of space for storing a maybe hundred of Firebees as spare parts, tools, etc.

  16. interesting, this one slipped me by.

    "US military tests swarm of mini-drones launched from jets"

    1. CNO was wondering what you thought of the above article.

      Its a USN project. Drones deployed from F18. Over 100 30cm drones.

      Obviously a maturing project. Its an interesting project goven budget restraints. Clearly for carrier deployment to a active ( or high risk ) enguagement zone ?

    2. CNO was wondering what you thought of the above article.

      Its a USN project. Drones deployed from F18. Over 100 30cm drones.

      Obviously a maturing project. Its an interesting project goven budget restraints. Clearly for carrier deployment to a active ( or high risk ) enguagement zone ?

    3. It's interesting, as are a thousand other technological marvels that have no relevance to military matters. My question after viewing a video on it was, what's the military application?

      Is it just software testing leading to control of very large UAVs or swarms of missiles? If so, that might be worthwhile. Honestly, I can't think of any other purpose.

      The drones are too small and too short ranged to carry any useful payload. I noted that a controller somewhere was providing aim points for the drones (did you see that in the video?). Given the small size of the drones, that would suggest the need for a controller in the very near vicinity of the drones. That means either inserting a clandestine controller behind enemy lines (that ain't going to happen) or the swarm can't be controlled at any great distance.

      What can these things do from a combat perspective? I can't think of anything useful.

      This is kind of like the announcement some while ago about launching a small UAV from a submarine. Under what circumstance is that useful? Can you imagine a sub sticking its radar/comms masts above water just to control a single, small UAV or receive telemetry? The benefit doesn't begin to justify the risk to the sub.

      Or using an F-35 to reroute a Standard SAM. When is that ever going to happen? There is just no reasonable scenario for that.

      Do you see a use for tiny drone swarms? If all 100 drones suicided into a ship they'd be lucky to scratch the paint (of course, they might sink a LCS!). Maybe I'm missing something.

    4. I feel the same way. I'm not directly seeing the application. Or the need to deploy from an F18. That was a surprise.

      I'm wondering if its a sensor net thing for SEAD\DEAD, simply to find a distributed air defence network some of which is inactive for prompt strike.

      Perhaps a hand-grenade sized charge is enough to disable complex electronic arrays in this way ? A little like those anti - material rifles ?

      Its slightly weird the way they have held this quiet for N months then done a world wide release of info with pictures. Stinks of a PR job doesn't it !?

    5. Really? The possibilities seem endless. Perhaps not as much in high end ship-to-ship combat, but...

      Close support, especially in urban areas? The 60 Minutes story says they can autonomously search, even claiming they can do facial recognition and identify individuals. Even if they just can recognize humans, that makes this the ultimate anti-personnel cluster bomb.

      SEAD? It wouldn't be hard to give a drone the same RCS as a stealth aircraft. Even if they're little more than smart, loitering chaff, they'd be useful.

      For actual attack, it'd be interesting to see just how precise they can be. From what I've seen, it's not out of the question that a drone this size could, for example, fly into a gun barrel.

      CNO, I'm surprised you don't see a use for the F-35/SM-6 combo..IMHO, that's the next step in the "shoot the archer" strategy. OTH fleet defense without the ships ever turning on their radar sounds like a non-brainer to me.

    6. "ultimate anti-personnel cluster bomb."

      Really?! You're going to turn loose a bunch of autonomous cluster bombs in an urban area packed with civilians?

      That aside, this is a navy blog so I'm focused on naval applications rather than ground warfare. There may be some ground warfare applications. I just haven't given it any thought. On the other hand, if you're willing to turn loose a bunch of autonomous killers, why not just use artillery and blanket the area? Cheaper, faster, simpler.

      Fly into a gun barrel? Well, I can't fault you for outside the box thinking!

      "F-35/SM-6 combo"

      If you have an F-35 in range of a target why not just shoot the target with the F-35's own weapon? Why make a complicated kill chain out of a simple one? I suspect you think the F-35 has some kind of near-AWACS sensing capability. It doesn't. It can search a relatively small chunk of sky. Also, I've not heard how many missiles it could control simultaneously but I suspect it's only one. Again, not an effective means of killing planes and missiles.

      "SEAD"???? In order for that to work, these small, short ranged, very slow drones would have to be released on top of the target. If you can get a launch aircraft on top of the SEAD target to release drones, you may as well release actual weapons and be done with it!

      These drones seem like the classic example of an answer in search of a question!

    7. Ok, so staying away from the fire support uses...

      "Well, I can't fault you for outside the box thinking!" Just extrapolating--think about how much things changed when ordnance accuracy went from kilometers to meters. What do we do differently when accuracy is measured in centimeters?

      You're right, I am thinking of an F-35 as sort of a mini-AWACS. But lots of reasons for the F-35 not to shoot its own weapons: it doesn't carry that many , it might not want to give away its position, it doesn't have to move to a firing position, the SM-6 is a superior missile.

      SEAD: The Perdix has endurance >20min, speed >40-60kt, so they can be dropped 20nm or so from the target. The drones can be released from just about over the horizon from the target, timed to arrive just as the main attack engages.

    8. For starters, where did you get those specs? I've looked and been unable to find any.

      Also, I think you hugely overestimate tiny drone performance!!! Accepting your specs, for the sake of discussion, remember that these are hand size units that must weigh next to nothing. Thus, that 40-60 kt speed is highly subject to wind. I have a personal quadcopter drone and in anything more than a gentle breeze it gets pushed significantly off the track that I want it to go. In a stiff breeze, I can direct it full throttle into the wind and it makes no headway! Consider the winds at altitudes above 100 ft. 40-100 kt headwinds are common. Passenger jets are affected by head/tail winds!

      Now consider the concept of adding a payload like an explosive. For anything more than a tiny firecracker, the added weight will cut the endurance, speed, and range down to nothing. Just guessing but I suspect that drone would be lucky to be able to lift a few hundred grams! That's not going to blow up enemy weapons.

    9. "might not want to give away its position"

      I swear, so many people think the F-35 is magic! How is the F-35 going to find targets and then guide missiles without using its own radar and comms? And if it does that, it gives away its position. C'mon, think this through!

      Also, you understand that if we launch SM-6 missiles we give away the position of the ship, right? Backtracking missile launches and/or detecting massive thermal blooms is not exactly hard. Is it better to let the F-35 shoot and then move or to give away the position of a multi-billion dollar ship?

    10. Specs are from here:

      They weigh 290g, so yes, they won't carry much. That's why we need out-of-the-box ideas if they're to to direct attacks. To get a feel for the precision drones are capable of, take a look at
      It's not crazy that these things will be able to attack particular components of a target.

      For the F-35, knowing where the other guy is without him knowing where you are is sort of the idea for stealth fighters, no? And since the whole idea is that the ship is 200+NM away, shouldn't the missile launch be over the horizion?

      I saw something claiming that the cooperative missile launch was against a mock re-entry vehicle target. If that's true, it might also be that you need launch at longer range just to make the intercept work.