Unmanned is the fad of the day and using drones for anti-submarine
warfare (ASW) is an idea that keeps cropping up. So many people seem enamored with the idea
and yet no one has examined the actual use.
Typically, proponents enthusiastically cite the usual drone
characteristics, such as extended endurance, without understanding what that
means … or doesn’t. Let’s take a closer
look and see if ASW drones are a good idea or not.
Let’s examine the characteristics, claimed and actual, of an
ASW drone.
Endurance.
Proponents claim that the greater endurance of drones will revolutionize
ASW, however, that ignores the reality that once any ASW aircraft, drones
included, have expended their weapons/sensors they become useless regardless of
their remaining endurance. A drone with
a month’s worth of flight time is finished as soon as its sonobuoys and/or
weapons are expended. It doesn’t matter
how much longer it can fly. ASW aircraft
are sensor/weapon limited, not endurance limited. Thus, drones offer no advantage and, in fact,
depending on the exact drone, would very likely have fewer sensors/weapons than
a manned aircraft and would, thus, have LESS
effective endurance than a manned aircraft!
Signal Processing.
Manned ASW aircraft such as the S-3 Viking, P-3/8, and helos carry
on-board computers and analysts to interpret the sensor signals. Drones have no on-board analysis capability
and must continuously communicate with the host ship. This is a continuous, broad band transmission
which is highly susceptible to detection and localization by the enemy. The host ship, in turn, has to broadcast
control signals to the drone. Elementary
analysis of the locations of drones performing ASW reveals the likely location
of the host ship to the enemy.
Size and Operation.
Proponents never quite specify the size of the drone they’re calling
for. There are only two possible drone
platforms, currently: Burkes and
amphibious ships.
The reality is that destroyer size ships (like a DDH) have
no large flight deck or recovery area and are limited to something in the Scan
Eagle size (5 ft long, 10 ft wing span, 30-40 lb empty weight, 11 lb payload)
or the somewhat larger helo-type UAV such as Fire Scout.
Helicopter carriers could operate larger drones but even
they have limits. For example, the Wasp
class LHD has a [roughly] 104 ft wide flight deck. A commonly cited drone is the MQ-9 Reaper and
I'm moderately sure that, in theory, a 1000 ft x 104 ft flight deck would allow
a Reaper to take off. The caveat is that the Reaper does not have an immensely
powerful engine so the acceleration might, actually, turn out to be
insufficient. That would have to be tested but, for the sake of further
discussion, let's assume it could take off.
More problematic is that the Wasp island extends close to
half way across the flight deck amidships. That reduces the usable flight deck
width to around 50 ft, at that point. The Reaper has a wingspan of 65 ft. which
puts the nose wheel at 33 ft from either wing tip. Allowing for, say, 10 ft of
wing tip clearance from the island (the Navy would probably insist on a greater
safety margin than that), that would put the nose wheel 43 ft away from the
island which would be within 10 ft or so of the deck edge. That, in turn, puts
the wing wheels within a few feet of the deck edge. The slightest deviation and
the aircraft is off the edge! In short, it would seem that the full length of
the deck cannot be safely used. That leaves only a few hundred feet forward of
the island for takeoffs. Now, I'm really not sure an unassisted takeoff is
possible! In fact, it seems unlikely.
So, a Reaper would be the maximum size drone that could
operate off a big deck amphibious ship with catapults and arresting gear and
it’s likely even that is too big.
Of course, we could purpose design a drone carrier that
could operate large UAVs but that would be decades down the road and, likely,
unaffordable if we continue buying $20B Fords.
Then, there's the issue of storing/hangaring large UAVs
(Reaper is 36 ft long x 65 ft wide, for example). It would need some serious
wing folding to get an acceptable spot factor so that we could operate more
than one UAV.
Carrier Adaptation.
Adaptations such as beefed up landing gear, arresting hooks, folding
wing mechanisms, etc. all add weight to the aircraft and negatively impact
already limited payload capacities as well as unaided takeoff and landing
distances.
Payload. To
give some frame of reference as we talk about drone payload capacities, here
are some relevant sensor/weapon weights:
Sonobuoy - 35-40 lbs, depending on specific type
Torpedo - The Mk54 lightweight torpedo weighs a little over 600 lbs.
What is a useful payload size and composition? A reasonable minimum would be something on
the order of 40 sonobuoys and 2 lightweight torpedoes. Thus, a payload of two torpedoes plus 40
sonobuoys = 2800 lbs without launchers, pylons, and associated equipment.
Drone payload capacities vary widely, depending on the size
of the aircraft.
Drone Types. With
the characteristics we just discussed in mind, let’s now review the basic drone
types and see how they mesh with the characteristics.
Small. Small UAVs, such as Scan Eagle or RQ-21
Blackjack size, can operate off destroyers.
The drawback is that they have a very small payload to the point of
being incapable of effective ASW work.
Scan Eagle, as an example, is very small and has a mere 11
lb payload capacity. A standard A-size
sonobuoy is around 5” diameter x 36” long and weighs 39 lb. A Scan Eagle size drone couldn’t even carry
one sonobuoy!
Medium. Intermediate size UAVs such as the vertical takeoff and landing MQ-8B Fire Scout has a theoretical maximum payload of around 500 lbs, however, the practical payload is around 100 lbs. Thus, it could not carry even a single torpedo and only a few sonobuoys. This is simply not an effective payload.
The larger MQ-8C Fire Scout has a maximum theoretical
payload capacity of around 700 lbs with a practical payload of around 300 lbs.
These drones are capable of operating off a destroyer but
cannot carry a combat-useful payload.
Torpedo - The Mk54 lightweight torpedo weighs a little over 600 lbs.
Medium. Intermediate size UAVs such as the vertical takeoff and landing MQ-8B Fire Scout has a theoretical maximum payload of around 500 lbs, however, the practical payload is around 100 lbs. Thus, it could not carry even a single torpedo and only a few sonobuoys. This is simply not an effective payload.
Large. Larger UAVs such as the MQ-9 Reaper,
Predator, Global Hawk, etc. have payload capacities that begin to be useful but
they require actual aircraft carriers with catapults and arresting gear to
operate from. The MQ-9 Reaper, for
example, has a theoretical maximum payload of 3800 lbs and a practical capacity
of around 1000 lbs on a wingspan of 65 ft and a length of 36 ft.
Does the flight of the GA-ASI Mojave off HMS Prince of Wales alter your analysis? https://www.ga-asi.com/ga-asi-demonstrates-short-takeoff-landing-of-uas-on-uk-carrier
ReplyDeleteNo, not in the slightest. The GA data brochure for the Mojave has a performance graph that shows that a max 2800 lb payload requires a 1000 foot takeoff roll and limits endurance to 3.5 hrs. That's not useful in our context.
DeleteWhat do you think about a small drone with MAD sensors? The size of magnetic anomaly detectors has gotten quite small, some are only a few kilograms. Obviously the aperture is smaller so there would be a need for hundreds of the drones for wide area coverage. MAD should be easier for software to interpret than sonar, especially if background signals are pre-mapped.
ReplyDeleteLink to one example:
Deletehttps://www.cae.com/media/documents/DM044_MAD-XR-_EN_Feb2020.pdf
"small drone with MAD sensors"
DeleteMAD is NOT an area search sensor and a small MAD is even less useful. MAD is useful for a final fix on an already detected target. You'll note that the Navy has pretty much abandoned MAD. The S-3 Viking had it and it was deleted.
My understanding was that the Navy ditched it because their sonar improved and the MAD requires flying close to the the surface, which they wanted to avoid.
DeleteThe math suggests a drone in the $100,000 type class (<200 km/h) and a 500 meter detection width can cover around 1000 sq km per day operating 12 hours. A P-8 cost worth of drones could cover a lot of ground.
"My understanding was that the Navy ditched it because their sonar improved"
DeleteI've never heard that. Do you have a source?
"500 meter detection width "
I've never heard that kind of detection area for a MAD. Do you have a source?
"You'll note that the Navy has pretty much abandoned MAD. The S-3 Viking had it and it was deleted."
DeleteI didn't know. Thanks for mentioning that.
Hunters and Killers (volume 2) by Norman Polmar and Edward Whitman says the P-8 flies at too high an altitude to make good use of MAD, though India's P-8s have MAD. The book says in 2015 the US Navy contracted with BAE to develop a drone with a MAD sensor to send information to the Poseidon. The book calls this "a very expensive after-thought."
The Wikipedia article about the Boeing P-8 Poseidon says, "Unlike the preceding P-3, the P-8 lacks a magnetic anomaly detector (MAD) due to its higher operational altitude;[41] its acoustic sensor system is reportedly more effective at acoustic tracking and thus lacking a MAD will not impede its detection capabilities;[43]"
I did a little searching and couldn't find anything newer than 2015 about the proposed MAD drone accompanying the Poseidon. Maybe the project was canceled.
Here is a paper about the detection distances:
Deletehttps://www.researchgate.net/publication/292239873_Detection_Range_of_Airborne_Magnetometers_in_Magnetic_Anomaly_Detection
500 meters is reasonable for a low altitude run. Which way the sub is facing relative to the detector matters along with altitude.
I agree with Phil on the P-8, they want to fly higher where the aircraft is more efficient and has wider range for their radar (if that is the most combat effective way to go, is another question). This source also mentions the air launched drone with the MAD, seems like a smart idea to cancel that.
http://www.b737.org.uk/mma.htm
So a theoretical drone that cost $100,000 and had decent software onboard to limit transmissions and could do repeat runs to better confirm targets before transmitting could be interesting. I am sympathetic to the counter argument that the Navy would end up with a $5 million drone and it wouldn't be useful at that price.
This confirms what has long been understood: MAD is NOT a search and detection sensor. It is a pinpoint 'fixing' sensor for final attack positioning. Whether a MAD is mounted on a manned aircraft or a drone makes no difference. It is simply not a viable search option.
DeleteI think at this point the Conops is to use the UAV to be the dogs and flush the subs out for the manned vehicle to make the kill.
ReplyDeleteWe should be specific about the what we are talking about. General Atomics has pitched a concept for a STOL, folding wing version or the MQ-9B. It is not a detailed proposal. We know MQ-9B has a 4750 lb payload capacity and that they typically show the ASW loadout as 2 x 10 Sonobuouy dispensers, but advertise a 4 x 10 capacity, along with srface search radar. Some recently have showed the 2 Sonobuoy launchers with 2 Sidewinders. If that were true those sidewinders could be CVLWTs instead. Their proposal also shows the aircraft taking off and landing using the Helo take off line further to port placing it close to the deck edge. Seems like a bad idea but with JPALS we seem to have automated carrier landings licked. Good article on War on the Rocks today about how to improve the pilot training pipeline by embracing this reality.
Processing - GA indicates the processing can happen on board. It can provide the details needed in transmission, not pump all th raw data.
Wingspans. Mohave with STOVL wing has a 52 foot wingspan. Its smaller than a reaper larger than a Gray Eagle. One wold assume the MQ-9B with similar wing might have fairly significant changes in endurance and payload and also still have a wider wingspan than Mojave's 52 feet. I agree wingspan is a problem. LHA/LHD have only ever opertionalized AV-8, F-35 and OV-10. OV-10 had the biggest wingspan at only 40 feet.
Rolling length. Mojave needs 1000 feet on land. Headed into the wind with 20-25 knots ship speed should keep it in the 800 feet of an LHA/LHD. Obviously a ship with a ramp is a sure thing, they just demo'd it.
Agreed, MQ-8B/C can't really get it done. It seems the Navy knows this at this point. They probably need an unmanned version of the future scout helo.
Things we here and then see and hear nothing more. They were working on a Coyote drone variant with MAD sensor to drop from P-8 Sonobuoy dispenser. Anyone know if it really happened?
" use the UAV to be the dogs"
DeleteHow would they do this?
"We know MQ-9B has a 4750 lb payload"
No. We know the manufacturer claims a maximum theoretical capacity of that when the aircraft has unlimited runway and bare minimum fuel load with no concern for range, speed, or endurance. In other words, the claimed payload is what the aircraft can theoretically get off the ground with without any concern for negative impacts on performance. On any realistic mission with a normal fuel load and shorter runway the payload will be a quarter of max.
" pitched a concept"
As the saying goes, that and a quarter will get you a cup of coffee.
"with JPALS we seem to have automated carrier landings licked"
Until battle damage or normal electrical, software, or mechanical failures cause it to be unavailable.
"GA indicates the processing can happen on board"
Pure fantasy. If that were true, our ships wouldn't need entire compartments filled with computers and analyst/operators ... and yet they do. This is just like DARPA's claim that Sea Hunter can track a sub. If a small, low power sonar with no large amount of computer data analysis and no human interpretation could unfailingly track a sub then we would be engaged in a crash program to remove the Burke's large sonars, compartments, computers, and operators with a single, small, low power sonar ... but we aren't.
23 years ago my company needed a whole NOC to control our network. Not today. Things change. The thing that is slowest to change would appear often to be the USN.
DeleteThe thing that I've always disliked about ANY drone is the reliance on transmissions.
ReplyDeleteIn addition to giving away locations, control can be reduced/removed through destruction of satellites, jamming or hacking. I wouldn't fancy relying on total automation, so a remote pilot needs to stay in control of the drone and that makes them vulnerable.
In related news, there are proposals to fit the RN QE class with cats and traps (finally!), in order to deploy drones. Hopefully though, the equipment will be suitable for launching proper aircraft too.
https://ukdefencejournal.org.uk/project-ark-royal-plans-for-angled-decks-and-drones/
Lofty.
That article was all 'might be', 'someday', and 'proposed'.
DeleteIt also illustrates one of the biggest failings of Western militaries and that is the tendency to leap into new technologies and programs without bothering to analyze it to see if it would actually be useful. Just because we CAN do something doesn't mean we SHOULD. As I've repeatedly demonstrated, the case for unmanned in combat is very limited and yet we're basing our entire future warfighting on it. No analysis. No realistic wargames. No foundation. Just do it.
CONOPS, CONOPS, CONOPS then WARGAME, WARGAME, WARGAME. Then see how these new fads can be used most effectively along with how they can be neutralized.
ReplyDelete"along with how they can be neutralized."
DeleteExcellent. We utterly ignore the enemy's likely responses and countermeasures while assuming that anything we do will work flawlessly. On the rare occasion that we allow for an enemy response - such as Millenium Challenge 2002 - we get spanked and then ignore the lessons.
Even if ASW drones do work, can we afford them? How many can we afford to lose in a war? How fast can we build and replace them? Because we will lose some if not most of them pretty fast....then what happens?!?
DeleteI'm not offering any solutions to the signal processing or endurance issues, but the old DASH drones in the 1960's were able to carry 1 or 2 torpedoes, so it doesn't seem impossible that a little drone helicopter could carry a useful payload. To be fair, they were gloried torpedo delivery systems, so they had very short endurance. Apparently instead of taking this idea and making it better, we have more expensive, less capable drones today? I'm just saying that if I was on a FRAM destroyer, it wouldn't seem crazy to me that in 2030 there would be improved versions of the DASH doing useful work cheaper and safer than normal helicopters.
ReplyDeleteBut the DASH added a capability that wasn't otherwise available, and did it at a reasonable price. For a modern ASW drone to make any sense, it would need to do something a helicopter can't (operate off smaller ships without existing hangers or replace current helicopters 2 to 1) or be way cheaper and deliver similar capability. I don't see that happening.
Interestingly, the DASH carried the Mk46 torpedo which was a hundred pounds lighter than the current torpedo. Note that it could only carry one torpedo. It is highly debatable how useful a single torpedo would be in an engagement. I would expect that several torpedoes would be needed to secure a single kill of a sub.
Delete"For a modern ASW drone to make any sense, it would need to do something a helicopter can't"
Bingo! We keep wanting to invent new systems that duplicate existing capabilities rather than bring something new to the table.
How about one (two) of those new very light weight torpedoes - looks like about 220lb each? Could work with an improved MQ-8B, maybe with a 5-bladed rotor?
DeleteIf I needed to come up with a way to use drones in ASW I would probably use them in the following way:
DeleteUse a quad-copter to carry a dipping passive sonar and use it to work a pre-programmed search pattern.
After every dip the drone would micro-burst back the results of the passive search (possibly done at multiple depths at each stop to get above and below thermocline layers).
Any possible contacts could be prosecuted with manned assets.
The value of this would be that you don't need to burn up pilots' flight time, and the maintenance on a quad-copter should be less than that of a helicopter.
You could have multiple drones working the perimeter of a battle group or convoy without needing to expend a lot of resources.
Lutefisk
"How about one (two) of those new very light weight torpedoes - looks like about 220lb each? "
DeleteI'm not familiar with that but what could a 220 lb torpedo with [I'm guessing] a 20 lb warhead accomplish? It won't sink a sub or even damage it much. The Navy was already convinced that the current lightweight (600 lb) torpedoes were very questionable as far as significantly damaging Russian subs. A torpedo one third that size isn't going to do anything.
"Use a quad-copter to carry a dipping passive sonar "
DeleteHave you worked out the weight and strain (which requires vertical power to overcome) on a quadcopter in order to operate a dipping sonar with a several hundred meter cable and sensor????! No quadcopter I've ever heard of could handle that.
I like that we're considering out of the box options but we can't ignore engineering and physics!
As far as unmanned/drone ASW work, i think unless theres some huge changes in concepts, ASROC-style is as much as we need. Most of this unmanned stuff is doing it for the sake of doing it, and usually with less quality than the manned version...
ReplyDeleteConsider, however, that the range of VL-ASROC is only 12 nm. If a sub is within 12 nm of you, you're probably already dead! Now, if you're calling for a significantly increased ASROC range, I'm all for that. Whether or not it can be achieved is, of course, unknown.
DeleteWhat's our intent here? Bending current platforms to current needs? Can't do it. Or contemplating better ways of doing things? Which implies different doctrine.
ReplyDeleteWhat do we want a drone to do? Detect. Communicate. Kill chain finishes by VLS ASROC. Drones don't need to carry torpedoes. They DO need to carry sonobouys. Weight rules.
What's the limiting factor on DETECTION? Weight of sonobuoy because numbers over area divided by re-loadable drone capacity . How many Sonos do you need to lay a pattern and how much area does that pattern cover? Any TACCOs out there? Yeah it depends; temp/salinity/convergence zones and all that.
Can we get under 40lbs per sono? I think so. Refer to my comments at
https://navy-matters.blogspot.com/2023/11/ddh-hayler.html
How much range do we need to cover? MK-48, per publicly published stats effective range is 27nm at 40kn which translates to
approx. 18 minutes detection from firing at max range. Did I get the math right?
Let's double detection range to 54 nm. (sono deployment) which gives me 18 additional minutes to prosecute the target with ASROC before MK-48 launch (MK-48 is my example; can't be bothered looking up SOV torps right now). That's 36 minutes to react after launch- That enough range post-detection to react/prosecute?
So I need, off the top of my head, sono-reloadable drones with 300nm range (108nm there and back plus loiter for comms relay and whatever). Plus much much lighter sonos. Which don't exist.
But. What are we discussing here? The war that we know or the war that is coming? Which, regrettably, is 3-7 years out. Not enough time to implement what I've proposed. But-
Want another alternative? Cargo ships loaded with VLS ASROCs. Recon by fire; periodically launch torpedoes set in active homing mode in all directions ahead of (and all around) the battle group. Yeah, yeah Blue Force deconfliction.
Nope. I haven't done the math on this. I could. But I'd rather direct NAVSEA or COMNAVPAC to run the numbers. Give me 3 LTs over a weekend.
$1MM per torp times X torpedoes per hour X 24 hours per day? Maybe not. Versus a multi-billion dollar battle group? Maybe. Unit torp costs declining on volume (1000 per cargo ship? as a starter). I need cheaper torpedoes. You want to win? Or not?
Oh no! Cargo ships can't keep up with 30kn carriers! Don't have to. Think about what's coming in the South China Sea. Walk up and sanitize the area by launching active homing torpedoes ahead. Let the stupidly-robot active torpedoes snake out and kill anything they see in front of them.
Bottom line on any discussion regarding drones is CHEAPER AND MORE vs. FEWER AND EXPENSIVE.
It's the future.
USA has a history of trading treasure to save (our) blood. I'm all in favor of that. We need to rethink surface ASW, IMHO. The key here is lighter sonos deployed by drones combined with cheaper surface-launched torpedoes.
"Kill chain finishes by VLS ASROC"
DeleteVL-ASROC has a limited range of 12 nm. If a sub is that close, or closer, you're probably already dead. If you're suggesting a much longer range ASROC, I'm all for that but it would have to be something on the order of 20-30 miles, minimum.
" Recon by fire"
DeleteHave you done the calculation on the area/volume of interest around a naval group? It's immense! And that's an understatement! You'd need thousands per day!
Oh. And another thought.
ReplyDeleteCluster sonos deployed by VLS like a cluster-munition round like DPICMS (it's an Army artillery thing).
Because of large form factor, even with my proposed mini-sonos, I don't think munition footprint works on Burkes et al. But on a container ship...
Volume. We should be looking at multiplying the area that any single helo can cover by using different sensor deployment methods.
You really need to switch to decaf.
DeleteLutefisk
Could, I suppose-
DeleteBut then you'd lose all the (my-) math.
I was planning to look into this, as you suggested, and actually did do some investigation, but hadn't completed it because of computer problems this week. Sorry about that. You've done a better job anyway.
ReplyDeleteI would like to add one point. There actually is an existence proof of an aircraft in the size and weight range of the Reaper that can operate off a flat deck Amphib. The Vietnam era OV-10 Bronco, is similar in size and weight to the Reaper, and DID successfully operate off Amphibs. In fact, there are photos in the Wikipedia article of Broncos operating off USS Nassau, which was a Tarawa class Amphib.
The Bronco had a wingspan of 40 feet, not 66 feet, so the deck width wasn't a problem. This suggests you don't actually need that huge wingspan to get STOL capability on an amphib. My understanding (very limited) of the aerodynamics is that the long skinny wings give better aerodynamic efficiency which provides greater range. But if the application could trade off some range, shorter wings might be a possibility.
General Atomics has announced a "STOL kit" for the MQ-9B which they claim will be able to operate off carriers and amphibs. As you properly noted, right now it's vaporware, and very little information is available about it. I haven't found anything specific about the planned wingspan, for example. Apparently the plan is to have a kit which would replace the wings and tail surfaces of an otherwise standard Reaper B. The wings would supposedly also fold. One article suggested that the STOL wings might be somewhat shorter and fatter (larger "chord") than the standard wings. But no specifics were given. So I guess we'll see in a few years what they come up with.
This doesn't prove, of course that a drone version of the Bronco is the answer. For example, the Bronco had less range than the Reaper. It just shows that something is possible in that weight range.
Of course, your concern about the lack of local analysis capability and the concomitant need for massive communication with the ship remains a problem which, I agree, is very dangerous. I suppose that the magic of "AI" may help to address this at some point in the future, but that time is not now.
"you don't actually need that huge wingspan to get STOL capability on an amphib"
DeleteBe careful, here. It's one thing to take off, it's another to take off with a combat-useful load of fuel and sensors/weapons. Could an OV-10 take off from an amphib with a useful load? No idea. Do you have any information on that or information on weight versus take off distance?