What is the practical maximum size for USVs?

[B.Smitty]

Most current efforts focus on converting RHIBs and other small craft into USVs.

How well does USV technology scale? Can a practical, sustainable 100 ton USV be developed? 200 ton? 600 ton?

Would such a thing be useful enough to pursue? Does the technology exist to enable safe, unmanned navigation in congested waters? Are datalinks robust enough? Are modern ship systems reliable enough to make this concept feasible?

How much space/tonnage can be saved by removing the requirement to carry humans? Can seakeeping requirements be relaxed?

The Navy balked at Cebrowski's Streetfighters because they didn't want "disposable" ships, and small ships pose logistics challenges.

Unmanned ships would be far more disposable.

At first glance, developing a system to sustain such vessels at sea would appear to be a big risk.

Obviously USVs wouldn't be viable for certain missions. There are no boarding parties aboard, so they may be limited to sensor nodes in maritime security operations.

[dundonrl]

well.. in Pt Hueneme, the ex USS Foster was converted too a remote controlled target/self defense test ship.. she's about 8000 tons and 560 ft long..

[kato]

How well does USV technology scale? Can a practical, sustainable 100 ton USV be developed? 200 ton? 600 ton?

100-ton USVs have been in service for decades. The German Seehund class (and the prototype Seepferd for the MJ2000 MCM concept), and the Danish SF-100 class.
These are minesweeping drones usually controlled two to four at a time from a controller ship; Seepferd was supposed to be a full unmanned minehunter, project cancelled due to budget cuts.

[B.Smitty]

Interesting. How long could these vessels operate without crew? Hours? A few days? A week? Two? How often do they require maintenance?

Is a large, self-deployable, USV "Streetfighter" viable, given today's technology constraints?

[kato]

Interesting. How long could these vessels operate without crew? Hours? A few days? A week? Two? How often do they require maintenance?

Jane's gives a maximum range of 520 nm at 9 knots cruise speed, so at least 2-3 days i guess. More likely something like a week if they run slower, which would well meet the control ship's mission endurance before they would regularly hit their tender again anyway.
In theory they can operate unmanned the entire time, with up to four drones controlled from one ship (standard missions are with three drones). Afaik, the control team onboard that ship consists of two people per drone.

However, for simplicity (especially in harbours), the Seehund is usually brought into the operation area by a 3-man crew, which will then transfer to the control ship, and the Seehund operates unmanned controlled from there. For weather protection for these temporary crews, the USV has a small bridge/cabin with minimal control facilities.

The Seehund USVs are used to sweep minefields by "simulating" a larger ship through artificial signals (magnetic field, acoustics) that will trigger mines (they also carry some mechanical minesweeping gear as well). The USV itself is built to withstand nearby mine explosions.

Any maintenance is performed by the same "System Support Group" as for their control ship, onboard the squadron's tender. With the frigate-sized tender in the operation area, the squadron with 5 control ships and up to 18 drones can keep up active operations for 3 weeks, although it's rare that all control ships and drones would sortie for a mission.

Couple pics:

[kato]

Is a large, self-deployable, USV "Streetfighter" viable, given today's technology constraints?

Self-deployable... depends. Doubt it, if we're talking something LCS-sized and above. Something along the lines of a 500-ton vehicle would definitely be doable i'd guess.
Of course, once you bring in armament and such, the control section vastly increases in complexity.

Note that the above Seehund drones are based on prototypes originally built in the late 60s. They've been upgraded with pretty modern COTS C2 systems though (to replace the obsolete 70s equipment), around 2000/2001.

This new C2 system includes a couple autonomous routines without operator control even, in which for example the drones can be set to do a simple "follow the leader" - i.e. they assume a formation behind the control ship for automatic transits, and regulate that formation through collision avoidance routines. Within limits (preset sweeping plans through a minefields), even the actual minesweeping is done automatically nowadays, with automatic contingency planning for drones that need to break out of formation for some reason or another.
The drones have a number of automatic "emergency routines" regarding their reaction on losing the data connection to the control system - these drones will usually simply stop and try to keep position (circling, anchoring etc) until a connection is reestablished.
On "manual control", the operators onboard the control ship have the exact same controls as if they were on the bridge of the drone itself.

[B.Smitty]

Self-deployable... depends. Doubt it, if we're talking something LCS-sized and above. Something along the lines of a 500-ton vehicle would definitely be doable i'd guess.
Of course, once you bring in armament and such, the control section vastly increases in complexity.

I wouldn't expect LCS-sized. The point of Streetfighter was to present the opponent with numerous, low-value targets. LCS, to me, is not a low-value target. It's a 3000-tonne corvette/frigate with a real crew onboard.

Even a 500 ton or smaller craft could be self-deployable, if it had the range. Perhaps certain seakeeping requirements could be relaxed because there's no crew to worry about.

How about deploying and recovering towed arrays or variable depth sonars? Is that something that requires human interaction? I know there's a USV towed array in the works in the US, so i imagine it's be doable.

I wonder how much volume and tonnage of a ship of this sized is taken up by crew spaces?

How hard would it be to develop an autonomous refueling capability?

[FOG3]

Streetfighter strikes me as an obsolete concept.

A single Arleigh Burke-class DDG has 96 VLS cells each of which can, at least theoretically, handle two quad packs of Harpoons. That makes for the capability of having 768 Harpoons on tap, not including the dedicated mounts on just one ship.

Nobody ever could make that many expendable ships. A expendable low value target pretty much by definition couldn't mount advanced sensor or defense systems, so wouldn't be exactly swatting the Harpoons down. Nor could it possibly be cheaper then Harpoons, Mk 46, or Mk 48 torpedoes without being a glorified moving target.

By tonnage a Arleigh Burke would have been a Battleship-class vessel at the turn of the last century at around 9,200 tons and 500 feet long. The USS Dunderberg was only 5,090 tons and 376 feet long able to take on any battleship in 1864. The American Battleships of 1890 were 10,288 tons, and Dreadnought herself was 17,900 tons, which really got Battleship tonnage as we think of it going.

What can Streetfighter do better that improving detection, targeting, and fire control systems won't? Every smart torpedo and missile is already an expendable one way system that can be rapidly deployed. Not to mention stuff like Captor mines and the equipment used by Helicopters and aircraft like the P-3.

[B.Smitty]

Streetfighter strikes me as an obsolete concept.

A single Arleigh Burke-class DDG has 96 VLS cells each of which can, at least theoretically, handle two quad packs of Harpoons. That makes for the capability of having 768 Harpoons on tap, not including the dedicated mounts on just one ship.

Nobody ever could make that many expendable ships. A expendable low value target pretty much by definition couldn't mount advanced sensor or defense systems, so wouldn't be exactly swatting the Harpoons down. Nor could it possibly be cheaper then Harpoons, Mk 46, or Mk 48 torpedoes without being a glorified moving target.

By tonnage a Arleigh Burke would have been a Battleship-class vessel at the turn of the last century at around 9,200 tons and 500 feet long. The USS Dunderberg was only 5,090 tons and 376 feet long able to take on any battleship in 1864. The American Battleships of 1890 were 10,288 tons, and Dreadnought herself was 17,900 tons, which really got Battleship tonnage as we think of it going.

What can Streetfighter do better that improving detection, targeting, and fire control systems won't? Every smart torpedo and missile is already an expendable one way system that can be rapidly deployed. Not to mention stuff like Captor mines and the equipment used by Helicopters and aircraft like the P-3.

A single Burke is a single target. Get lucky with an AShM or torpedo and all that capability goes down the drain. Spread that out over many smaller ships and your capability can survive longer in the face of hits.

Plus, we only have a finite and relatively small number of Burkes. Will we risk them close to an enemy shorline protected by mines, AIP subs, small boat swarms, land-based missiles and airpower? I doubt it. Burkes will stand far off, limiting their advanced sensor suite to a purely defensive role.

A USV is a vessel you can afford to lose. They can mount advanced enough sensors and weapons such as small towed array/VDS system and AShMs to be valuable sensor and shooter nodes. They can screen larger combatants from small boat swarms, and if a few die in a minefield, oh well. We may actually want an AIP sub to take a shot at one. That confirms its presence and general location and makes it easier to kill.

If inexpensive enough, they can saturate an area with small sensors and munitions. Quantity does have a quality of its own.

They force an enemy commander to make a decision: take them out and risk asset exposure, or let them whittle down his defenses one sub or minefield or small boat swarm at a time.

Streetfighter, as is, is dead because the Navy doesn't want "expendable" manned platforms. They also don't want to support large numbers of small combatants.

But if you take away the crew, you take away a large support cost, and you take away the main argument against "expendability" - the potential for lost lives.

For an illustration of this, just look at UAVs. Losing a Predator today is a small, page 10 story in the paper. Losing an F-16 is front page news.

[FOG3]

The thing is Burkes aren't being purchased in the concept of battleships, there's over 50 of them already built.

Your key word:
Get Lucky

Of course, someone already got ridiculously lucky against the USS Cole of the Arleigh Burke-class and failed to do anything that wasn't repairable seems as how the USS Cole is presently still in active service.

Sensor and weapon drones are sensor and weapon drones. Something P-3s drop in large quantities. Designing an intentionally expendable vessel flies in the face of the established and proven N-square law. There's a reason the drive was towards Battleships and not simply swarms of frigates, before the Carriers came in to their own.

I'm not sure why you're focusing solely on USVs. UUVs would tend to be able to pick up on anything on or under the waves with less fuss, and would probably be a cheaper alternative to sonobouys in the long run due to ease of pulling them back in. Captor mines already give the ability to sow areas with intelligent munitions capable of taking out things that ride on or below the waves. VTUAVs (Helicopter drones) are in develop in addition to continued UAV development otherwise.

So what you're really saying is you're full behind the DD(X) and Virginia programs, because the Burke and Los Angeles are both maxed out already and thus not really able to take advantage of developing UUV tech.

[B.Smitty]

Of course, someone already got ridiculously lucky against the USS Cole of the Arleigh Burke-class and failed to do anything that wasn't repairable seems as how the USS Cole is presently still in active service.

It is still in active service but at the time it was a mission-kill (or close to one).

Sensor and weapon drones are sensor and weapon drones. Something P-3s drop in large quantities. Designing an intentionally expendable vessel flies in the face of the established and proven N-square law. There's a reason the drive was towards Battleships and not simply swarms of frigates, before the Carriers came in to their own.

P-3s drop immobile sonar buoys of limited capability.

Attrition happens. Better it happen to a cheap USV than a billion-dollar Burke.

I'm not sure why you're focusing solely on USVs. UUVs would tend to be able to pick up on anything on or under the waves with less fuss, and would probably be a cheaper alternative to sonobouys in the long run due to ease of pulling them back in. Captor mines already give the ability to sow areas with intelligent munitions capable of taking out things that ride on or below the waves. VTUAVs (Helicopter drones) are in develop in addition to continued UAV development otherwise.

I was only focusing on USVs for the purposes of this thread. I specifically wanted to know if there were any show-stoppers for developing larger USVs than what is presently under development.

The problem with non-self-deployable USV/UUVs is that they require a mothership to carry them too and from the theater. This means either you build a large ship to carry them, or accept limits on USV/UUV size (and thus capability).

[kato]

One show-stopper for a large USV might be simply bandwidth.

You'd conceivably need a certain amount of bandwidth for any weapon station onboard the USV (in particular for man-in-the-loop systems such as Phalanx - keyword "video feed"), plus some more for the pilot, some for general sensor data (battlefield awareness), and some for damage control. Not so much on the receiver side of the ship, but on the sender side.

This quickly goes beyond what's available with current datalink technology - and physics.

This is also an issue with UCAV development already btw (yes, i know, Carlo Kopp - he's right there though). The physics simply don't support satellite microwave links that can transfer a couple dozen megabit with the antenna technology available for such limited sizes. Hence why UCAV development strives towards more autonomous systems.

Now for a USV it's perfectly possible to make its software autonomous enough that it can e.g. patrol a certain stretch on its own. Collision avoidance routines and such are highly developed enough for that.
The problem is with actual combat though. You don't want a system that shoots on its own - always keep a man-in-the-loop. And that means that for every single weapon system, you'll need to transfer the necessary information for that man to identify the target.