Pages

Friday, June 14, 2024

MQ-25 Control

The Navy’s new, not yet active, MQ-25 unmanned tanker is a fascinating control scheme case study.  You may recall that it began life as a combat UCAV concept which then morphed into a combined strike/ISR, then a pure ISR, then a combined ISR/tanker, and, ultimately, into a pure tanker … with occasional rumors of ISR or strike capabilities still being possible with minor modifications.  That convoluted development path alone makes for a fascinating story but there’s another aspect of the MQ-25 that is equally fascinating and, as best I can tell, completely ignored and that is the control scheme required to operate the tanker and how that control scheme impacts the concept of operations (CONOPS).
 
The closest I’ve seen to a CONOPS is the vague, general requirement that the tanker should be capable of delivering 14,000 lbs of fuel at a distance of 500 miles (variously reported as 15,000 lbs at 500 nm, depending on the source).  Of course, that’s not even remotely a CONOPS;  it’s a capability and an ill-defined one at that.
 
Moving on …
 
The MQ-25 consists of two main components: the MQ-25 air vehicle and the MD-5 Ground Control Station (GCS).
 
In a bit of a first for a major program, the government is acting as the lead integrator.  I applaud that, however, there won’t be any manufacturer to blame if it does not go well!
 
The Navy’s Unmanned Carrier Aviation program office (PMA-268) is moving forward with integrating its two key elements—the MQ-25 air vehicle and the MD-5 Ground Control Station (GCS) at the program’s System Test and Integration Lab (STIL) at Patuxent River.[1]
 
PMA-268 is the lead systems integrator, working closely with its two prime industry partners, Boeing  and Lockheed Martin Skunk Works … [1]
 
“This will be the first time we are integrating an air vehicle and GCS from two different prime contractors,” said T.J. Maday, MQ-25 labs and integration manager.[1]

 
Airframe development aside, the challenge is to integrate the aircraft and the GCS with the various control ship’s sensors and software.  Many levels of integration are required – no easy task.
 
 
Control Scheme
 
There is no direct ‘pilot’ control of the MQ-25.  A ground ‘pilot’ does not fly the aircraft as is done with other UAVs.  Instead, the MQ-25 will be controlled via general commands which the aircraft’s software will attempt to implement … eventually … as the immediate situation allows.  The analogy would be someone telling you to buy milk from the grocery store but they don’t give you exact, second by second instructions.  You’re given a general command and left to figure out the details and exact timing of how to go about it yourself.
 
… the AVO [air vehicle operator] is never intended to directly input singular controls to the AV, combined with the expected signal delay, this is omitted … [2]
 
AVOs will input large scale commands such as a flight path or holding pattern, an altitude or direction change while running concurrent systems like the Stingray’s fuel pod or landing gear. “The logic within the aircraft will resolve [these] requests as compatible with its current phase of flight … [2]
 
This kind of ‘execute when you can’ control is fascinating.  Consider the simple example of a command to turn to a new heading, say, 90 degrees off.  Seems simple enough, right?  But, what if an aircraft is being currently refueled?  The UAV might be wise to delay execution of the heading change until after the current aircraft finishes tanking.  On the other hand, what if the turn command is the result of an enemy threat dead ahead?  Maybe the UAV should turn very soon and very sharply?  In fact, maybe it should terminate the refueling?  Maybe there’s another friendly aircraft that needs refueling on a fairly high priority but not an emergency?  Should the UAV continue tanking or break off and disrupt the current aircraft’s plan and timing?  What’s the current aircraft going to do with the fuel it receives?  What’s the priority?  And so on … 
 
As you see, the variations to even this ‘simple’ command are infinite.  Can we write software that can correctly assess and evaluate all the possibilities?  That strikes me as no easy task considering that, for example, we’ve been working on the ‘simple’ F-35 logistics software (ALIS) for decades and have failed miserably and the F-35 Block 4 software has been largely abandoned due to failure to complete it.
 
Alternatively, what if the UAV receives no command but there is a threat dead ahead?  Does the UAV have the sensors and software to detect and interpret a threat on its own and then make an intelligent response?  While we’d like to believe that the person controlling the tanker will be omniscient and aware of all threats and command the UAV accordingly, that’s pure fantasy in actual combat.  Some threats will be detected but others will be missed or detected too late.  What if an aircraft in distress needs fuel but can’t contact whoever the UAV controller is?  With a manned tanker they might be able to contact the tanker pilot directly and request help but you can’t talk to a UAV.
 
 
Signal Delay
 
Did you note the reference in the quote to ‘expected signal delay’?  This, too, is intriguing.  We’ve come to believe that any remote, unmanned control is instantaneous and this would appear not to be the case, at least not for the MQ-25.  I don’t know what particular component of the control scheme introduces delay or what the length of the delay is. 
 
This signal delay is similar to the widespread and misguided belief that satellites provide instantaneous detection and weapons launch control against ships.
 
Consider the example of a late detected threat described above.  The pilot of a manned aircraft can react instantly when the undetected threat eventually materializes.  A UAV, especially one with a signal delay built in, cannot react instantly.  We may lose tankers while the UAV flies blithely on, uncomprehending and uncommanded.
 
It is also unclear to me whether the expected signal delay is an inherent, unavoidable characteristic of the system components or whether it’s a conscious decision that real time control is not needed.  Fascinating, either way!
 
Regardless, the approach is a wise one, in the sense that trying to control a UAV in real time in combat is not consistently possible and it is probably counter-productive to even try.  Of course, this only works if the software can be made smart enough to resolve and manage the potentially conflicting or contradicting commands the UAV will receive.
 
It is significant that there is no official mention of MQ-25 control by other airborne assets although the Navy has expressed interest in such alternate control.  At the moment, the only control is via the GCS stations which will reside on the carriers.  However, consistent with its obsession with the ‘any platform/sensor/weapon can network with any other platform/sensor/weapon’ philosophy, it now appears that the Navy is trying to make the MQ-25 controllable by other aircraft.
 
Boeing is also working with the Navy so that an airborne platform can control the MQ-25 Stingray and not just from its aircraft carrier home. When speaking about this, Rear Admiral Telford said:  “MQ-25 needs to have the ability to talk and be managed by any airborne platform, including those of our allies and partners.”[3]

 
Communications Security
 
We noted in a previous post that the desire to control the MQ-25 from other airborne assets was rooted in a fundamentally illogical assumption about communications security (see, "MQ-25 Control Concept"). 
The value in pilots being able to task MQ-25s mid-flight lies within a core assumption the Navy — and more broadly the Pentagon — has about the future battlefield: all communications will be subject to attack. The shipboard controllers may not always have contact or permission to communicate with the MQ-25 depending on the situation. If that’s the case, then a pilot of a nearby manned aircraft may need to redirect the unmanned tanker without assistance from the ship.

Of course, this raised the question, if the shipboard controller can’t communicate with the unmanned tanker due to enemy disruption of communications, why would we think that we’ll be able to communicate with the manned aircraft to tell the pilot to redirect the unmanned one?  That’s a logical inconsistency.  Military thinking just teems with this kind of logical inconsistency.
 
 
Issues
 
Communications – Regardless of the degree of communications with the MQ-25, how secure are the communication links?  Will the regular, if not constant, communications, back and forth, betray the UAV, carrier/control asset, or both locations?  Every person I’ve talked to who knows anything about signals intercept states unequivocally that our comms are nowhere near as secure as we like to believe.
 
For that matter, what type of communication signal will the MQ-25 use?  Satellite relay?  LOS?  Omni-directional?  Multiple modes?
 
Cyber Security – Anything that can receive a signal can be cyber attacked and the MQ-25 certainly qualifies.  At a minimum, the aircraft will have sensors taking in external signals and dedicated communication and data link receivers.  We don’t want a Battlestar Galactica scenario but, as we’ve seen repeatedly, even the best protected network or computer can be hacked and on a fairly regular basis.  Hardly a month goes by that I don’t receive a letter from some company saying that my customer data has been compromised and that’s from major corporations who claim to have secure networks!  China is working every day to find and develop cyber vulnerabilities in our assets.  Can an unmanned platform function reliably in the face of cyber threats? 
 
Ground Control vs. Aircraft Control – Both approaches have pros and cons, as we’ve discussed.  One further aspect of the discussion is that if you need an aircraft to control the tanker, you’ve essentially turned the ‘one-man’ tanker operation into a multi-aircraft procedure.  Requiring two aircraft to enable one to be a tanker is horribly inefficient and, essentially, doubles the cost while requiring twice the resources.
 
CONOPS – I desperately hope the Navy has thoroughly worked through the CONOPS under realistic conditions before concluding that the MQ-25 was the best solution.  My fear (near certainty) is that they hopped on board the unmanned tanker in a technology-for-the-sake-of-technology move and that an unmanned tanker is not the best solution.
 
 
 
 
 _____________________________ 
 
Note:  Fleet service timeline has been pushed back to 2026 or later.
 
 
______________________________
 
[1]Navair website, “MQ-25 team preps for first air vehicle, control station integration test event”, 18-May-2022,
https://www.navair.navy.mil/news/MQ-25-team-preps-first-air-vehicle-control-station-integration-test-event/Wed-05182022-0715
 
[2]Forbes, “Developing The MQ-25’s Ground Control Station Means Thinking Like A Mission Commander - Not A Pilot”, Eric Tegler, 12-Jan-2021,
https://www.forbes.com/sites/erictegler/2021/01/12/developing-the-mq-25s-ground-control-station-means-thinking-like-a-mission-commandernot-a-pilot/?sh=643ac901557a
 
[3]Simple Flying website, “Pushing Boundaries: What Is The Boeing MQ-25 Stingray?”, Mark Finlay, 13-Feb-2024,
https://simpleflying.com/boeing-mq-25-stingray-guide/

10 comments:

  1. Seems like a large waste of time and money while we continue burning up Hornet airframe hours as tankers. How many A-6s could've been refurbished and put back in service as tankers for whats been spent on the new gadget??

    ReplyDelete
  2. I wonder if this MQ25 has been tested in a flight with windy conditions that an aircraft can tolerate having good pilot. Also I believe that the S3 Viking had a tanker variant and some could be made airworthy ?

    ReplyDelete
    Replies
    1. It's always windy at altitude! An automated flight control system can probably do a better job of remaining steady than a pilot can.

      S-3 Vikings have been put in storage. Whether they can be reactivated is a matter of opinion. Whether the cost to do so would be worth it is, again, a matter of opinion.

      Delete
    2. https://theaviationist.com/2024/05/03/new-sotware-f-18-pilot-commands-mq25/
      Boeing is testing software to allow the Super Hornet to control the MQ25 ( No mention of this software use with the F35 C )

      Delete
    3. "allow the Super Hornet to control the MQ25"

      That largely defeats the purpose of the MQ-25. The intent was that the MQ-25 would operate autonomously. If every Hornet has to have additional comm gear, secure comms, and software installed in order to refuel, that somewhat negates the MQ-25's supposed benefits. Broadcasting also allows the possibility of signals detection by the enemy in what should be a silent operation.

      Delete
    4. My mistake earlier-said A-6, but S-3 was what I was thinking of. Have to agree that this project does feel like a tech-for-techs-sake scenario. If its waffling between autonomous and being controlled than its clear its progressing without solid CONOP, because the two are quite different. Moreso considering that at least initially ISR and or strike missions were in there too. So will there be a dedicated control aircraft or will that be a widespread capability? And is there any info on how many are envisioned to be deployed per carrier or air wing?

      Delete
    5. "waffling between autonomous and being controlled than its clear its progressing without solid CONOP"

      One thing we need to keep in mind is that the military's default, one-size-fits-all CONOPS is ... do everything! Their answer to any CONOPS related question is, YES!, do it, do everything! Recognizing that, we see that adding the ability to control the UAV from a Hornet makes perfect sense .. to them.

      For the Navy/military, CONOPS = EVERYTHING and every weapon, ship, plane, or piece of equipment has the same CONOPS: do everything. Saves time; you only have to write one CONOPS because they're all the same! Logical and efficient! Saves time so you can focus on diversity, climate, gender, and woke issues which is the real purpose of today's military!

      With that in mind, we can understand the military's decisions. They're still stupid but we can understand the stupidity.

      Why do they have a default 'yes' position? Aside from the innate desire to always say yes because that's how you get promoted, saying no involves taking a risk and making a stand by saying something isn't needed. Today's promotion boards do not reward risk-takers or stand-makers. Hence, the default answer is always yes.

      Delete
    6. Also a strike package could have a multitude of aircraft lining up to be refueled. The MQ25 tanker will service one aircraft at a time. So the buddy system, using the super hornet for tanking. will still be an option. ( but perhaps needing the F18 super hornet tanker less.) But communication with the MQ 25 could potentially be a problem.

      Delete
  3. few points

    - as a practical matter, with only 14,000 pounds of give the MQ25 is going to spend its life in the “immediate” vicinity of Mom and top of aircraft as soon as they get to altitude. This should alleviate concerns about unexpected threats materializing right in front of it.. And if one *does* magically show up that close to the boat, the MQ25 sucking up a missile is actually a great outcome.

    - tanking in a turn is a non-event. It happens everyday

    - “ Of course, this raised the question, if the shipboard controller can’t communicate with the unmanned tanker due to enemy disruption of communications, why would we think that we’ll be able to communicate with the manned aircraft to tell the pilot to redirect the unmanned one? That’s a logical inconsistency. Military thinking just teems with this kind of logical inconsistency.”

    Because it’s much harder to jam communications everywhere all at once than in very specific areas (ie where the boat is). And there are lots of ways to get messages to manned aircraft. And it’s harder to jam the comms of things close together (like an aircraft and the MQ25 for example) than for things far apart (like maybe the boat and an MQ25.

    All that being said it seems sort of like a dumb capability.

    ReplyDelete
    Replies
    1. "as a practical matter, with only 14,000 pounds of give the MQ25 is going to spend its life in the “immediate” vicinity"

      No, the 'specification' is 14,000 lb at 500 miles. The MQ-25 is intended for delivery at distance.

      "tanking in a turn is a non-event."

      Never said it couldn't be done. Straight and level is easier. This simply illustrates one of hundreds or thousands of possible command situations that require a nuanced response. Don't be pedantic.

      "harder to jam communications everywhere"

      There are many reasons why communications can be disrupted. For example, the UAV may have lost comm capability and NO ONE can communicate with it.

      This was an extremely low quality, pedantic, argumentative comment that did nothing to further the discussion. There are many discussion-worthy aspects to unmanned tanking, control schemes, software direction, etc. but you addressed none of them. If you'd like to comment again, please offer something informative and useful that furthers the discussion. Thank you.

      Delete

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