Airbus is preparing two uncrewed combat aircraft
airbus.com174 points by phasnox a day ago
174 points by phasnox a day ago
There are multiple interesting developments wrapped together here.
First, these are intended to be "loyal wingman". They'll be commanded (but not really remotely controlled) from manned fighters nearbyish. Presumably, the "shoot authorization" will be delegated down to the pilots.
Secondly, the actual unmanned platform (the Kratos Valkyrie) is also part of a program of record for the USMC (US Marine Corps) to act as a partner SEAD (suppression of air defence) vehicle.
Thirdly, the "MARS" system chattered about looks to be Airbus' open architecture /system of systems pitch that they were developing for FCAS (the European 6th generation fighter program). MARS and all pitches like it are about ways to make individual platforms as software defined as possible, and to get different platforms/instances to really data/function share as much as possible.
If this program goes well, it shows that Airbus' MARS has the flexibility and capability required to just... layer into/ontop of some random other vendor's hardware/software and then "just work". I think it would be major demonstration/validation of the work.
Why do I get the feeling that the market shifted beneath their feet to drones and these old aircraft companies are using "loyal wingman" to make a half-hearted half-way play between old/new products to stay relevant, which just buys them time to keep selling expensive jets... until pure drone upstarts start eating their lunch.
Like when Blackberry tried to make BlackBerry Storm after iPhone and Blockbuster tried to make Blockbuster Online after Netflix.
Technology shifts rarely wait for these stodgy middle ground transitionary products to find a market.
Roughly everyone expects the 6th generation fighters (the ones currently in development like F-47) to be the last manned generation. Most observers expect many/most 6th gen fighters to become optionally manned within their life span.
The real question is basically - is full autonomy both technically possible and culturally/politically acceptable within 5, 10, or 20 years? Because full autonomy isn't really ready now (or else we wouldn't need hundreds to thousands of drone operators in the Ukraine war). And at least the USAF doesn't think remote control will let them do what they need (which is to fly from Japan to Korea or Taiwan, or Philippines to Taiwan, and contest/control the skies in the face of a basically peer adversary).
Because no one knows that answer, everyone (governments, militaries, manufacturers) is hedging, and CCA is part of that hedge.
I think we are underestimating and/or forgetting that the enemy gets a vote, and remote piloting something from Virginia all the way out to Japan or Korea or Taiwan involves many signals integrity steps along the way. This is to say that you should assume these signals are interrupted and you will not be able to maintain continuous control of the aircraft from whatever datacenter box the "pilot" sits in. That means fully autonomous decision making, functionally for the entire journey, and independent release authorization.
>or else we wouldn't need hundreds to thousands of drone operators in the Ukraine war
I don't think this is the reason the systems are not fully autonomous right now ("fully autonomous" here meaning that they can complete the kill chain independently, no HITL). Even if we assume it true that the drones are not "good enough" to be at parity with a human operator, if you had an essentially limitless amount of them, would you really waste the manpower on operating them in FPV mode? You would not, you would completely saturate the battlefield with them. Thus, as it was in the beforetimes and ever shall be, logistics wins wars.
The reason that FPV drones are so easily disrupted is that they are too light to carry anything more than a radio and fly low.
Disrupting the signal for a normal-sized aircraft is much harder. If you're flying at 10s of thousands of feet and have a line of sight to multiple satellites it's going to take some serious weaponry to disrupt that.
True. But the next rung up the escalation ladder is of course disrupting the satellites.
I envision them all gone seconds into any large scale war.
The G forces are another thing. I wonder why they aren't stsrting wth missle platforms instead.
Sure, winged flight has uses, but taking a missle platform, adding small munitions instead of a big bang?
I'm not sure about this. Space is big and these satellite constellations are getting very large, with lots of redundancy. I know I'm sort of arguing against my previous point, but bear with me for a sec. You'd need an anti-satellite system that either destroys them kinetically (accepting the cost of the debris field) or one that breaks them electronically (an EMP or another device that defeats them electronically). The United States' underlying philosophy on advanced weapons has, for a long time, been precision so I could see the emergence of in-orbit interception & defeat/disable platforms. But you'd need a lot of them for the doctrine to be effective, which means a lot of mass-to-orbit logistics. Adversaries do not have this, so I would expect e.g. PRC to have an alternate strategy of rendering entire orbits unusable or dangerous, which I think is easier.
Regarding your missile platform question, there are several companies that already manufacturing loitering munitions, and long-range loitering cruise missiles are on the roadmap, so to speak.
> I wonder why they aren't stsrting wth missle platforms instead
Price and ease of manufacture. Missiles are expensive and hard to build.
Latest FPV drones in Ukraine became much more resistant to electronic countermeasures. Plus other drones are used as retranslators.
Seems they are using kilometers of fibre optic cables, so they fly tethered and communication can't be disrupted.
I'd hate to be part of the clean-up crew when that war ends. Broken fibre is nasty stuff.
I believe they’ve also deployed hybrid solutions: FPV fibre drones launched and piloted via link to an unmanned platform.
So a drone boat with good/secure signalling pulls up and a bunch of fibre optic drones launch from that point penetrating inland.
FPV drones cannot have powerful GPU yet to enable truly autonomous flight. And the issue is not only weight/energy restrictions, but also cost.
Autonomous flight is significantly easier than autonomous driving. You just fly between points in space, and there's nothing but air inbetween. The ground control handles most of collision avoidance, and if that's not available, it's easily achieved by moving 300ft/100m up or down.
True, but take into account that plants need to be able to fly/fight with instruments only and without vision.
Also dogfights are much rarer now, most people just fling rockets at each other (so you know how much these cost, a b200 seems cheap in comparison)
You don't need a super powerful GPU to do computer vision. There are cheap small devices that can do it.
> This is to say that you should assume these signals are interrupted and you will not be able to maintain continuous control of the aircraft from whatever datacenter box the "pilot" sits in. That means fully autonomous decision making, functionally for the entire journey, and independent release authorization.
Only if every mission is absolutely critical. If disruptions are rare then you don't need autonomy.
Or more interestingly with the low-earth sat/data network. Seeing as projectssuch as starlink are basically mil in nature with a side of barely profitable civilian use. The whole data centers in space makes more sense. These are not for running cat blogs and video streaming , which is waht they are/will be marketed as. Realworld application will always be a command and control node spanning the globe for the mil use. And as its rolloed out globally can basically provide jammingfree links for the autonomous commands from space.
>Roughly everyone expects the 6th generation fighters (the ones currently in development like F-47) to be the last manned generation. Most observers expect many/most 6th gen fighters to become optionally manned within their life span.
The said that about the 5th though. Like I've personally talked to people who were actively working on the F35 and they were saying "last manned aircraft" in like 2011ish.
I expect autonomy to be a long steady improvement of taking on additional subroutines of increasing complexity of decisions being made along the way. Fly here, land there, kill that, go over there without being detected, etc, etc, until humans are making only a select set of decisions that will probably be randomly sprinkled at the high and low levels.
Kind of like how when we build a brick wall the "vision" and the actual laying of bricks still get done by human but all the intermediary steps are drudgery that can be trivially automated (not to say they are all automated, just that they could be if labor $$ vs software $$ penciled out that way)
Yeah, I remember thinking that about the 5th gen myself. But I think part of that way unclarity around when the 6th gen would appear. 2011ish would be a weird timeframe.
Obama didn't announce the "pivot to Asia" until 2012. A lot of the world was still believing in the whole unipolar moment thing. The F-35 hadn't even started training squadrons in service yet. 6th generation probably felt really far away. 2016 onwards has been a major acceleration in all sorts of ways.
In my eyes, 6th generation was really ignited by the focus on China. The PCA/NGAD/F-47 was first out the gate, and really set the tempo and got everyone else going.
I agree with your estimation of the likely development path. I expect that approach to merge/converge from the other direction - I expect there to also be a parallel path of fully autonomous systems growing to occupy ever greater mission sets. Imagine telling a drone "I need an ELINT mission in the area and timeframe" (and it plans its own launch time, flight route, calculates fuel loads, communicates with ground staff), or "I need you to airdrop this supply pallet".
> And at least the USAF doesn't think remote control will let them do what they need (which is to fly from Japan to Korea or Taiwan, or Philippines to Taiwan, and contest/control the skies in the face of a basically peer adversary).
I mean, they wouldn't think that, would they? It would put their pilots out of a job. But most flying has been done by autopilot long before AI, and even if/when you need a human in the loop, why would you want to put that human in the cockpit rather than safely in Virginia?
My read is that the "loyal wingman" thing is a ploy to get around all the pilots / expilots in Air Force brass who might otherwise gatekeep anything they think is a threat to the careers of human pilots. These people want the Air Force to be about hotshots flying planes; this is part of the reason the US spun off the space stuff, because Air Force brass is institutionally incapable of taking anything other than manned flight seriously.
Really loyal wingmen are just an extension of jets.
Ultimately a carrier strike group could achieve many of the missions a F35 can through cruise missiles, ballistic missiles etc.
What an F35 provides is a sensory platform deep into enemy airspace.
But with the F35 being very expensive, and required to stay silent in RF to maintain stealth it's further desirable to have a loyal wingman out there.
The whole thing becomes one large sensor network with th added weapons.
It makes most sense in a near adversary situation and as it stands that particular scenario is why the F22 is not exported.
Manned-Unmanned teaming is not a new concept created in the last couple months to placate fighter pilots in the age of ai. With 5th generation fighter using datalink they to use the active radar in far away AWAC planes for targeting so the stealth fighter can get closer to the enemy without breaking cover by turning on active radar.
If you can outsource the radar on a jet it is not a huge leap in logic to put the very hot missiles onto a unmanned aircraft. All of these concepts where written up 20 years ago by both china and the US
> With 5th generation fighter using datalink they to use the active radar in far away AWAC planes for targeting so the stealth fighter can get closer to the enemy without breaking cover by turning on active radar.
What benefit does a human pilot offer in this case? Are going to be using their eyes to track their location or see a Chinese fighter jet launching a standoff antiair missile at them? Drones can do AWACS and deep sensor roles, with a pilot and sensor operators far away from the planes.
> All of these concepts where written up 20 years ago by both china and the US
Defense contractors and the gov planners are often the same group of people, the same small community, there's not that many vendors. They show up at defence trade shows and see what industry is offering them. They tend to stick with the same big safe companies that change slowly. Bold ideas are infrequent. The smaller countries can take those risks easier than America.
mistakes in A/A combat can have serious repercussions. not only loss of expensive air vehicles, but things like civilian airliners.
'loyal wingman' gives the kill / no kill decision to an Air Force officer. And having the decision maker geographically close eliminates jamming, delays, and the requirements to have a satellite infrastructure (like is required for Predator UAV's).
i hope we never assign a piece of code, AI or not, to be the decision maker.
> i hope we never assign a piece of code, AI or not, to be the decision maker.
This is already a past station, just not at Airbus.
> just not at Airbus
Airbus has publicized that it is working on a Project Maven style project with France's DGA [0][1].
Thales also publicly launched and demonstrated SkyDefender a couple days ago [2].
Mistral AI also announced in January 2026 that it is working with the DGA to productionize it's models for military applications [3] - ironically similar in manner to how the DoD was using Claude but is now using Gemini and GPT.
No country is going to leave networked, autonomous offensive and defensive capabilities on the table.
[0] - https://www.reuters.com/business/aerospace-defense/airbus-wi...
[1] - https://www.janes.com/osint-insights/defence-news/defence/ai...
[2] - https://www.janes.com/osint-insights/defence-news/air/thales...
[3] - https://www.linkedin.com/posts/marjorietoucas_were-happy-to-...
Yes, we are ~4-5 years into AI kill-chains now, though maybe only 1-3 with full autonomy.
You haven't been paying attention. We are at least 47 years into AI kill chains.
https://www.vp4association.com/aircraft-information-2/32-2/m...
That's not really what they meant. They meant that the weapon is guided by software that decides which targets to pick and autonomously makes that decision without a human in the loop. The device seeks you instead of you going to it.
A landmine has no friend-or-foe-or-noncombatant decision engine, it will kill you or maim you just like it will kill or maim the guy that laid it or any other passer by.
You missed the point. The Mk 60 Captor is not a "land" mine. It is guided by software and autonomously makes the decision to launch a homing torpedo without a human in the loop.
A homing torpedo/release mechanism is not AI by the normal definition of the word. You're welcome to redefine words as much as you want but it's a bit silly. We also don't use that term for heatseekers or for line followers.
The 'signature' bit is interesting, but I'd still not label that AI, and neither does anybody else. It is a loitering munition, I'll give you that, and I think that that brings it closer to the 'mine' definition of things than the 'AI killbot'.