J-31 in PLAAF service makes perfect sense. It is your post that doesn't make any sense.
I'll try to be as short as I can but it's not something that can be explained in a single post very well.
Aerial warfare has never been a problem of individual combat. It's a collaborative effort through and through. And currently that collaboration will extend to unmanned systems some of which will follow the manned fighter while others will act autonomously. It's also a very complex problem requiring a lot of preparation and you don't understand it at all.
The current plans for F-35s for example involve a standard set of two UCAVs developed under CCA (Collaborative Combat Aicraft) program. In the future there might be more e.g. two UCAV carrying weapons or jammers and four UAVs serving recon or decoy roles.
NGAD and B-21 is already being programmed with the capability to coordinate several CCAs.
All that to perform the same mission that 30 years ago would be executed by a single aircraft. All because the capabilities of radars, missiles and C4 systems has grown well beyond what a single aircraft can do. So for the aircraft to be able to perform the mission successfully the counter-capability needs to be distributed across several other aircraft because one is not enough.
There simply isn't room in the future for manned aircraft not capable of collaboration with unmanned systems because of scale of attrition and the complexity of systems that need to be present in combat space to achieve success with any likelihood. And even then the basic mission team is always going to be two pilots i.e. two F-35s mean four CCA and likely eight light UAVs or however many MALD-type drones.
Another element that most people can't imagine is the vulnerability of comm networks relaying information over distance. Presently EM spectrum is a separate domain of warfare where you fight for the ability to use EM spectrum to carry information. You physically have to fight for the ability to transmit signal between nodes in your network.
Information itself is also a domain of warfare and the ability to process information in limited time is limited by physics. You can put only so much computing power into an aircraft and it isn't much. Therefore any significant computing operation must be conducted by a dedicated system on a larger aircraft or on the ground or it is conducted by a human present on the spot.
All of these have their risks but the risk to disruption of communication between UCAVs and a distant and expensive command post in the air or on the ground is significantly higher than the risk to a pilot in a fighter managing the decision in the battlespace where the pilot has constant awareness and needs only to control a handful of UAVs.
Largely it's a question of signal strength and decision time. A ground target has the capability to generate more power and more intelligence than an aircraft. Therefore - in a hypothetical scenario - it can intermittently alternate between high power wideband jamming and scanning the battlespace for targets. It can identity them faster and engage them more effectively because physics allows for more energy and computing power on the ground than in the air which will reflect the scenario in most cases.
Whenever that happens any long range network of drones will be neutralised by the jamming because as soon as the jamming is on they have to act autonomously and as soon as the jamming is off they have to reacquire their mission parameters while being engaged by enemy defense systems.
It's a mathematical problem that can be easily solved by either allowing the drones to act completely autonomously or having a man in the loop in the area. So far nobody is risking full autonomy because we're nowhere near an AI that can replicate human decision and which fits in a standard UCAV. Without this level of intelligence "full autonomy" is as good as full autonomy of an unguided missile. The AI that you are familiar with requires gigantic clusters operating in climate-controlled conditions to achieve their efficiency and is trained on gigantic datasets. Any autonomous system can be easily deceived by spontaneous random masking of target information with active protection measures. It's too easy to deceive an AI that for an AI to break the deception. It's a physics problem and whoever has to battle gravity while solving it loses by default.
Which is why we're nowhere near getting rid of pilots actually flying the mission. And if you need to have pilots you need to have manned aircraft.
And this is where we come around to J-20 and J-31 and this:
The following is a collection of maps, tables and quotations from publicly available sources with heavy emphasis on Government Accountability Office's report "Operation Desert Storm - Evaluation of the Air Campaign". I divided the whole set into five posts: air superiority over Iraq -...
www.sinodefenceforum.com
I made that thread because from what I've seen you people love to fantasise about unrealistic scenarios and never bother to sit down and try to crunch actual numbers. So there, Desert Storm in actual numbers. This is closest to what any future war will be like.
War never changes. It is a numbers game. The side with better numbers wins. But you have to know which numbers matter and which don't.
Every weapon is a machine that has its life cycle both long-term and short-term between mandatory checks and likely breakdowns. The more you fly the more you fight. The more you fly the more you wear them out. Pilots are trained
before the war. Machines wear out
during the war. A pilot can fly fewer hours than a machine over a day but more days than a machine over a week of constant operations. So if for every 3 J-20s PLAAF can get 5 J-31s these additional 2 fighters can be the difference between success and failure of an operation. On the other hand there are plenty of missions that require VLO (J-16 can't do them) and J-31 can perform them as well as J-20 so building J-31s instead of J-20 to run those missions frees up resources for more J-20s for other missions.
Also J-31 is likely to be
better aircraft than J-20. It will have worse
performance but it will have better
reliability and
economy which means that not only will there be more of them but they will be easier to fix, maintain and replace. Those things win war. Don't believe me? Here:
View attachment 125348
Who won this one?
Desert Storm too was won not by F-117s or F-15s but by F-16s. Lots and lots of sorties by F-16s. Close to twice as many sorties as the other two combined. And also there were more F-16s in theater than the two planes combined because they were cheaper to make and maintain.
I will be very surprised if PLAAF doesn't get J-31 by the end of the decade. 2030s will be the era of UAV collaboration and you can't do it with non-VLO aircraft. That's a recipe for getting sniped and wasting every single UAV.
I also don't understand the fetishisation of J-20. It's the same ridiculous circlejerk as with F-22. It's the first VLO/5gen China made. It simply can't be better than the next one. You missed the whole
learning from use process for reasons that I really don't want to get into. Why wouldn't PLAAF a cheaper plane that is better?
That's all.
