F-22 now critical to survival of USAF

[Androsphinx]

You can't make the maths any simpler. So why do the rich, white middle-aged folk in Washington, D.C. not grasp this fact? They're quite happy funding the LCS, and not something that actually has far more relevance today and urgency in deployment.

Because they don't care. Sitting on a Defence committee, or even being in the Defence department, is nothing to do with a genuine interest or concern with national security and the military, but rather a way to gain contacts, favours, influence and a media profile.

[Ritterin Sophia]

Wasn't the F-35 supposed to be like a replacement to the F-18?

For the US it's like this. USAF: F-16 & A-10->F-35A, USMC: AV-8B & F/A-18->F-35B, USN: F/A-18A/B/C/D->F-35C. For the Royal Armed Forces the F-35B is to replace both the RAF and RN's GR7/GR9's.

And why the hell is it so crappy? What did they remove from it?

I heard most of the problem's with the JSF Program come from the F-35B and STOVL part.

[Illuminatus Primus]

Let me ask a silly non-military person question. In light of how expensive the F-22's are is there a negative to ordering a second tier fighter similar to the F-15. A far cheaper but still lethal fighter that can sort of be a "filler" fighter so that instead of a squadron of F-22's we can field say 3 squadrons of this cheaper fighter relegating the F-22 to an elite fighter status with this second tier cheaper fighter a "go to" work horse.

I know we as a military always want to fly the best and most awesome fighter but given that the F-15 is over 30 years old and still a front line fighter in some countries and obviously still considered a good enough fighter for us to field in large numbers why not order up a small fleet of modernized F-15 like clones.

Just because the technology and capability requirements are old, designing, testing, and fielding a from-scratch ersatz F-15 would make any relative financial advantage in per unit cost irrelevant. You might as well buy more F-22A's, as we have the fighter now and the production is already running.

Does that no sense from a military standpoint for some reason? Or if we positively need to have a killer plane how about a second tier F-22 "lite" not so many stealthy expensive features but with the same kind of lethality or at least better than most 3rd world militaries will ever field.

There are quite a few countries with Sukhois or other fighters which can compete favorably with the F-22, and that's disregarding whomever Russia sells its new generation fighters too.

[Illuminatus Primus]

[snip]

Also, the F-22A is so good that its literally worth more than 10 F-15s in the sky; it'll kill as many enemy fighters and without getting killed, either.

[Stravo]

[snip]

Also, the F-22A is so good that its literally worth more than 10 F-15s in the sky; it'll kill as many enemy fighters and without getting killed, either.

After reading through some comments on the F-22's performance I had no clue it was that good in comparison. Are these real tests or mock ups made to make the F-22 look good?

[The_Saint]

There are plenty of claims coming out of test pilots that in training scenarios the F-22's went up 2 vs a dozen F-15's several times and always won.. the juries out on what rules of engagement were used. No specific official results have been released.

The jury is still not definitively set on what the latest Flanker variants vs F15 results are as well.


For the RAAF the F-35 is to replace the F/A-18's which is arguably a fair trade .. what I don't find palatable is that the F-35 is also to replace the F-111.

[Wanderer]

The UCAVs I envision would be autonomous when need be and not something coming into service for a while yet. The idea of having UCAV wingmen isn't far fetched today, but a fleet of them being controlled from afar is not a good idea. The Predator and Reaper units are one thing and I'm under no impression they make anything more than police units in contested zones rather than full fledged combat units.

It would be nice, alas, no super AI for us yet.

And how superior will this be to a human pilot who has a radar to assist him, his eyes (if the plane is a two seater he has a buddy watching his back) if the radar is jammed and a wing man in an over watch position covering his blind spots against a drone that has a narrow field of vision and its radar jammed, plus unable to take advantage of regular maneuvers used by human pilots to check their blind spots?

What exactly am I missing here

[Lonestar]

And how superior will this be to a human pilot who has a radar to assist him, his eyes (if the plane is a two seater he has a buddy watching his back) if the radar is jammed and a wing man in an over watch position covering his blind spots against a drone that has a narrow field of vision and its radar jammed, plus unable to take advantage of regular maneuvers used by human pilots to check their blind spots?

What exactly am I missing here

Most UAVs are set to wipe their harddrives and ditch if they lose contact. Presumably a manned aircraft would turn around and return to base.

[Starglider]

And how superior will this be to a human pilot who has a radar to assist him, his eyes (if the plane is a two seater he has a buddy watching his back) if the radar is jammed and a wing man in an over watch position covering his blind spots against a drone that has a narrow field of vision and its radar jammed, plus unable to take advantage of regular maneuvers used by human pilots to check their blind spots?

What exactly am I missing here

A computer has the potential to react much, much faster than humans. It can have literally 360 degree spherical vision, in multiple areas of the EM spectrum, and pay attention to it all at once. It is effectively immune to G force, weighs much less than a human+cockpit+ejector seat and knows no fear. It can be programmed with a library of every aerial combat maneuver ever divised and records of every exercise and simulation ever conducted by its home nation. It can simulate thousands of possible outcomes to the dogfight and make a calculate probability assessment of the best way to kill the enemy.

In theory at least. In practice, we've got a lot of software engineering still to do. But I'm pretty sure air combat doesn't require 'real AI' (i.e. sentient AI) any more than playing chess at grand master level does. I'm sure pilots will tell you that no machine could cope with it, but they're not qualified to predict future trends in software. A gradual increase in capability at the fundamental level becomes visibile to non-specialists as thresholds being crossed with relative suddeness; self-driving support vehicles are going from vapourware to practicality in a decade, automated gun systems are going to make a similar transition very soon (South Korea rushed the alpha versions into service on the DMZ). UAV AIs capable of beating most human pilots could come into being quite quickly - particularly given that existing hardware will be easy to retrofit with it.

[Illuminatus Primus]

After reading through some comments on the F-22's performance I had no clue it was that good in comparison. Are these real tests or mock ups made to make the F-22 look good?

One thing is it really isn't so much about a man's skill or luck of the draw like in the movies. Not tight-maneuvering and so-on. The plane with the best sensors and computers and missiles will win. The F-22A has a much more advanced radar than the F-15; is equipped with the newest radar-guided beyond-visual-range missiles; can launch them from further than the F-15 (it can supercruise, that is, go supersonic for extended periods without using afterburners - the F-15 cannot and speed of the plane is added to the speed of the missile obv); and the F-22A itself hampers the performance of the enemies computers, sensors, and missiles because it is so fast and stealthy. Basically old planes are bringing knives to a gun fight when they're fighting a F-22A.

[Sidewinder]

Now if only the troops were brought home from Iraq, the US might have enough money to pay for the Raptors they need.

[brianeyci]

A computer has the potential to react much, much faster than humans.

Even if that happens, you will still need aircrews for abort. I believe Stuart said one of the advantages of bombers compared to ICBM is bombers can be recalled. And if there are aircrews, they had better be highly trained enough to dogfight and not rely on technology and be prepared to manual override. Everything you listed: fear, g-force, weight, life support, probability, faster reaction, are all trivial factors compared to ability to recall and having a human override. Even if AI's are not necessary for fighter planes, according to you it would take just the right software to turn the plane sentient assuming the right hardware. Star Trek hologram syndrome could come into play, with some software engineer who has no fucking idea he's created a sentient being uploading software into a machine of death. This is not something anybody should want.

[Admiral Valdemar]

The last thing anyone should do is upload an AI like that into a nuclear capable aircraft and let it off the leash. Hell, even a conventionally armed one could pull an EDI from Stealth and end with several international incidents, but that's a damn sight better than a city going up in an instant sunrise because your computer pilot felt overly patriotic and single-minded in some training exercise.

[thejester]

[snip]

Also, the F-22A is so good that its literally worth more than 10 F-15s in the sky; it'll kill as many enemy fighters and without getting killed, either.

After reading through some comments on the F-22's performance I had no clue it was that good in comparison. Are these real tests or mock ups made to make the F-22 look good?

They're 'real' exercises as far as I can tell. It makes sense: the F-22 is simply invulnerable if it's beyond eyesight, and it has the power to make sure it remains beyond eyesight.

[phongn]

They're 'real' exercises as far as I can tell. It makes sense: the F-22 is simply invulnerable if it's beyond eyesight, and it has the power to make sure it remains beyond eyesight.

Even close in, the F-22 has so much thrust available - and its 2D TVC - that it has a marked advantage in a dogfight, IIRC.

[Siege]

Even if that happens, you will still need aircrews for abort.

How is that? If you can tell the human operator to abort, then you can radio the drone to abort as well. Drone bombers are not one-shot missiles, they're unmanned aircraft. I don't see any reason why you shouldn't be able to recall them at your leisure.

[The_Nice_Guy]

A friend of mine once remarked that each generation of aircraft for the US airforce seems to reduce their numbers by a factor of ten(or something like that). Where we once had about 500~600 F-15s, we might now only have 200 F-22s.

Which means that in 2 or 3 more generations, the USAF might only have a single super-fighter...

Quality is important, but numbers count too.

[TheMuffinKing]

A friend of mine once remarked that each generation of aircraft for the US airforce seems to reduce their numbers by a factor of ten(or something like that). Where we once had about 500~600 F-15s, we might now only have 200 F-22s.

Which means that in 2 or 3 more generations, the USAF might only have a single super-fighter...

Quality is important, but numbers count too.

Maybe we'll have a"Firefox" of our own?

[Wanderer]

A computer has the potential to react much, much faster than humans. It can have literally 360 degree spherical vision, in multiple areas of the EM spectrum, and pay attention to it all at once. It is effectively immune to G force, weighs much less than a human+cockpit+ejector seat and knows no fear. It can be programmed with a library of every aerial combat maneuver ever divised and records of every exercise and simulation ever conducted by its home nation. It can simulate thousands of possible outcomes to the dogfight and make a calculate probability assessment of the best way to kill the enemy.

In theory at least. In practice, we've got a lot of software engineering still to do. But I'm pretty sure air combat doesn't require 'real AI' (i.e. sentient AI) any more than playing chess at grand master level does. I'm sure pilots will tell you that no machine could cope with it, but they're not qualified to predict future trends in software. A gradual increase in capability at the fundamental level becomes visibile to non-specialists as thresholds being crossed with relative suddeness; self-driving support vehicles are going from vapourware to practicality in a decade, automated gun systems are going to make a similar transition very soon (South Korea rushed the alpha versions into service on the DMZ). UAV AIs capable of beating most human pilots could come into being quite quickly - particularly given that existing hardware will be easy to retrofit with it.

Okay and what will it do to counteract jamming of its sight or protect itself against such jamming or even hacking?

[TheMuffinKing]

Okay and what will it do to counteract jamming of its sight or protect itself against such jamming or even hacking?

That is a good question.

I suppose multiple redundant systems an encrypted datalinks, mush like what we have today could still be useful. If we go with automated fighters, they will probably have to have similar sensors to whats in service now, at least. Weapons deployed by these planes could be guided by infrared, gps, active/passive radar (from onboard or another aircraft or ground station), or even remote viewing from a secure location like a video game. Thats just some of the stuff I could think of. Many of these could suffer interference from jamming, but this could be mitigated by the use of an A.I. allowing the unit to operate somewhat independently.

As for jamming/hacking, I suppose the only real defenses would be sophisticated encryption of data concerning operation of the aircraft, communications security guys could better answer the question.

[tim31]

There are plenty of claims coming out of test pilots that in training scenarios the F-22's went up 2 vs a dozen F-15's several times and always won.. the juries out on what rules of engagement were used. No specific official results have been released.

The jury is still not definitively set on what the latest Flanker variants vs F15 results are as well.


For the RAAF the F-35 is to replace the F/A-18's which is arguably a fair trade .. what I don't find palatable is that the F-35 is also to replace the F-111.

Let's just see where the Rudd Government takes us on these issues... I don't think the ADF is going to get half the stuff that was in the pipeline.

[Sarevok]

Okay and what will it do to counteract jamming of its sight or protect itself against such jamming or even hacking?

Starglider is talking about autonomous AI not remote controlled kites of doom like the Predator. Presumably these would be no more vulnerable to jamming than a manned jet.

[General Trelane (Retired)]

An investigation of the Nov. 2 crash shows the F-15 that broke apart over eastern Missouri had a fault in a crucial support component called a longeron, a structural beam that serves as part of the spine of the aircraft. F-15s have four longerons around the cockpit.

Air Force officials have not yet learned how a defective beam came to be installed in the plane, which was manufactured in 1980. But Air Force officials emphasized that the age of the airframe, combined with the faulty part, puts the older F-15s at risk.

"This airplane broke in half because of a fatigue crack," the officer said.

The Air Force has found cracks in nine of about 180 planes that remain grounded, but it thinks more have faulty structural beams.

(Emphasis mine).

What makes the reporter think that a defective beam was installed in the plane? Probably have no idea what fatigue cracking is. After more than 25 years of service, we really shouldn't be surprised that highly stressed structural components are suffering fatigue. The 'defective' beam was likely perfectly fine when it was installed.

[Beowulf]

It should be noted that the USAF wants at least 381 F-22As. This number is driven by the number of squadrons required by each Air Expeditionary Force, as well as the number of AEFs the USAF has, supplemented by aircraft for training and test purposes. This is a difference of two hundred aircraft from the currently allocated buy.

[brianeyci]

Even if that happens, you will still need aircrews for abort.

How is that? If you can tell the human operator to abort, then you can radio the drone to abort as well. Drone bombers are not one-shot missiles, they're unmanned aircraft. I don't see any reason why you shouldn't be able to recall them at your leisure.

Well if you can tell an ICBM to launch you can tell an ICBM to disarm, right? I always just assumed this was for security reasons.

Regardless, AV is right -- there's absolutely no reason to upload an AI or even near-AI into F-22. The example Starglider listed -- DMZ -- that is because the DMZ is no-man's land and has to be constantly watched, a perpetual state of war. It makes sense to automate there for repetitive, dangerous tasks. This is a world of difference to accepting AI pilots, pilots who have to be unquestioningly loyal and prepared for dynamic situations. If an AI increases the F-22's kill ratio from 30 to 1 compared to fourth generation aircraft to 32 to 1, big fucking deal. Third world nations do not field that many modern aircraft, and there is more than one F-22. Because of n-squared law air defenses will be completely destroyed if it's 30 to 1 or 100 to 1 with no losses to the American side. It's possible AI gives no real kill ratio advantage. No advantage, no point.