UAV- discussion from way back continued

[Skipjack]

Immaterial with respect to my argument since the F-35 is not an air superiority machine, rather a "jack of all trades and master of none".

Well that would mean that carriers wont have any air superiority jets on board in the future?
Because the carrier version of the F22 was cancelled IIRC and F35s were put in its place, no?

For surface attack, recon, etc., not air-to-air combat.

I think that this is only a matter of time. Seeing how this trend is ever accelerating, it is matter of VERY little time too, at least in my opinion.
Of course all the fans of the gay Tom Cruise "you can always ride my tail" Top Gun romanticism will cry and whine about it. Personally I dont care. It is the age of the geek baby and now we get to fly the fast machines! Ooooh yeah!

[DeltaV]

Well that would mean that carriers wont have any air superiority jets on board in the future?

They don't have any now, unless the opponent doesn't happen to fly the latest Sukhoi models. "Air superiority" carries a different meaning between USN and USAF, mainly because the limited carrier deck space means USN fighters must be multirole by design. Only USAF has (had) the luxury of fielding a no-compromise air-superiority fighter.

I think that this is only a matter of time. Seeing how this trend is ever accelerating, it is matter of VERY little time too, at least in my opinion.

So you've abandoned your initial assertion?

...I was actually thinking of the next generation of UAVs, but I guess the current one is good enough

Of course all the fans of the gay Tom Cruise "you can always ride my tail" Top Gun romanticism will cry and whine about it. Personally I dont care. It is the age of the geek baby and now we get to fly the fast machines! Ooooh yeah!

Have fun playing with your joystick as you win air battles using time-delayed, jammable data links.

[krenshala]

Only USAF has (had) the luxury of fielding a no-compromise air-superiority fighter.

Why is it the guys with control of the purse strings always seem to be the type that think you need only an adjustable wrench, and nothing else, in your toolbox? How can you fix things if you don't have the correct (best) tool for the job at hand?

At least we got some of the F22s in service before the F15s starting falling out of the sky from structural fatigue (... again).

[Skipjack]

They don't have any now, unless the opponent doesn't happen to fly the latest Sukhoi models. "Air superiority" carries a different meaning between USN and USAF, mainly because the limited carrier deck space means USN fighters must be multirole by design. Only USAF has (had) the luxury of fielding a no-compromise air-superiority fighter.

Ok, so can we agree that I see things with the eyes of the Navy and you see it with the perspective that the airforce may (or may not) have.

So you've abandoned your initial assertion?

My initial assertion was that UAVs are the future and manned jets, be it fighters or bombers are already last generation. I predict that the F22 an the F 35 will be the last manned fighter planes the US will build.
That was my original assertion and I stick with that.
I started this thread to bring a recent example for this trend.

Have fun playing with your joystick as you win air battles using time-delayed, jammable data links.

Today's manned fighters already target jammers and radar stations semi automatically and these things make wonderful targets for missiles. But I am sure that you knew that already.
The delays are minimal, as you also know and can be further reduced by moving the controler closer to the combat zone (could still be way out of danger though).
Again, I am sure you know that too.

[DeltaV]

My initial assertion was that UAVs are the future and manned jets, be it fighters or bombers are already last generation. I predict that the F22 an the F 35 will be the last manned fighter planes the US will build.
That was my original assertion and I stick with that.

I was referring to this assertion from the opening entry in this thread:

I guess the current one [generation of UAVs] is good enough [for air-to-air combat, since you posit UAVs replacing F-22s]

Today's manned fighters already target jammers and radar stations semi automatically and these things make wonderful targets for missiles. But I am sure that you knew that already.

And the missiles always hit their target, yes?... and a smaller UAV can carry more missiles than a larger manned fighter, right?... and no adversary would ever think of jamming UAV data links from orbit, or from aerial platforms outside the combat zone, that would just be unsporting.

The delays are minimal, as you also know and can be further reduced by moving the controler closer to the combat zone (could still be way out of danger though).
Again, I am sure you know that too.

Lockheed-Martin - Design of UAV Systems: Time Delays and UAVs
Slide 39: Communications Latency, msec
LOS 2
LEO 10
MEO 100
GEO 200-300

When OTH control latency > 40 msec, direct control of a UAV is high risk (except through an autopilot)

And that is for a slow turning ISR/ground-attack type of UAV, not the >10g type needed to defeat an F-22 class manned fighter.

A commercial fly-by-wire flight control system has control loops on the order of 100 Hz (10 msec frame time). A no-compromise, 9-g manned fighter has a higher frame rate, but that number is probably classified, so let's assume 200 Hz (5 msec) for its FCS. The dogfighting UAV is smaller, with higher dynamic rates than the manned fighter, so let's assume its FCS runs at 250 Hz (4 msec). Under extreme maneuvers a loss of even one FCS frame is usually considered unacceptable, but lets be generous and assume your killer UAV can tolerate a one-frame loss while pulling high g and high rates in battle. To be extra nice to you, we'll also ignore latencies introduced by encryption, background noise, hardware delays, counter-jamming algorithms, etc.

So any latency over 4 msec means your UAV from Hell is toast.

That rules out LEO, MEO and GEO satellite links, and limits you to a Line Of Sight data link. But I am sure that you knew that already.

"Oh, please, Mr. PLAAF General, sir, please delay your J-20 attack until our UAV Ground Command Units are shipped in and set up." Nyuck.

[ladajo]

I have a pdf of the most of the Janes stuff. Just need to get it posted somehow.

You are arguing for man in the loop with the latencies. The trend and current practice is for autonomy. Mouse click vice joystick. We have fielded systems that do most of the grunt work without man in the loop. So the delays are not that important.

That said, policy still dictates weapons release by man in the loop. This is the sticking point. As for air to air, once weapons release gets sorted out, it is not a big leap from where we are with current tech. To paraphrase a flag officer, "when is the last time we put a man inside an ICBM, SLBM or Tomahawk?". The next gen AF long range bomber is fragged to optional manned. Contested environment is what it is all about in unmanned circles today. Some of the experts argue the tech is there, but the will is not. From what I've seen I agree with them.

[Skipjack]

LOS 2
LEO 10
MEO 100
GEO 200-300

Your numbers match what I had calculated for myself as well. In contrast to you, I do not find them too problematic, however.
First of all, you are assuming closeup dogfights in a time when most air to air combat is done with rockets over long distances. Whoever sees the enemy first on the radar screen wins (disregarding counter measures for a second). Second, you are thinking of using a UAV like a manned plane. I am sure that normally you would not do that. For a manned fighter jet, it is about survival most of all. For a UAV you calculate with loosing some and simply adapt your strategy accordingly. In fact you will probably even plan on loosing some, sacrificing them like pawns.
At half the price of a F35, you can accept loosing half your UAVs in combat from a purely economic POV and you dont have to worry about loosing any pilots or having to recover them from behind enemy lines, etc.
All this can improve morale quite a bit, which has its benefits as well (remember that Vietnam was lost because morale was down and not because of any other reason).
Plus, you can still have manned planes for supporting the UAVs. Nobody said that you cant do that.
In any case, this is a trend that I predict to go on and that I regard as a positive development.

[hanelyp]

When OTH control latency > 40 msec, direct control of a UAV is high risk (except through an autopilot)

And that is for a slow turning ISR/ground-attack type of UAV, not the >10g type needed to defeat an F-22 class manned fighter.

A commercial fly-by-wire flight control system has control loops on the order of 100 Hz (10 msec frame time). A no-compromise, 9-g manned fighter has a higher frame rate, but that number is probably classified, so let's assume 200 Hz (5 msec) for its FCS. The dogfighting UAV is smaller, with higher dynamic rates than the manned fighter, so let's assume its FCS runs at 250 Hz (4 msec). Under extreme maneuvers a loss of even one FCS frame is usually considered unacceptable, but lets be generous and assume your killer UAV can tolerate a one-frame loss while pulling high g and high rates in battle. To be extra nice to you, we'll also ignore latencies introduced by encryption, background noise, hardware delays, counter-jamming algorithms, etc.

So any latency over 4 msec means your UAV from Hell is toast.

A pilot has reflexes far slower than that. So the control loop described would not include the pilot, and need not contend with communications link delay.

[DeltaV]

You are arguing for man in the loop with the latencies.

No, I'm arguing for man in the loop with minimal latencies, the kind you get when the pilot and the aircraft are co-located.

The trend and current practice is for autonomy. Mouse click vice joystick.

For observation and attack of fixed or slowly moving surface targets. What's that got to do with air combat?

We have fielded systems that do most of the grunt work without man in the loop. So the delays are not that important.

Again, for fixed or slowly moving surface targets. Dynamic air combat is extremely more demanding.

To paraphrase a flag officer, "when is the last time we put a man inside an ICBM, SLBM or Tomahawk?".

When is the last time an ICBM, SLBM or Tomahawk was launched against against anything other than a fixed or slowly moving surface target?

The next gen AF long range bomber is fragged to optional manned.

Another surface attack system. What bearing does that have on air combat?

[DeltaV]

First of all, you are assuming closeup dogfights in a time when most air to air combat is done with rockets over long distances. Whoever sees the enemy first on the radar screen wins (disregarding counter measures for a second).

Well then, by your own flawed logic, the victory will go to the platform with the most surface area for sensors and the most internal volume for missile stowage.
Manned fighters, in other words.

Second, you are thinking of using a UAV like a manned plane. I am sure that normally you would not do that. For a manned fighter jet, it is about survival most of all. For a UAV you calculate with loosing some and simply adapt your strategy accordingly. In fact you will probably even plan on loosing some, sacrificing them like pawns.

Oh, the UAVs will be sacrificed, alright, once the DEW technology is in place: viewtopic.php?p=34763&highlight=#34763

In any event, my gripe is about your insinuation that F-22s will be made obsolete by UAVs in the near term.

[DeltaV]

A pilot has reflexes far slower than that. So the control loop described would not include the pilot, and need not contend with communications link delay.

Yes, far slower:
Human eye flicker detection 33 msec
Computer screen refresh 13 msec
Keyboard buffer 10-20 msec
Typical human reaction 150-250 msec

But the issue is pilot induced oscillation (PIO). The likelihood of PIO gets worse with greater latency. A human pilot is a very sophisticated, predictive controller, which no onboard AI has been able to match yet in the air combat realm. If the onboard AI is not good enough (and it's not, at this time) that leaves only a data link to a human pilot.

A typical fighter's fly-by-wire system makes hundred(s) of corrections per second just to keep the airplane stable, with or without pilot inputs. There is a tradeoff between stability and response time. A fighter requires the fastest possible response time. The cost for maneuverability is a reduction in inherent stability, hence the need for fast control loops.

The 4 msec I stated above is more applicable to the onboard stabilization loops than to PIO, although I suspect the PIO latency threshold for a remote-controlled UAV dogfighter would be far less than the 40 msec PIO threshold for a typical, non-dogfighting UAV.

[Skipjack]

Well then, by your own flawed logic, the victory will go to the platform with the most surface area for sensors and the most internal volume for missile stowage.
Manned fighters, in other words.

Why would that have to be manned fighters? Where is the logic in that.
Plus you can send a UAV ahead as a pawn to scout the area and relay targetting information to other UAVs or manned fighters following behind. Lots of options there, especially if you do not have to worry about the life of a pilot.

Oh, the UAVs will be sacrificed, alright, once the DEW technology is in place: viewtopic.php?p=34763&highlight=#34763

Not quite sure what that has to do with anything. I would guess that if it is effective against a UAV, it would also be effective against a manned fighter, which would be quite a tragedy then, no?

In any event, my gripe is about your insinuation that F-22s will be made obsolete by UAVs in the near term.

I guess we both understand something different, when we say near term. To me near term is by the time when the Chinese will have a large amount of fighter jets equivalent to the F22, as you proposed they would have. I think that this is still far enough out for UAVs to become such a major part in operations that you will be able to have a smaller amount of F22s, F35s and other fighters and still outmatch the Chinese due to having a larger amount of UAVs for support. This is what I meant.
And I am still standing by my assessment that UAVs are the future of all types of air combat, air to air, air to ground and recon.
For the types of wars that the US is fighting RIGHT now, they are actually superior already. And that is IMHO quite important, because these are not hypothetical wars against hypothetical enemies (that are IMHO never going to happen), but wars in which american soldiers are dying right at this very moment.

[ladajo]

You are arguing for man in the loop with the latencies.

No, I'm arguing for man in the loop with minimal latencies, the kind you get when the pilot and the aircraft are co-located.

The trend and current practice is for autonomy. Mouse click vice joystick.

For observation and attack of fixed or slowly moving surface targets. What's that got to do with air combat?

We have fielded systems that do most of the grunt work without man in the loop. So the delays are not that important.

Again, for fixed or slowly moving surface targets. Dynamic air combat is extremely more demanding.

To paraphrase a flag officer, "when is the last time we put a man inside an ICBM, SLBM or Tomahawk?".

When is the last time an ICBM, SLBM or Tomahawk was launched against against anything other than a fixed or slowly moving surface target?

The next gen AF long range bomber is fragged to optional manned.

Another surface attack system. What bearing does that have on air combat?

The problem on the table is Contested Airspace. That includes ground and air threats. Thus includes air combat considerations. At its far right end, air combat is dogfighting. Most engagements now are about missile shots. Current unmanned tech can support air combat up to missile shots. We are not far off from dogfighting. One can argue that missiles themselves are unmanned vehicles engaged in dogfighting during the terminal phase.

[ladajo]

PM me if you want to see the pdf I wanted you to read.

[Skipjack]

The problem on the table is Contested Airspace. That includes ground and air threats. Thus includes air combat considerations. At its far right end, air combat is dogfighting. Most engagements now are about missile shots. Current unmanned tech can support air combat up to missile shots. We are not far off from dogfighting. One can argue that missiles themselves are unmanned vehicles engaged in dogfighting during the terminal phase.

Quoted for agreement
Wow, me and ladajo agree on something. Gotta mark that day in my calender