A.I. DOWNS EXPERT HUMAN FIGHTER PILOT IN DOGFIGHTS

[suzutsuki]

http://goo.gl/IjFpsS

 

In the military world, fighter pilots have long been described as the best of the best. As Tom Wolfe famously wrote, only those with the "right stuff" can handle the job. Now, it seems, the right stuff may no longer be the sole purview of human pilots.

 

A pilot A.I. developed by a doctoral graduate from the University of Cincinnati has shown that it can not only beat other A.I.s, but also a professional fighter pilot with decades of experience. In a series of flight combat simulations, the A.I. successfully evaded retired U.S. Air Force Colonel Gene "Geno" Lee, and shot him down every time. Lee called it "the most aggressive, responsive, dynamic and credible A.I. I've seen to date."

 

And "Geno" is no slouch. He's a former Air Force Battle Manager and adversary tactics instructor. He's controlled or flown in thousands of air-to-air intercepts as mission commander or pilot. In short, the guy knows what he's doing. Plus he's been fighting A.I. opponents in flight simulators for decades.

 

But he says this one is different. "I was surprised at how aware and reactive it was. It seemed to be aware of my intentions and reacting instantly to my changes in flight and my missile deployment. It knew how to defeat the shot I was taking. It moved instantly between defensive and offensive actions as needed."

 

The A.I., dubbed ALPHA, was developed by Psibernetix, a company founded by University of Cincinnati doctoral graduate Nick Ernest, in collaboration with the Air Force Research Laboratory. According to the developers, ALPHA was specifically designed for research purposes in simulated air-combat missions.

 

The secret to ALPHA's superhuman flying skills is a decision-making system called a genetic fuzzy tree, a subtype of fuzzy logic algorithms. The system approaches complex problems much like a human would, says Ernest, breaking the larger task into smaller subtasks, which include high-level tactics, firing, evasion, and defensiveness. By considering only the most relevant variables, it can make complex decisions with extreme speed. As a result, the A.I. can calculate the best maneuvers in a complex, dynamic environment, over 250 times faster than its human opponent can blink.

After hour-long combat missions against ALPHA, Lee says,"I go home feeling washed out. I'm tired, drained and mentally exhausted. This may be artificial intelligence, but it represents a real challenge."

 

The results of the dogfight simulations are published in the Journal of Defense Management.

[MasterZelgadis]

Implement in DCS please :D

[NineLine]

And this is how Skynet gets air superiority over us....

[Pikey]

Implement in DCS please :D

They used the P51 AI?

[Rangi]

I for one welcome our new Psibernetic overlords.

[kontiuka]

I for one welcome our new Psibernetic overlords.

They aint so smart. They can't even spell cybernetic right.:music_whistling:

[txmtb]

E.D.I is the future. :pilotfly:

http://th06.deviantart.net/fs70/PRE/f/2013/001/1/2/edi_and_talon_by_fighterman35-d5q10da.jpg

[Bushmanni]

It's an AI commanding a simulated red force in BVR fight and not dogfight AI. Regardless they have done good job. As the system runs on a regular 3.2 GHz desktop PC maybe some day we can have something like that in DCS.

[Tirak]

E.D.I is the future. :pilotfly:

http://th06.deviantart.net/fs70/PRE/f/2013/001/1/2/edi_and_talon_by_fighterman35-d5q10da.jpg

 

Oh please god no. The long hours in the processor mines i can handle, but not the shitty music, anything but that!

[Fishnuts]

The human pilot was not an actual fighter pilot, but just an Air Battle Manager who flew in the back of aircraft and gave callouts. This article is a bit misleading.

[OutOnTheOP]

The human pilot was not an actual fighter pilot, but just an Air Battle Manager who flew in the back of aircraft and gave callouts. This article is a bit misleading.

 

The AI was also given a 2:1 numerical advantage. Doesn't take a genius to win 4v2.

[aaron886]

The human pilot was not an actual fighter pilot, but just an Air Battle Manager who flew in the back of aircraft and gave callouts. This article is a bit misleading.

 

What's more, it doesn't describe the method of the test. I've seen real military aviators in desktop-type flight simulators struggling with the screen field of view and toy-ish flight controls. I've also watched DCS AI eliminate real military aviators... does that mean ED is creating SkyNet?!

[lunaticfringe]

Col. Lee's last assignment in the air 20 years ago was in E-3s. While he was doing instruction at the FWS in the 80s on Soviet tactics, anything he did in the cockpit prior to his controller position (and later wing command roles) is so atrophied in the way of a skillset to be invalid for the matter of debate.

 

Stick a current FWS instructor of the type of your choice F-22, F-15, F-16, F-35 in that simulator, or from a squadron that just ran through Red Flag, and watch that AI burn.

[yellow_jacket]

Actually it does say he was also a pilot at one time in the third paragraph. He was asked to drink a 72 ounce slurpy while flying though.

Edited June 29, 2016 by yellow_jacket

[Fishnuts]

Actually it does say he was also a pilot at one time in the third paragraph. He was asked to drink a 72 ounce slurpy while flying though.

 

I think that is an error on the author's part.

[captain_dalan]

It's an AI commanding a simulated red force in BVR fight and not dogfight AI. Regardless they have done good job. As the system runs on a regular 3.2 GHz desktop PC maybe some day we can have something like that in DCS.

 

That makes much more sense . I knew something was fishy, especially in the second part of the article when they stated the AI beat the human with handicapped missile performance or turning ability. That is especially dubious in a symmetrical ACM. Thanks for the clarification.

[SinusoidDelta]

The most interesting part is that "ALPHA" wasn't ran on a 3.2 Ghz desktop. It was ran on a $30 Raspberri Pi. http://magazine.uc.edu/editors_picks/recent_features/alpha.html