Skunk Works Declares Success Using AI for Air-to-Air Combat | Aero-News Network
Aero-News Network
RSS icon RSS feed
podcast icon MP3 podcast
Subscribe Aero-News e-mail Newsletter Subscribe

Airborne Unlimited -- Most Recent Daily Episodes

Episode Date

Airborne-Monday

Airborne-Tuesday

Airborne-Wednesday Airborne-Thursday

Airborne-Friday

Airborne On YouTube

Airborne-Unlimited-11.10.25

AirborneNextGen-
11.11.25

Airborne-Unlimited-11.05.25

Airborne-Unlimited-11.06.25

AirborneUnlimited-11.07.25

LIVE MOSAIC Town Hall (Archived): www.airborne-live.net

Sat, Jun 29, 2024

Skunk Works Declares Success Using AI for Air-to-Air Combat

AI Wingmen Still a Ways Off, but the Future is Closer Than Ever

Lockheed Martin's Skunk Works gamed out the effectiveness of Artificial Intelligence along with the University of Iowa's Operator Performance Laboratory using one of their L-29 Delfin jets, to reported success.

The test wasn't too intense, though, with only half of the combatants being a real aircraft from the start. That being said, a full-scale, live jet is a fine starting point, and a higher level of fidelity than similar tests often enjoy. The Delfin in question was given simulated-to-real transfer test objectives against a virtual adversary across defensive and offensive risk postures.

All in all, the team tried out 8 different test cases, to see how the AI agent would handle a variety of situations. That included head-to-head fights, off-aspect encounters, missile support for friendlies, and missile defeat scenarios against enemy aircraft.

"This was the first live exercise of the new flight interface; it's thrilling to see the separate components successfully integrate on the L-29 to demonstrate new capabilities. The complete system performed even better in live flight than in simulation," said Dr. Tom "Mach" Schnell, OPL professor at Iowa Technology Institute.

"Live flight tests are a crucial aspect of advancing our expertise in AI and autonomy. These flights are powerful demonstrations of our ability to quickly and affordably develop and test operationally relevant AI capabilities," said Matthew "Gabe" Beard, Lockheed Martin Skunk Works autonomy and machine learning engineering manager.

There's still plenty more to do, but the team feels they're on track to figure out the best way forward in no time.

FMI: www.lockheedmartin.com

Advertisement

More News

NTSB Final Report: Piper PA-44-180

While On The Base Leg Of The Airport Traffic Pattern The Right Main Landing Gear Did Not Fully Extend Analysis: Both pilots reported that after performing airwork they returned to >[...]

Classic Aero-TV: The Bizarre Universe of Klyde Morris Cartoons

From 2023 (YouTube Edition): Putting the ANT in Antihero A Beech Starship speeds along at altitude. “Deflectors on!” a voice from within the aircraft cries. “Look>[...]

ANN's Daily Aero-Term (11.09.25): Minimum Friction Level

Minimum Friction Level The friction level specified in AC 150/5320-12, Measurement, Construction, and Maintenance of Skid Resistant Airport Pavement Surfaces, that represents the m>[...]

Aero-News: Quote of the Day (11.09.25)

“Beginning this aircraft subsystem testing is the culmination of more than a decade of focused engineering and certification refinements. This is the moment where our intende>[...]

Classic Aero-TV: Falling for Para-Phernalia’s Softie Emergency Parachutes

From 2023 (YouTube Edition): The Best Option for A Pilots’ Worst Days Since its 1979 founding, Para-Phernalia, Inc. has designed and manufactured the Softie line of pilot eme>[...]

blog comments powered by Disqus



Advertisement

Advertisement

Podcasts

Advertisement

© 2007 - 2025 Web Development & Design by Pauli Systems, LC