WASHINGTON \u2014 Northrop Grumman and three artificial intelligence firms \u2014 Shield AI, Accelint and Applied Intuition \u2014 showcased how different AIs could swap control of a single aircraft \u201cseamlessly\u201d mid-flight in recent testing, the companies said, which could offer US forces unprecedented flexibility in future fights.<\/p>\n
The flight tests \u2014 one last month<\/a> involving Shield, the latest Wednesday<\/a> with Accelint and Applied \u2014 were part of a Northrop initiative called Talon IQ (formerly Beacon<\/a>),\u00a0which turned a manned demonstrator, Scaled Composite\u2019s Vanguard Model 437, into a testbed for both Northrop\u2019s own Prism autonomy system and AI software from a growing group of partner companies<\/a>.<\/p>\n \u201cWe just completed our eighth flight test of the Talon IQ platform,\u201d said Dan Salluce<\/a>, Northrop\u2019s senior director for aerospace systems. \u201cWhile the aircraft was flying, the software was queued up so that we could have different companies\u2019 behaviors take control of the platform and fly [it].\u201d<\/p>\n In essence, the AIs took turns controlling the aircraft, the companies said. What made that possible is a layered and modular open architecture<\/a> that lets the airplane plug-and-play different software programs \u2014 either specialist AIs for specific tasks or generalist ones to run an entire mission \u2014 without disrupting the microsecond-by-microsecond operations of the flight controls that keep the airplane from falling out of the sky.<\/p>\n \u201cThe mission autonomy is really about what do you do with that airplane to perform the mission,\u201d Salluce explained. \u201cWhere do you go? How fast do you fly? What direction? What altitude?\u201d<\/p>\n In this week\u2019s flight test, Northrop\u2019s Prism handled the overall mission, but it handed control to Applied Intuition\u2019s Acuity AI at one point, and to Accelint\u2019s AI at another, for them to execute specific functions or \u201cskills\u201d such as performing a Combat Air Patrol<\/a>.<\/p>\n In the March test flight<\/a>, by contrast, once the aircraft was aloft, Prism handed full control to Shield\u2019s Hivemind AI<\/a>. Shield\u2019s AI then put the aircraft through some standard military maneuvers \u2014 \u201cCombat Air Patrol maneuvers, then doing simulated target engagement maneuvers,\u201d explained Shield\u2019s Vice President for Hivemind, Todd Wesley<\/a> \u2014 before returning control to Prism.<\/p>\n Because the baseline software handles the flight controls, the higher-level mission-autonomy AIs don\u2019t have to be tailored to the specific characteristics of a given aircraft, the executives said: Instead, you can test AI on one plane and then port it over to another, much like an experienced pilot can fly several kinds of (related) aircraft.<\/p>\n