Air Force asks industry for artificial intelligence (AI) and machine learning for aerial battle planning

Nov. 3, 2021
Stratagem seeks new AI capabilities that can reason in real-time about the battlespace, and assist planners responsible for reacting.

ROME, N.Y. – U.S. Air Force researchers are asking for industry's help in developing artificial intelligence (AI) and machine learning for planning systems that help commanders adapt quickly to changing conditions in high-intensity aerial battles.

Officials of the Air Force Research Laboratory's Rome Research Site in Rome, N.Y., issued a broad agency announcement (FA875020S7007) last week for the Stratagem: Applying State-of-the-Art Artificial Intelligence and Machine Learning Approaches to Air Battle Management project.

Stratagem seeks new AI capabilities that can reason in real-time about developments in the battlespace during wartime engagements and assist with battle planning and decision making in reacting to those developments.

As the Air Force begins to operate in contested environments against peer and near-peer adversaries, the demands on planners and warfighters will increase quickly enough to require AI and machine learning decision support, researchers say.

Related: Air Force considers gaming-developed artificial intelligence (AI) and machine learning for battle management

One source of this kind of technology is advanced video gaming. In recent years, game designers had made major developments in AI for non-human game-playing agents; this technology might be considered for battlespace decision support.

The Stratagem project has three parts: machine intelligence support for complex military operations; capturing human expertise to augment warfighter capability; and transferring machine-learning models from video gaming to Air Force battle management and simulations.

Only one part, machine intelligence for strategy development, is part of this specific solicitation. This part seeks to explore new and existing algorithms and learning methods that involve a model architecture for an intelligent agent, determining which pieces that absolutely must be included in the design of a reasoning agent in complex environments.

Related: Artificial intelligence and machine learning for unmanned vehicles

This model should be based computational game theory and machine learning. The goal is to develop an approach to match and then exceed expert human gameplay in advanced wargames.

It seeks to produce blue/red agents that can assist and play against planners in real-time and run many games to produce plan options for operators.

Air Force researchers are considering wargames with an operational area of as little as 50 by 50 miles, and as large as 500 by 500 miles, that last for as long as 100 days, at the squadron, brigade, battalion, and naval ship level of granularity. Games will include terrain, cities, and weather.

Related: Raytheon to develop large-scale battle management software to enable military planning at the theater level

Companies interested should email white papers no later than 31 March 2022 to the Air Force's Nathaniel Gemelli at [email protected], with copies to [email protected].

Those who send promising white papers will be invited to submit formal proposals. The project stays open through 29 Sept. 2023. Email technical questions or concerns to Nathaniel Gemelli at [email protected], and contracting questions to Amber Buckley at [email protected].

More information is online at

About the Author

John Keller | Editor-in-Chief

John Keller is the Editor-in-Chief, Military & Aerospace Electronics Magazine--provides extensive coverage and analysis of enabling electronics and optoelectronic technologies in military, space and commercial aviation applications. John has been a member of the Military & Aerospace Electronics staff since 1989 and chief editor since 1995.

Voice your opinion!

To join the conversation, and become an exclusive member of Military Aerospace, create an account today!