DARPA looks to video games to develop warfighting strategies

DARPA looks to video games to develop warfighting strategies

Technology News |
The U.S. Defense Advanced Research Projects Agency (DARPA) has launched a program that seeks to develop and apply AI to existing commercial open-world video games - such as StarCraft II and FreeCiv - in an effort to develop "winning warfighting strategies."
By Ally Winning

Share:

Such games are designed to be balanced, allowing players of equal skill to have equal chances at winning and results. In the agency’s Gamebreaker Artificial Intelligence (AI) Exploration program, teams would test AI methods to quantitatively assess the game balance, identify underlying parameters that significantly contribute to balance, and explore new capabilities, tactics, and rule modifications that are most destabilizing to the game.

These could include the introduction of a new weapon, capability, or rule change, and the AI system must be able to predict the outcome of the change. In future conflicts, says the agency, Department of Defense (DoD) investment is designed to maximize imbalance to create an advantage or to seek equilibrium when an adversary is seeking an advantage.

AI algorithms inspired by Gamebreaker could, says the agency, help develop winning warfighting strategies when the adversary’s objectives – i.e. the ‘rules of the game’ – are not clearly known. By exploiting game balance, Gamebreaker is designed to address an existing gap in AI and data analytics research as applied to current wargaming and simulation.

“If we can figure out a generic method to assess and then manipulate balance in commercial video games,” says Lt. Col. Dan ‘Animal’ Javorsek, the Gamebreaker program manager in DARPA’s Strategic Technology Office, “my hope is that we might then apply those AI algorithms to create imbalance in DoD simulated war games used to train warfighters for real-world battle.”

A number of advanced commercial video games, says the agency, involve realistic command and control, campaign planning, and strategy development, which are very relevant to the military. Games of strategy like chess or weiqi – the Chinese name for Go – often related to military strategy also serve as the developmental playground for AI research. However, after conquering such sequential perfect information games, AI has moved into the much more dynamic environment of real-time strategy (RTS) video games.

Rather than starting from scratch, says the agency, Gamebreaker aims to leverage the gaming industry’s significant AI development in these areas and build on them for purposes unique to the DoD. The teams in the program each selected two games that they will try to break, with the AI used to break the first game to be then tested on a second game.

The two-game approach is designed to validate that an AI methodology can be extended beyond a single video game and could potentially be applied to advanced DoD wargaming scenarios. The Gamebreaker teams and their two video games include the following:

  • Aurora Flight Sciences is teamed with MIT and will use StarCraft II and Google Research Football
  • BAE is teamed with UC Santa Barbara and AIMdyn and will use StarCraft II and the AFRL Strategem Wargame
  • Blue Wave AI Labs will use SpringRTS: 1944 and OpenRA
  • EpiSci will use miniRTS and StarCraft II
  • Heron Systems will use DeepRTS and StarCraft II
  • Lockheed-Martin is teamed with Cycorp and will use Multi-agent Particle Environment and SpringRTS: 1944
  • Northrop-Grumman has teamed with Hazard Software and Matrix Games and will use Command: Modern Operations and TORCS
  • Purdue University will use microRTS and StarCraft II
  • Radiance Technologies is partnered with BreakAway Games and will use FreeCiv and Zero-K

“We received many incredibly rich proposals from a diverse base of universities and small and large companies,” says Javorsek. “Since the teams aren’t competing against each other, we’re looking forward to a collaborative effort as performers develop their AIs across an array of games.”

At the conclusion of the first phase of the program early next year, the teams will demonstrate their methods of empirically assessing game balance in the first of their two games.

DARPA

Related articles:
Self-taught Google AI achieves ‘superhuman’ chess proficiency in hours
Google’s Kaggle opens AI attack/defense competition
Citizen science effort recruits gamers in fight against COVID-19
NASA drone race pits human pro vs AI
Autonomous AI racing drone designed to defeat humans

Linked Articles
eeNews Embedded
10s