AUSA 2021: US Army's ‘Scarlet Dragon' project aims to use AI, satellites for targeting

The US Army is running a series of training and experimentation events for using artificial intelligence (AI) to identify targets, in an effort to drastically reduce so-called sensor-to-shooter times.

The ‘Scarlet Dragon' series is designed to provide AI-augmented decision-making in targeting for large-scale combat operations, Colonel Joseph Buccino, XVIII Airborne Corps public affairs director, told reporters at the annual Association of the United States Army (AUSA) conference in Washington, DC.

XVIII Airborne is a rapid reaction force based at Fort Bragg, North Carolina, and is leading the exercises. “We think the immediate response force mission makes us the ideal host” for this sort of experimentation and innovation, Col Buccino said.

The driving technology behind the Scarlet Dragon series is software created under the US Department of Defense's Project Maven, which was originally designed to use AI for selecting targets from aerial video feeds.

Scarlet Dragon, however, aims to use AI for targeting based on data from satellites, such as electro-optic and synthetic aperture radar sensors, Colonel Melissa Solsbury, Project Ridgway director for XVIII Airborne Corps, told reporters at the event.

Scarlet Dragon IV ran during the week of 4 October, and was the fourth iteration of the series, which began last year. It is intended to inject new partners and concepts every 90 days, so the army can better understand how to better leverage data in warfare, Col Solsbury said. “We're already starting to think about the next … exercise in the February or March timeframe.”

Officials declined to describe how data is being transmitted from satellites to the ground, citing security reasons.