The war in Ukraine is showing, in vivid and deadly detail, drones are now a permanent fixture on the battlefield. The United States military wants swarms of drones at its disposal in the event of future conflicts. But operating swarms of drones in the wars of tomorrow means building that capacity today.

Live from Devils Lake, North Dakota, it’s Autonomous #VBAT Teams! 
“Wait, you’re flying those three aircraft, doing the recon, and at the same time briefing us?!’

We recently came together with customers from across the #DoD enterprise to demonstrate real autonomy on real… pic.twitter.com/GoJbWSATas

— Shield AI (@shieldaitech) August 30, 2023

The defense technology company Shield AI is working toward that goal, and just completed a major step along the way. Shield AI touts its artificial intelligence agent, called Hivemind, as the best AI pilot in the world.

Up until now, Hivemind has only flown one aircraft at a time. However, Shield AI recognized if someone is creating something called Hivemind, it really needs a hive.

In August, the company announced the successful completion of what it called an “autonomous teaming demonstration.”

The Hivemind AI pilot launched three V-Bat craft and flew them in a scenario to monitor virtual wildfires. The V-Bat is a vertical takeoff and landing (VTOL), unmanned aerial system (UAS). The U.S. Navy and Marines currently use V-Bats for a variety of missions.

During the August test flight, the team of three autonomous craft, flown by a single AI pilot, conducted detect, identify, locate, and report (DILR) missions.

Shield AI worked with the Air Force Research Lab and AFWERX to field the drone teaming technology. Shield AI said in a release on its website the successful testing puts the company on a path to deploy V-Bat teaming capabilities in GPS-and-communications-denied environments in the next year.

US Air Force finding future airmen at FPV drone races

“Intelligent, affordable mass that can see everything on the battlefield, execute the mission even when GPS and comms are denied or degraded, and put all our adversaries’ military assets at risk at all times is the holy grail of deterrence. This milestone brings us closer to achieving that reality,” said Brandon Tseng, Shield AI’s president, co-founder and former U.S. Navy SEAL. “We had many customers from across the DOD enterprise attend the event and my favorite customer quote was ‘Wait, you’re flying those three aircraft, doing the recon, and at the same time briefing us?!’ The customers genuinely appreciated that this isn’t merely talk, or just computer simulations, or a science project leading nowhere. This represents real autonomy on actual aircraft that, most importantly, will be deployed imminently.”

The Hivemind is also platform agnostic, meaning it can be taught to fly a slew of aircraft in use by the Department of Defense, not just V-Bats.

Hivemind learned how to fly on small quadcopters, but it also successfully flew an F-16 for a few hours.

US military letting artificial intelligence ‘Hivemind’ fly fighter jet

Hivemind can be trained to carry out a variety of different missions like breaching an enemy’s air defense network, hunting for mobile missile launchers, countering enemy aircraft, or extending a fleet’s attack range.

When it comes to carrying out military missions, a human is still involved for the most critical decisions; like whether to engage with an enemy. Now, with its ability to fly in teams, Hivemind really is on another level compared to other AI pilots.

Production at our Batcave is not just up & running, the pedal is to the metal! The new Replicator initiative aims to deploy thousands of autonomous systems in the next 18-24 months. V-BAT, which is operating live missions under a number of different allied and international… pic.twitter.com/a146qqsxP1

— Shield AI (@shieldaitech) August 29, 2023

And that’s good news for the U.S. military, since it will be the biggest benefactor of Shield AI’s breakthrough. Current and potential conflicts, coupled with the rise of highly mobile cruise missiles, are forcing virtually every branch of the U.S. armed services to rethink how it would conduct a war.

In addition to spreading out lethal assets, the U.S. military also wants to add mass, lots of mass, and quickly. The only real way to do that is with unmanned systems. But one pilot operating one drone at a time isn’t conducive to this strategy, which makes the breakthrough in teaming capabilities Shield AI just had that much more important.

To beat China, US may need to win without runways