This is the Bolt-M, the autonomous drone with AI presented by the United States Army

Since 2015, when Russia and Ukraine began to demonstrate that they were an unbalancing force, small drones have demonstrated their capabilities in different scenarios. And the latest addition is an attack and intelligence quadcopter that its manufacturer hopes will do more things with much less operator attention. This is the Bolt-M, presented today by the company Anduril and its objective is to demand less from the operator and offer more information than first-person view (FPV) attack drones.

This is not the first time that the US Army has considered using FPV drones for infantry platoons, but until now they faced the problem that they needed special training to use and had many operational limits. The Bolt-M, on the other hand, works “without requiring specialized operators,” according to a statement from Anduril. The company has a contract with the Organic Precision Fires – Light program (Organic Precision Fires – Light) of the US Marine Corps, or OPF-L, to develop an attack variant.

The key feature of Bolt-M is its autonomy and artificial intelligence software powered by the Lattice platform by Anduril. The operator can draw a bounding box on a battlefield screen, set some specifications, and send the drone on its way.

“Once a target is identified in Lattice, an operator can specify a standoff position for Bolt-M to maintain, tasking the system with Stalk the target from beyond the visual or acoustic detection range even when the target is moving and hidden – explains the statement -. When it is time to attack, an operator can define the angle of attack to ensure the most effective mode, while onboard vision and guidance algorithms maintain terminal guidance even if connectivity to the operator is lost.”

But the system is also intended to handle some reconnaissance tasks that humans do, but other small, inexpensive attack drones do not. “The Bolt-M aims to help its operator understand what is happening on the battlefield – notes Anduril’s director of strategy, Chris Brose -, whether on targets known or recognizable to onboard systems, or if they are unknown objects that that operator can select through their interaction with the autonomous systemtell it to track, tell it to follow, eventually, if you want, based on the human saying ‘Go’, to fly up and attack that target.”

According to Brose the Bolt-M could even detect new variants of older weapons. “If the Russians in this case start modifying them and building these types of turtle tanks, maybe computer vision hasn’t seen it already… it can still show that information to an operator,” adds Brose.

Over the next six months, the Marine Corps will put the Bolt-M ammunition variant through “a fairly rigorous testing and evaluation campaign,” Brose concludes. The drone can fly towards GPS waypoints, but In places where GPS is under attack, operators can control it manuallyand can maintain custody of the target and execute commands previously delivered by the operator, even when connections are interrupted.

Finally, another advantage of the system that Bolt-M uses is that the Lattice platform can integrate data from various sensors and sources and the goal is for it to work with a variety of drones, even from other manufacturers.