“This event marked the first-ever kinetic drone swarm on U.S. soil,” a defense official said of a Camp Blanding, Florida, live-fire demonstration that placed multiple armed first-person-view drones under the direction of a single operator.
This is less significant from an accomplishment perspective than it is indicative of the future that tactical control has in store: from one person physically controlling one drone, to one person controlling multiple drones that can move, deconflict, and arrive on target all at the same time. During the Florida trial, Auterion and Kraken Kinetics showcased a one-to-many engagement, where one person controlled three autonomous strike drones that struck three different targets almost at the same time.
The tech solution was the Auterion stack: Skynode onboard computing and the Nemyx swarm engine. Rather than needing joystick control inputs all the way through every phase of flight, the operator set up intentions and approved engagement while the software controlled formation, collision avoidance, routing, and terminal guidance. This sets up an operator-to-effect ratio that is rapidly approaching that of fire direction rather than flight, which has dramatic implications for small units that don’t always have extra bodies to spare for one-shot rockets.
One of the aspects of the demo is the reason for the correlation between autonomy and the type of warhead. Each of the platforms had an explosively formed penetrator. EFPs are intended to create a high-speed metal projectile that has the ability to destroy armor, an ability that in the past led to decades of research and development of countermeasures to protect vehicles. Dr. Michael Knights explained the threat in general terms: “they have a very powerful, armor-piercing effect that can penetrate even the most modern main battle tanks.”
This combination swarm drones and terminal effects against armor is an evolution from swarm control as a novelty to the world of combined arms mathematics. An attack that is coordinated and multi-axis does not have to be numerically large to be considerably troublesome if it comes quickly, from multiple directions, and with sufficient accuracy to require a response from vehicles, crews, and air defenses at the same time. The message of the demonstration was not size it was concurrency. It was also a part of a larger effort by the U.S. government to institutionalize experimentation with AI-enabled systems.
The exercise was framed as an early look at Swarm Forge, which is an initiative of the Pentagon that aims to iterate on tactics and integration decisions for swarms of autonomous systems while also stress-testing how these systems will have to function in a noisy, contested, and rule of engagement-constrained environment. This is consistent with the guardrails that have been established within DoD Directive 3000.09, which mandates that autonomous and semi-autonomous weapon systems be designed to enable “appropriate levels of human judgment over the use of force.”
What gives Auterion’s U.S. range event more than a one-off is the feedback loop that the company says it has already established. Auterion says that it has already shipped more than 30,000 Skynode Strike kits to Ukraine, where small drones are being produced in large numbers and where GPS and control are often degraded by electronic warfare. In another report, Auterion’s SkyNode S is reported to use onboard processing to mitigate GPS and external communications, such as real-time image processing for target recognition and control.
From an engineering perspective, the Florida demo illustrates an emerging center of gravity: the “operating system” level that enables a mixed air vehicle fleet to be considered as a single strike platform. Hardware, warheads, and radios are all well and good, but the bottleneck in small drone warfare is increasingly the human bandwidth. Swarming architectures seek to solve this problem by making the human operator an authorizer and controller of intent, while the machines implement the geometry of arrival.
