Four robotic combat vehicles hunted targets together at Fort Knox without a single hand on the wheel, and the Pentagon says that’s just the beginning.
Quick Take
- Four autonomous ground vehicles built under a DARPA program worked as a team at Fort Knox’s St. Vith Range on October 15, 2026.
- DARPA says the vehicles moved, searched, and found targets on their own, with no human steering them.
- Humans still had to approve any strike — the robots don’t “pull the trigger” without a person’s order.
- The test used simulated targets on a controlled training range, not a real battlefield.
- U.S. European Command is reportedly looking at using similar robots ahead of American troops overseas.
Robots Hunt Targets Together At Kentucky Training Ground
Four unmanned ground vehicles built under the Defense Advanced Research Projects Agency’s RACER program rolled across Fort Knox’s St. Vith Range on October 15, 2026. They worked as a team, scanning rough terrain for enemy positions. The Army says the machines operated “independently operating together,” hunting for simulated threats without a soldier steering them from a control box.
DARPA technical adviser Dr. Jason Gregory made clear this wasn’t remote-control driving dressed up as innovation. “All the mobility, all the movement, all the detections, all of that is autonomous,” Gregory said. Operators gave the machines a location on the map. The vehicles then used their own sensors to read the land and drive there, he said, “comparable to a human.”
Team Tactics, Not Just Solo Driving
This wasn’t one robot wandering a field. The Army described a real tactical sequence: an operator used satellite imagery to task the team, one vehicle pushed forward and searched for a possible enemy position, and two more vehicles rolled up the road behind it to provide cover and support. That’s small-unit teamwork, run by machines instead of soldiers.
The vehicles carried different equipment packages, and the Army says the demonstration linked that autonomous movement and sensing directly to “kinetic and non-kinetic effects” against simulated targets spotted by both ground sensors and satellites. In plain terms, finding the enemy and striking the enemy were shown as one connected process, not two separate jobs.
Humans Still Hold The Trigger
For readers who value accountability and the chain of command, this matters most: a soldier still makes the call to fire. Gregory was blunt about where the autonomy stops. “They don’t pull the trigger until a human says pull the trigger,” he said. The system flagged targets for human review before any simulated engagement happened.
That distinction should matter to anyone worried about machines making life-and-death decisions without oversight. The Fort Knox test shows a machine that can move, look, and report on its own. It does not show a machine deciding who dies. Supporters of a strong, well-run military should see that as the system working as intended — technology serving the warfighter, not replacing his judgment.
A Program Two Decades In The Making
This demonstration didn’t come out of nowhere. DARPA has chased autonomous ground vehicles since its 2005 Grand Challenge, when five vehicles finished a 132-mile desert course with zero human intervention. RACER pushes that same idea into harder conditions: off-road terrain, poor maps, and spotty GPS, the kind of environment American troops actually fight in.
Stars and Stripes reported that U.S. European Command is exploring sending robots like these ahead of American soldiers overseas, scouting danger before troops ever reach it. That’s a straightforward, common-sense use of technology: let a machine walk into the minefield first. For a military stretched thin by years of budget fights and recruiting struggles, tools that protect troops without replacing their judgment deserve a hard look, not knee-jerk suspicion.
The Fort Knox test was a controlled exercise with simulated targets on a known training range, not combat. The Pentagon hasn’t released detailed numbers on how often the vehicles got it right, how they’d hold up against jamming, or how they stack up against a human-led team. Those are fair questions for taxpayers footing the bill, and for a Congress that should demand real answers before any wider rollout.
Sources:
dvidshub.net, army.mil, stripes.com
