Twelve to fifteen at a time
In the second week of March, security forces at Barksdale Air Force Base watched multiple waves of 12 to 15 drones cross sensitive areas of the base, including the flight line. Barksdale is home to B-52s. An internal briefing described non-commercial signal characteristics, long-range control links, and resistance to jamming. The drones came and went in patterns that suggested the operators were avoiding being located and testing how the base would respond.
Around the same time, Northern Command had one counter-drone fly-away kit for the entire homeland. Its commander said about a quarter of detected drones were being defeated, up from almost none a year earlier. Those two facts explain most of what has happened since.
The money arrived this month
On July 1, Joint Interagency Task Force 401 put a three-year, $500 million vehicle in place for its Domestic Shield program, and the first task order, worth $80.5 million, buys multi-sensor detection systems, electro-optical and infrared payloads, and counter-UAS radars for Air Force Global Strike Command sites. In May the task force picked five installations for a directed-energy pilot: Fort Huachuca, Fort Bliss, Naval Base Kitsap, Grand Forks, and Whiteman.
This is the right order of operations. You cannot defeat what you cannot see, and bases have been short on sensors for years. The task force's own counter-drone guide leans on passive detection: radar, EO/IR, and RF sensing that reads signal characteristics without reading content.
Sensors multiply faster than pictures
The trouble starts when the second vendor shows up. A base that buys radar from one company, RF detection from another, and a laser from a third ends up with three screens and three ideas of where the drone is. The watch officer becomes the integration layer, correlating tracks by eye and calling them out over a radio.
Multiply that by every installation in a region. A drone that crosses from one base's airspace into its neighbor's starts over as a new detection. Nobody can see whether the same operator is probing five sites in one night, which is exactly the pattern Barksdale's briefing was worried about.
What the picture layer has to do
EdgeLance treats every sensor as a source of typed observations: a radar return, an RF hit, a camera detection, an ADS-B report. Our multi-domain tracker fuses them into one set of tracks, refuses reports it cannot verify, and keeps contested identities visible instead of picking a winner. Any vendor's sensor that can emit an observation joins the same picture, which is how we built the sensor fabric in the first place.
The picture runs where the defenders are. A forward node at the base keeps tracking and keeps its delegated authority when the link to the regional operations center is cut, and it syncs back when the link returns. Every track, decision, and engagement lands in the mission record, so the review after a probing night can answer who saw what, when, and what was done about it.
The counter-UAS and base defense mission page lays out how this runs at a single site and across several.
Buy the sensors, then plan the layer
Domestic Shield is spending money on the right problem. The next purchase should be the layer that turns a base's sensors into one picture and a region's bases into one network. Without it, each new sensor adds another screen to the watch floor, and cheap drones keep winning the cost math because the defender spends minutes deciding what an attacker decided in seconds.