AT&T and Ericsson used three ordinary commercial 5G cell sites to detect and track drones flying near AT&T Stadium in Arlington, Texas, during the World Cup, with no radar and no new hardware anywhere in the system. The July 10 demonstration turned the network itself into the sensor. I have covered cellular drone detection at DroneXL since the first European experiments, and this is the technology’s loudest American arrival yet.
Three Cell Towers Watched The Sky Over A World Cup Venue
The demo ran in authorized airspace outside one of the tournament’s marquee venues, which was not a coincidence. Stadiums spent this summer at the center of the largest drone enforcement operation in US history, and the FBI’s 600-plus World Cup seizures made airspace awareness a front-page problem.
Two of the three participating cell sites sat roughly a quarter mile (0.4 km) from the demonstration zone. Working together, they detected multiple drones at 300 to 400 feet, located them, and tracked their movement as the aircraft approached the stadium area.
The range claims go further. The companies say the approach can reach out to about 3.7 miles (6 km), resolving objects into boxes roughly 4 inches (10 cm) on a side, all using mid-band 5G spectrum chosen as the compromise between resolution and reach.
Software, Not Hardware, Does The Watching
The system works passively, which is the detail that separates it from most counter-drone gear. The drone does not need to be connected to the network, broadcasting Remote ID, or cooperating in any way. Standard 5G transmissions bounce off the airframe, and Ericsson’s software fuses the reflections from multiple towers in what engineers call a multi-static configuration.
That matters because the non-cooperative aircraft is precisely the one security teams lose sleep over. Remote ID only identifies pilots who follow the rules. A sensing network that sees the airframe itself, transponder or not, closes the gap that compliance-based systems were never going to cover.
The underlying capability, known as gNB sensing, was standardized in 3GPP Release 19 under the banner of Integrated Sensing and Communication. The industry expected this class of feature to wait for 6G. It is arriving early, by software update, on infrastructure that already exists.
That economics argument is the entire pitch. Dell’Oro analyst Stefan Pongratz put it in market terms: “The ability to leverage the existing macro grid and minimize the incremental capex completely changes the risk/reward profile.” AT&T described its own investment as fairly modest, concentrated in a small technology group in Austin.
Europe got here first, a pattern DroneXL has tracked all year. Lockheed’s NetSense wants to turn 5G signals into a drone detector, and Rheinmetall and Deutsche Telekom are building a German drone shield on cellular detection. AT&T and Ericsson just brought the concept home to American spectrum.
The Fine Print Is Where The Birds Live
The companies were candid about what the demo did not prove. Neither offered a clear explanation of how the AI distinguishes a drone from a bird, published false positive rates, or shared performance data in bad weather. Dense urban cores present multipath problems, and fast movers are harder to hold.
The system also stops at detection. Under US counter-UAS law, actually doing something about a hostile drone, whether jamming, capturing or grounding it, stays reserved for a short list of authorized federal entities. The tower can point. It cannot shoot.
Commercialization is a slow road by AT&T’s own admission. Robert Soni of AT&T described the roadmap in three stages: “We’ll offer a skateboard version of this, then we’ll offer a bicycle version of this, and then we’ll offer an automobile version.” The skateboard, aimed at federal agencies and first responders, is a three-to-five-year project, with government customers already funding research.
Both companies are also opening APIs so third parties can fuse the tower data with camera feeds, radar, flight manifests and other systems, positioning the network as one layer in a larger airspace picture rather than a standalone product.
Ericsson, for its part, elevated the demo to its Q2 earnings release, the first time network sensing graduated from research footnote to investor material.
DroneXL’s Take
Here’s what nobody is saying out loud, your phone company just became a counter-drone company, and it did not have to buy a single new sensor. The towers were already standing. You paid for them.
This is the third time this year I have watched this exact movie. Lockheed wants your 5G signal to snitch on drones. Deutsche Telekom and Rheinmetall are assembling a national drone shield out of Germany’s cell grid. Now AT&T and Ericsson are teaching the American network the same trick, at a World Cup venue, in the same summer the FBI filled warehouses with seized quadcopters. When the phone bill people and the defense people start finishing each other’s sentences, you know where the counter-drone money is headed: not into new hardware, into infrastructure that was already paid for twice.
The unresolved half is the one the press releases skip. A tower can tell you something is there. It cannot yet swear that something is a drone instead of a bird, and until somebody publishes false positive rates, every security director should hold their applause. Detection was always the easy 90 percent.
Watch the skateboard. If federal agencies turn this into a funded program within Soni’s own three-to-five-year window, the counter-drone sensor market gets a competitor whose tower grid is already standing in every city, suburb and highway corridor in America. Nobody in history has been this excited about coverage bars.
