Drone Detection Gets a 5G Boost: Lockheed Martin’s NetSense Turns Cell Towers into Surveillance Sensors
Lockheed Martin’s NetSense Technology
According to НВ — Техно: With small drones becoming a growing concern around airports and other sensitive sites, affordable early detection is a pressing need. Lockheed Martin and its partners have completed initial field evaluations of NetSense, a new system that repurposes standard commercial 5G towers to monitor airspace. The tests took place near Miami, where the system detected a drone and followed its path in real time. By picking up subtle radio-frequency disturbances caused by a drone crossing a mobile network’s signal, the system uses artificial intelligence to identify these micro-interference patterns and flag the aircraft.
Partner Contributions
Several companies were involved in developing the technology:
- Verizon supplied the network spectrum;
- NVIDIA built the AI Aerial platform used to analyze signals;
- ODC handled the radio access software;
- Keysight Technologies was responsible for simulating the environment.
Lockheed Martin consolidated this input into a single data stream. Notably, the solution requires no physical changes to mobile towers, which makes it fast and cost-effective to deploy.
The first pilot projects are scheduled for the second half of 2026 and early 2027. The system’s rapid deployment capabilities make it suitable for protecting airports, power plants, stadiums, and government facilities. Its open architecture can also be adapted for 6G standards or integrated with additional interception tools. This is based on a report by Interesting Engineering.
NetSense could reshape airspace security by enabling faster detection of drones and other aerial objects. It holds promise for critical sectors where airspace management is essential. Because it relies on existing 5G infrastructure, adopting the technology is less expensive and quicker than building new detection networks, making it appealing to both government and commercial buyers.
Read also

