A team at the Swedish Defense Research Agency has developed WiFireNet, an AI system that spots concealed handguns by analyzing how they distort ambient Wi-Fi signals. The system uses a Wi-Fi transmitter and receiver to capture channel state information, which reflects an object's shape, orientation, and material.
To train the model, the researchers recorded people walking with a real metal handgun, a plastic handgun, a metal water bottle, and a phone, matching video frames with Wi-Fi signatures. In controlled trials with up to four moving people in a small indoor space, WiFireNet achieved an average handgun-detection precision of 72.5%. When the system also had to pinpoint the weapon's exact location on the carrier, precision fell to 53.8%.
The system is still experimental. It was built on a modified version of the 1999 IEEE 802.11a standard, and it struggles in unfamiliar indoor layouts. The team plans to move to modern Wi-Fi 7 and Wi-Fi 8 networks, which offer higher-resolution channel state information for better detection.
The researchers have released the code, models, and dataset as open source. While Wi-Fi sensing is gaining attention as a low-cost surveillance method, critics point out that its ubiquity and lack of regulatory frameworks raise serious privacy concerns.