Fort Worth law enforcement has deployed artificial intelligence-driven audio sensors across the city to identify potential gunfire. The technology, which has been in operation for several years, allows the police department to receive location data for suspected shots, potentially before any resident makes a 911 call. The system relies on two distinct vendors: Flock Safety and Acoem.
The Flock Safety Raven system functions through a network of microphones spread across a specific area. When a sound occurs, the sensors analyze the noise to determine if it matches the acoustic signature of a firearm. If three separate microphones identify the same sound as gunfire, the city’s Real Time Crime Center receives an alert along with the calculated location. This system can also integrate with existing public safety cameras to provide visual context.
Acoem’s technology operates differently, using individual pods equipped with multiple sensors that can locate a sound source independently. Artificial intelligence helps these sensors distinguish between actual threats and false alarms. The system identifies both the initial blast from the weapon and the subsequent sound wave of a bullet traveling through the air. Each Acoem unit is directly linked to a camera, allowing officers to view the area where the sound was detected in real time.
Fort Worth police installed Flock Raven equipment in five areas chosen based on the city’s history of gun violence. These zones include Northwest 25th Street, Las Vegas Trail, Rosemont, areas near South Riverside Drive and Berry Street, and the intersection of Stalcup Road and Berry Street. Together, these locations cover approximately five square miles of the city.
Acoem equipment is not permanently fixed in these zones but is deployed at large-scale events or in areas with increased reports of shots fired outside the primary coverage areas. Between January 1 and November 16, 2025, the Raven sensors generated roughly 6,074 alerts. However, this number does not represent 6,074 separate incidents. Multiple sensors often trigger alerts for the same event, and delays between shots, such as reloading, can create additional notifications.
