Acoem will showcase the latest version of its ATD Gunshot Detection system, featuring an expanded acoustic intelligence engine capable of analyzing 96 distinct attributes of a gunshot event in real time, at ISC West 2026, March 23 to 27 in Las Vegas, Nevada.

96-Attribute Detection Technology

The 96-attribute framework reflects how modern gunshot detection has moved beyond simple loud-noise thresholds.

Each acoustic event is evaluated across a broad range of waveform, spectral, temporal, and ballistic characteristics, such as impulse shape, rise time, spectral content, and shockwave geometry.

This allows the system to distinguish true gunfire from environmental noise with greater precision, even in complex outdoor environments such as parking lots, campuses, critical infrastructure sites, and public venues.

AI-Powered Sensor

By performing this analysis directly on a single AI-powered sensor, Acoem ATD delivers immediate, validated alerts without relying on cloud processing, backend servers, or multi-sensor triangulation. Detection, classification, and localization occur at the edge, enabling actionable intelligence in real time.

“Not all gunshot detection systems analyze sound the same way,” said Timothy English, Managing Director at Acoem ATD. “Having 96 unique attributes reflects decades of acoustic research and real-world data. It allows the system to evaluate the full anatomy of a ballistic event rather than reacting to a handful of characteristics. That depth of analysis is what enables faster decisions and greater confidence when seconds matter.”

From Detection to Visual Confirmation

Acoem ATD’s AI-powered sensor and Cadence software tightly integrates with video management systems, transforming detection into immediate situational awareness.

Upon confirming a gunshot, the sensor automatically slews a connected PTZ camera to the source of gunfire using open standards (ONVIF-S). Security teams receive both audio validation and live video of the scene, without manual intervention or delay.

This automated camera response is especially critical in outdoor environments, where fixed camera coverage and human monitoring alone often leave blind spots. By directing PTZ cameras instantly, the system ensures that critical moments are captured as events unfold, supporting faster response and stronger post-incident investigations.

Showing for the first time at ISC West, Acoem has developed an on-prem version of its Cadence software for mission-critical environments such as utilities, government, and remote facilities, where outside internet connections are restricted for cybersecurity reasons. Unlike other platforms that centralize control, Cadence gives organizations direct autonomy to monitor sensors, validate alerts, and manage deployments in real time.

Single-Sensor Accuracy, Flexible Deployment

Unlike legacy gunshot detection architectures that require multiple sensors arranged in fixed meshes, Acoem ATD localizes threats with a single sensor by analyzing both the muzzle blast and the ballistic shockwave of a projectile. This approach reduces infrastructure complexity while extending coverage beyond traditional perimeter boundaries, including shots originating outside fenced or controlled areas.

Because intelligence resides on the device, the system is well-suited for both permanent and portable deployments. Sensors can be installed on existing poles, infrastructure, or mobile security trailers, making it possible to protect temporary events, evolving risk areas, or remote locations without redesigning the network or adding server capacity.

Edge AI Built for the Real World

Acoem’s edge-based AI architecture is designed for noisy, unpredictable environments. The neural network continuously evaluates acoustic conditions and applies advanced denoising and pattern recognition techniques to isolate true ballistic signatures. This reduces false alarms while preserving sensitivity to real threats, even when background noise levels are high.

By avoiding continuous audio streaming and limiting recordings to brief, event-triggered clips, the system also supports privacy-by-design principles while maintaining operational effectiveness.