False positives remain an operational challenge for organisations using acoustic threat detection across campuses, critical infrastructure, utilities and communities, particularly in environments where hills, buildings and other environmental changes that affect the surrounding skyline can complicate classification. For security teams, those conditions can mean more alerts to review and more time spent determining which events actually require a response.

Customised environmental set up

Acoem’s latest Cadence software release addresses this challenge by allowing administrators to set elevation and horizon thresholds for individual sensors based on their surrounding environment. This is useful in areas where loud noises like fireworks are common, as their acoustic characteristics can be similar to gunshots. The elevation threshold helps the system more accurately classify fireworks and other skyline noise as non-threats. Across larger deployments, administrators can also tailor individual sensors to local conditions rather than applying the same settings across an entire fleet.

Refining acoustic signal and noise analysis

Acoem is also refining its AI methodology with changes to acoustic signal and noise analytics intended to further detect anomalies and reduce false positives. The update builds on the company’s data science expertise to improve how the system distinguishes potential gunfire from surrounding noise.

These changes promise better and more accurate information and build on existing Acoem ATD capabilities including:

• Dual-detection capability: Employs shockwave-enhanced muzzle-blast detection to analyse acoustic signatures, ensuring highly reliable gunshot verification across both urban and rural environments.
• 96-attribute analysis: Uses an onboard AI neural network to instantly analyse acoustic events and filter out ambient noise and urban distractions.
• Edge-based processing: Verifies threats locally at the sensor in real time, eliminating latency and reducing reliance on backend server infrastructure.
• PTZ camera integration: Instantly directs connected PTZ cameras toward the origin of a detected acoustic threat, delivering immediate visual confirmation for actionable situational awareness.

“The volume of security system alarms are one of the biggest drains on security resources today. This update highlights how we leverage our deep data science expertise to continuously evolve our technology alongside our customers,” said Mike ODea, Managing Director, Acoem ATD. “By giving teams the ability to adapt detection to their specific environments, we're delivering the most flexible, accurate acoustic threat protection available to keep communities and critical infrastructure safe.”

The software updates will be available in Q4 2026 for existing customers and they will have the option to apply the updates through the Cadence software platform. System administrators will receive official release notes ahead of the rollout, with over-the-network updates delivered directly to edge sensors without requiring additional backend server infrastructure.