Real-world audio collected at the Army’s Warden 2026 event gave Certus Innovations the opportunity to test its DART and FALCON TEAM-ML plugins in operationally realistic conditions. The results marked significant progress toward field-ready gunshot and UAS detection, classification and localization.

Certus Innovations has submitted new research to DCASE 2026 showing that choosing the right audio feature extraction technique can improve gunshot detection model accuracy by up to 20%. The paper, co-authored by Sinclair Gurny and Ryan Quinn, marks a major step forward in building a complete, high-accuracy gunshot audio ML pipeline.

Introducing C3GD; The Largest, Most Detailed Gunshot Audio Dataset Ever Released for Open Science

We tested gunshot detection across 12 calibers, 22 platforms, and 27 cartridges, including two never before seen in public datasets. The results revealed a critical gap between lab performance and real-world conditions, where background noise often mimics gunshots. By shifting from traditional audio analysis to spectrogram-based approaches, and ultimately prioritizing high-frequency data, we uncovered a more accurate and field-ready method for isolating true gunfire.

5 insights we learned about the future of TAK on the AllThingsTAK livestream.

TAK TALK: Translation Without Limits

DMSAT: From Vision to Execution

In May 2025, President Trump submitted his FY2026 discretionary budget recommendations to the Senate Appropriations Committee, proposing what the Administration described as “unprecedented increases for defense and border security,” including a 13 percent increase in defense spending to $1.01 trillion.¹

Certus Innovations has been selected by the U.S. Air Force for a DMSAT Blanket Purchase Agreement supporting force protection and base security at scale. The award reflects Certus’ continued role as a trusted technology partner delivering mission-ready AI, cloud, and IoT solutions that enhance operational awareness and accelerate capability for Air Force Security Forces.

Across multiple Air Force units, RHINO has already proven what happens when manual bottlenecks disappear and leaders gain instant visibility into the fleets they rely on every day. By transforming paper-based inspections into real-time digital intelligence, RHINO is giving Airmen and operational-level leaders the speed, accuracy, and situational awareness they need to keep missions moving without interruption.

In high-stakes environments, clarity isn’t a luxury — it’s survival. DART delivers real-time acoustic gunshot detection that gives military units and first responders the intelligence they need the moment they need it. From distinguishing friendly fire on the battlefield to providing instant alerts during active shooter events, DART transforms raw sound into actionable insight. Integrated with TAK, it creates a unified operating picture that strengthens coordination, accelerates decision-making, and supports life-saving response when seconds count.

After training our DART (Distributed Acoustic Reasoner for TAK) model to detect gunfire with up to 95% accuracy, our team expected smooth sailing as we deployed the plugin in new environments. 

At Certus Innovations, we’re redefining what’s possible in acoustic machine learning. This post dives into our latest advancements with the DART (Distributed Acoustic Reasoner for TAK) project—exploring how real-world data, rigorous training, and field-driven design have produced exceptional results in gunshot detection accuracy. Learn how DART is shaping the future of acoustic intelligence, situational awareness, and mission-ready AI.

Protecting bases and critical infrastructure from unmanned aerial systems (UAS) requires more than blind faith in radar.

Modernizing Mission Logistics: How ATLAS is Revolutionizing Asset Tracking

Top 5 Tech Gaps in Government Operations—and How We’re Solving Them

Why 'Security-First' is the Future of Government Tech

The Certus Approach: Agile, Veteran-Led, and Mission-Aligned