How It Works
A plain-language walkthrough of what happens between a drone entering your airspace and your team knowing about it.
Sensor Input
Camera by default — RF or audio optional as you add them (radar on the roadmap).
Edge Processing
Signal and video are processed locally, on-site, in real time.
AI Classification
Detections are classified and, when more than one sensor is present, cross-verified for confidence.
Alert & Recording
A verified alert reaches your team instantly. Behind it, an agentic AI layer is already working — tracking the incident, correlating it against your site's history, and assembling a report automatically. What would normally take an analyst time to compile is ready the moment you need it.
Start with what you already have
Joint-Brain Fusion Logic
Our term for how Herkeios handles multiple sensors when they're present. Herkeios detects reliably from a single sensor — camera alone is enough to get started. When RF or acoustic are added, Joint-Brain Fusion Logic cross-verifies signals across them, raising confidence further before an alert is triggered. More sensors mean stronger confirmation, not a requirement to function at all. Radar support is on the roadmap — see FAQ.
Detection and reporting — that's the scope
Herkeios detects, tracks, and alerts. It never acts on a drone directly — mitigation in the EU is restricted to authorized state actors by law. What you do with a confirmed detection is your organization's call.
Performance depends on real-world conditions
The specific device, network conditions between sensors and the local processing node, and environmental factors like weather and site layout all affect results — true of any detection system, and we'd rather say so upfront than have it surface as a surprise during a pilot. See the FAQ for the full breakdown.