Edge drone detection
A trained vision model runs on NVIDIA Jetson hardware for low-latency, camera-side inference.
See the detection pipelinePassive counter-UAS detection & tracking
Operational prototypeLamassu is building a passive, camera-first sensing system for detecting and tracking small drones at the edge - without relying on the target to transmit.
Demonstrated today
The current prototype combines real-time computer vision, edge inference, visual tracking, and physical pan-and-tilt control on field-relevant hardware.
A trained vision model runs on NVIDIA Jetson hardware for low-latency, camera-side inference.
See the detection pipelineDetections drive a calibrated optical pan-and-tilt prototype that keeps the selected object centered.
See tracking technologyA multi-node subscriber confirms observations across sensors, deduplicates alerts, and monitors node health.
View capability statusPlatform direction
Our architecture starts with what a sensor actually observed, preserves where it came from, and is designed to combine useful signals as integrations mature.
Passive cameras produce timestamped observations of objects in the low-altitude airspace.
Edge compute classifies candidates, maintains visual tracks, and limits unnecessary data movement.
Observations are corroborated across nodes today, with deeper multi-sensor fusion in development.
Structured track output for operator tools and command-and-control systems is on the roadmap.
Evidence over theater
We validate the hard parts directly: inference timing, target persistence, motor response, node health, and corroboration behavior. The result is an operational prototype—not a slide-deck simulation.
Conceptual interface shown. Current prototype output is an engineering visualization, not a deployed operator product.
Relevant environments
Lamassu is designed for evaluation in environments where small drones can create operational uncertainty and passive detection is valuable.
Evaluate passive optical sensing around fixed sites without adding active emissions.
Add a detection layer around energy, utility, data, and logistics infrastructure.
Explore deployable sensing for venues and events with defined operating boundaries.
Start with the environment
Share the site geometry, sensing constraints, and integration needs. We will tell you plainly where the current prototype fits and where it does not.
Contact Lamassu