Svitovyd
Field-tested in a US Navy drone trial

Passive, camera-only drone detection.

Autonomous drones fly by sight and emit no radio signal. Svitovyd sees them anyway: a low-cost 360° camera and the CCTV you already own, feeding edge software that spots drones and reports where they are.

The gap cameras close

The usual counter-drone sensors were built for drones that talk to a pilot or are big enough to reflect radar. The new threat is neither.

RF SENSORS

Nothing to hear

An autonomous drone navigates by its own camera. With no control link and no video downlink, there is no radio signal to detect.

RADAR

Too small, too slow

Small, slow targets are hard for radar to separate from clutter, birds and the ground.

SVITOVYD

If it can see, so can we

A drone that flies by sight has to be visible. Svitovyd watches the sky passively, emits nothing and cannot be jammed.

How it works

Cameras at the edge, one shared node to put the pieces together.

STEP 1

Watch

The low-cost 360° Svitovyd Camera covers the whole horizon. Existing CCTV cameras join the same network.

STEP 2

Detect

Edge software on each camera spots drones and rejects birds and clouds.

STEP 3

Point

Every detection is reported as a direction: bearing and elevation from that camera.

STEP 4

Triangulate

A shared node combines directions from several cameras into range, height and speed.

September 2026
US Navy
drone trial

Svitovyd has been field-tested at a US Navy drone trial, with Svitovyd Cameras recording and detecting live drone flights in the field.

The Svitovyd Camera

Four wide-angle cameras and an NVIDIA edge GPU in one compact, 3D-printed unit assembled with screws only, so it can be serviced in the field with hand tools.

Because detection runs on the unit, only detections leave it, not video. The same software also runs on feeds from CCTV cameras already on site.

Sensing
Passive, visible light only
Coverage
360° horizon per unit
Cameras
4 × wide-angle, plus existing CCTV
Compute
NVIDIA edge GPU
Output
Direction per camera; range, height and speed from the shared node
Calibration
Star-based, per camera

Technology

Built on proven robotics and edge-AI tooling.

NVIDIA edge GPUs TensorRT detectors ROS 2 / DDS AWS

Why “Svitovyd”

Svitovyd is a figure from Slavic tradition, depicted with four faces so that he looks in every direction at once. One watcher, all horizons. That is the design of the camera: four lenses, one view of the whole sky.

Talk to us

Interested in a pilot deployment, a field trial, or running Svitovyd on your existing cameras? Send us a note.

valentyn.stadnytskyi@svitovyd.com