The same lightweight VHF radio tag biologists already use — unchanged, no new hardware on the animal.
Carries the receiver aloft. An antenna hangs below on a cable; onboard software pulls the signal out of the noise.
One handheld controller shows the live flight track and detection map — no laptop needed in the field.
Wildlife tags are designed to be as light and long-lived as possible, so they transmit at low power and can't afford the circuitry that would hold pulse timing and frequency perfectly steady. That's the right trade-off for the animal, but it makes the signal faint and slightly unpredictable. A human listening on headphones, or a basic receiver, only catches one pulse at a time — miss it, and the tag might as well not be there.
UAV-RT's onboard software instead searches across a range of possible timings and frequencies, then mathematically combines several faint pulses into one clear signal — like stacking several blurry photos to get one sharp one.
The vehicle can fly a preplanned search mission or be piloted manually. Either way, flight control, telemetry, and the live tag-detection map all run off one handheld controller — no laptop in the field.
Full operational readiness takes about 2–3 minutes after power-up.
The drone's own motors and electronics generate radio noise that can swamp a faint tag signal. Hanging the antenna several meters below the airframe on a cable — either a slung monopole for maximum range, or a directional Yagi for bearing precision — moves it away from that noise.
Custom radio-shielding fabric around the airframe cuts self-generated noise further still.
Remote sites don't have parking lots next to them. The vehicle's arms fold down so the whole system — drone, antenna, and controller — packs into a single backpack for the hike in.
No case to carry, no vehicle required to get close. If you can hike to the site, you can fly from it.
MATLAB/C++/ROS2 detection algorithm
TagTrackerCustom QGroundControl-based ground-station app
MavlinkTagController2SDR → signal pipeline → MAVLink controller, with a no-hardware simulator mode
GitBook build docsFull vehicle build instructions
CSV → map/KML toolComing soon — upload flight detection data, view it on a map, export KML