What it is
Dronevade is an edge platform for detecting and tracking drones in environments where bandwidth is scarce and false positives are expensive — wildfire response, utility infrastructure, restricted airspace.
How it works
Custom YOLO detection models run on edge hardware (Jetson / embedded), fused with RF and thermal signals so the system holds up when any single sensor is unreliable. Telemetry is designed to be explainable — operators can see why something was flagged.
Why it matters
It’s the same pattern as the rest of my work: ML that thrives in the real world, on the edge, where the stakes are physical.