FireLens is an onboard burn severity intelligence application built on Little Place Labs' Orbitfy platform, a flight-proven, TRL-8 edge processing environment demonstrated in orbit in July 2025. Designed for SmallSats/MicroSats carrying NIR and SWIR sensors, FireLens computes a differenced Normalized Burn Ratio (dNBR) onboard using the NASA ABoVE Landsat-derived dNBR dataset methodology, classifies burn severity in real time, and transmits a compact alert packet, under 100 kilobytes, before a full scene reaches the ground. It then issues an adaptive retask request to the same or a coordinated satellite via OrbitfyNETWORK. Key features include Docker-containerized deployment on NVIDIA Jetson AGX Orin and Xilinx Versal class hardware, edge AI models within a 5-30 MB operational footprint derived from LPL's co-authored PyNAS research with ESA's Φ-lab (Scientific Reports, 2025), and a closed-loop CONOPS that converts a single satellite pass into a prioritized land resilience workflow; sensing, deciding, alerting, and retasking under real SmallSat constraints.
Bosco Lai, Gaurav Bajaj, Brian Barris, Matthew Whitley