Use a thermal infrared sensor to detect heat anomalies associated with wildfires. Process imagery on-board to generate alerts and downlink only confirmed hotspot data to reduce bandwidth requirements.
Use a thermal infrared sensor to detect heat anomalies associated with wildfires. Process imagery on-board to generate alerts and downlink only confirmed hotspot data to reduce bandwidth requirements.
This is an advanced-level project with an estimated timeline of 20-26 months using a 2U form factor.
Wildfires are detected faster when monitored from space, but current satellite systems have revisit gaps that can leave fires undetected for hours. A CubeSat-based wildfire detection payload uses a thermal infrared sensor to identify heat anomalies on the ground, processing the imagery onboard to distinguish actual fires from sun-heated surfaces, industrial heat sources, and sensor noise. Only confirmed hotspot detections are downlinked, conserving precious bandwidth for actionable data. The onboard processing chain thermal thresholding, spatial clustering, false-positive rejection introduces students to real-time embedded image processing under tight power and compute constraints. The project combines thermal imaging hardware, machine learning or rule-based classification, attitude control for nadir pointing, and a ground-side alert pipeline. While a student CubeSat cannot match the resolution of professional fire monitoring systems, it demonstrates rapid revisit cadence and autonomous detection capabilities that complement existing systems. This is one of the more challenging projects in the catalog, touching nearly every spacecraft subsystem, but the societal relevance of wildfire early warning makes it a compelling mission story for funding proposals and outreach.
FLIR Lepton 3.5 thermal camera module (~$250, 160x120 resolution, SPI, 150 mW) provides 8-14 µm LWIR sensitivity to detect thermal anomalies. Pair with OV2640 visible camera for dual-band confirmation. Onboard processing: threshold thermal pixels >350K, cluster adjacent hot pixels, reject single-pixel noise. Downlink only confirmed hotspot coordinates + thumbnail. Requires nadir pointing (±15°) for useful ground coverage. Consider ADCS or passive gravity gradient stabilization.
FIRMS (NASA Fire Information for Resource Management System) uses MODIS/VIIRS at 375m-1km resolution a CubeSat thermal sensor cannot compete on resolution but can demonstrate rapid revisit cadence. FLIR Lepton 3.5 at 500 km altitude gives ~2.5 km/pixel GSD sufficient to detect large wildfires (>1 km²) but not spot fires. Key challenge: onboard processing requires more compute than PyCubed SAMD51 provides need ESP32-S3 or Raspberry Pi Zero co-processor. Nadir pointing requirement adds ADCS complexity. EMI from thermal camera switching can affect radio needs careful PCB layout. Cost: $1,500-$4,000 for camera + co-processor + ADCS assist. Complexity: advanced. Combines imaging, ML, ADCS, and thermal engineering challenges. 2U recommended for camera + optics + processing.
This project spans 4 disciplines, making it suitable for interdisciplinary student teams.
Ready to take on this project? Here's a general roadmap that applies to most CubeSat missions:
Connect with a Blackwing chapter for mentorship, platform access, and a path to orbit.