Sensors that measure. Pipelines that validate.
Atlas runs a fully calibrated sensor stack — multispectral, radiometric thermal, high-resolution RGB and LiDAR — fed into a photogrammetry and AI pipeline benchmarked against in-field ground truth on every mission.
Accuracy you can defend in an audit
Every claim Atlas makes about a field is cross-checked against physical ground truth. Below are the standing performance benchmarks we hold ourselves to.
Payload catalogue
Mission profile determines payload selection. The right sensor pays for itself in fewer flights.
MicaSense RedEdge-P Dual
- Bands
- 10 bands · 444–840nm + Panchromatic
- Resolution
- 12 cm/px @ 120m AGL
- Coverage
- 180 ha/flight
- Use
- NDVI · NDRE · RECI · CIRE indices
DJI Zenmuse H20T
- Bands
- LWIR 8–14μm + 20MP RGB + 1200m laser
- Resolution
- 640×512 thermal · ±0.1°C sensitivity
- Coverage
- 120 ha/flight
- Use
- Irrigation audit · disease cold-zones · livestock counts
Sony α7R IV (61MP)
- Bands
- Full-frame RGB · 9504×6336
- Resolution
- 1.8 cm/px @ 80m AGL
- Coverage
- 90 ha/flight
- Use
- Stand count · weed mapping · orthomosaics
DJI Zenmuse L1 LiDAR
- Bands
- 905nm pulsed · 240k pts/s
- Resolution
- 5 cm vertical · 3 returns
- Coverage
- 80 ha/flight
- Use
- Biomass volumetry · canopy height · drainage modelling
From photon to decision
Six stages, each instrumented and version-controlled. Raw and processed artefacts are retained for audit.
Capture
RTK-corrected autonomous flight at 120m AGL. Reflectance panel calibration pre- and post-flight. Lossless storage on dual SD with on-board hash.
Radiometric Calibration
Per-band reflectance correction against calibration target. Atmospheric scatter compensation. Channel co-registration to sub-pixel accuracy.
Photogrammetry
Structure-from-motion bundle adjustment. Orthomosaic generation at native GSD. DSM/DTM computation for slope-aware indices.
Index Computation
NDVI · NDRE · CIRE · RECI · CWSI · canopy temperature differential. Zoned via K-means clustering with agronomist-tunable bands.
Anomaly Detection
Computer-vision pipeline flags emergence gaps, thermal hotspots, disease cold-zones. Bounding boxes + confidence + GPS coordinates.
QA & Delivery
Agronomist sign-off, format packaging (GeoTIFF / SHP / KML / GeoPDF), signed flight log, Ops Center push within mission SLA.
Every claim has evidence
Aerial indices are meaningless without verification. These are the four standing protocols Atlas runs on every mission.
Ground Sampling Distance
All flights flown at 120m AGL with RTK positioning, delivering a consistent sub-5cm absolute spatial accuracy confirmed on every flight run.
Spectroradiometer Calibration
NDVI/NDRE values are validated against in-field handheld spectroradiometer readings on grounded calibration targets.
Sensor Benchmarks
We verify sensor arrays within standardized climate wind chambers to maintain consistent spectral measurements.
John Deere Integration
Raw deliverables automatically packetize to John Deere Ops Center, QGIS, ArcGIS, and standard Trimble controller files within 24 hours.
See what the pipeline produces.
Sample deliverables, format matrix, and what your team can do with them.