Regional
QA-masked vegetation patterns from a real historical Sentinel-2 study. Plantation classification remains unvalidated.
Needs: Sentinel-2, boundaries, reference labelsOpen · Agentic · Community-driven
OpenPalm AI connects satellite imagery, geospatial evidence and AI research for oil palm landscapes, with a real historical study and clearly labelled autonomy simulations.
Research prototype · real study + labelled demo fixtures
Honest by design
QA-masked vegetation patterns from a real historical Sentinel-2 study. Plantation classification remains unvalidated.
Needs: Sentinel-2, boundaries, reference labelsIndividual crown detection needs finer imagery. Sentinel-2 cannot resolve single palm crowns.
Needs: Very-high-resolution imageryMeta-predicted CHMv2 height summaries are available, with dates and uncertainty kept visible.
Needs: CHM provenance; validation for accuracySimulated coverage routes on synthetic parcels, with computed geometry and route lengths.
Needs: Suitable spatial inputs; operational validationGeo Intelligence
Inspect real Sentinel-2 NDVI and Meta-predicted canopy height, explore pixel values and download the scientific run. Synthetic parcel fixtures stay separately labelled.
Explore the workspaceAgent Studio
Explore a visual pipeline draft, then inspect evidence, citations and a saved Muse Spark Open response in the Research Assistant. Live Baseten research requires protected development access.
See the studioAutonomy Lab
Generate deterministic coverage routes on synthetic parcels and inspect their computed lengths. The lab is a planning simulation, not an operational flight plan.
Open the labOpen research
Inspect source attribution, provenance and validation status alongside each dataset. Real study outputs and synthetic examples stay distinct; community publishing and measured benchmarks remain future work.
No account needed to explore the real study, saved research and demo workspace.