Season‑long vegetation tracking (NDVI/related indices), cloud‑robust compositing, stress detection, and anomaly alerts.
The Anomaly Detection product helps insurers and farmers quickly identify unexpected patterns in crop performance.
By benchmarking fields against similar parcels and multi-year historical profiles, it pinpoints areas showing unusual growth, stress, or disease symptoms.
This enables early interventions, reduces claim investigation times, and improves the accuracy of underwriting decisions.
It’s an essential product for scaling modern agro-insurance programs.
At-a-glance #
- Business Processes: Product Development; Product Sales; Underwriting; Loss Adjustment; Claims Handling
- Thematic Sector: Land – Agriculture
- Service Objective: Detect abnormal crop performance to flag potential stress, disease, or risk areas.
Main Steps #
- Build cloud-free parcel-level vegetation-index time series
- Benchmark vs similar fields and historical profiles
- Rank/flag anomalies
Inputs #
- Optical: Sentinel-2
- Supporting: Meteorological; in-situ crop/yield (optional)
Resolution & Coverage #
- Spatial: 10 m; farm → watershed
- Coverage: Global
Frequency / Timeliness #
- Frequency: Daily (imagery-dependent)
- Timeliness: Near real-time
Accuracy / Constraints #
- Indirect indicator; confidence rises with converging evidence (weather, ground)
Formats / Delivery #
- CSV parcel aggregates; time-series dashboard/API
Limitations / Notes #
- Variety/soil/management heterogeneity may trigger non-damage anomalies