Case Study: Descartes Labs achieves petabyte-scale satellite analysis to improve crop forecasting and global food security with Google Cloud Platform

Descartes Labs Advancing global food security

Descartes Labs is a Santa Fe–based company that applies machine learning to satellite imagery to model complex systems like agriculture, forestry, and water resources to improve global food security. Faced with continuously growing petabytes of historical and incoming satellite data, the company needed a way to process years of imagery (including the full Landsat archive) quickly and cost‑effectively without investing millions in physical infrastructure or waiting months to run analyses.

By adopting Google Cloud Platform—using Cloud Storage, Compute Engine (including preemptible VMs), Kubernetes Engine, BigQuery, Cloud SQL, Pub/Sub and Cloud Bigtable—Descartes Labs scaled on‑demand processing to run the entire Landsat archive in about 15 hours. The move cut processing time from months to hours, accelerated time‑to‑market by over six months, enabled faster and more accurate crop‑yield forecasts (even outpacing government reports), avoided large capital expenditures, and positioned the company to analyze 15–30+ petabytes for early famine warnings and global forecasting.

Descartes Labs

Mark Johnson

CEO