Space Intelligence is a UK-based organization that produces high-quality forest maps by combining petabytes of satellite data with machine learning and ecological expertise, transforming raw mapping data into trustworthy insights so decision-makers can protect and restore forests globally. The team faced a critical scaling problem: desktop computing infrastructure could only map small areas in weeks at a time, making it impossible to address urgent global needs. As Chief Scientist Ed Mitchard put it: “Every decade we lose 10% of the world’s forest. We need to move more quickly.” The EU’s new Deforestation Regulation, which requires mapping 39 tropical countries, made cloud infrastructure essential.
Space Intelligence migrated to Microsoft Azure and integrated the Microsoft Planetary Computer ecosystem — which provides fast, API-driven access to petabyte-scale environmental data from NASA, ESA, and JAXA — alongside Microsoft Foundry’s Azure Machine Learning capabilities. The infrastructure enabled the team to process massive optical, radar, and LiDAR satellite datasets at scale with dramatically improved compute power. Ben Ritchie, Head of Engineering, noted: “Azure provides very, very high scale, remarkably affordable compute.” Machine learning models reached 99% accuracy distinguishing cocoa agroforestry from natural forest in West Africa, a critical capability for verifying zero-deforestation claims.
The results were transformative. Space Intelligence increased data production by 100 times and reduced mapping time from six months to six weeks, a 75% reduction. In a single year, the company mapped 3 billion hectares across more than 50 countries — work that would have taken approximately six years under the previous workflow. Head of AI Stephanie Earp confirmed: “With Microsoft Foundry and Planetary Computer we’ve increased data production by 100 times.” The advances accelerate global capacity to verify forest conservation efforts and unlock financing for nature-based climate solutions.