Building Scalable Food Security Solutions with Satellite Data and AI

A new public–private collaboration will harness satellite data and AI to strengthen food security across Sub-Saharan Africa, starting in Kenya and designed to scale across the region.

Reliable agricultural data is foundational to food security, economic stability, and sustainable growth. Project CroME (Crops Measurement and Evaluation) brings together the Kenyan Ministry of Agriculture and Livestock Development, Kenyan Space Agency, the East Africa Grain Council, NASA Harvest and the Microsoft AI for Good Lab, to build an open, government-anchored geospatial agricultural data system. The initiative aims to improve the quality, consistency, and accessibility of agricultural information, enabling better decisions across government, industry, financial services, and development partners.

The system will be Kenya-led and built for long-term use: affordable, adaptable, and structured so national institutions can maintain and evolve it over time. By prioritizing open data and national ownership, CroME is intended to function as durable public infrastructure, supporting innovation, transparency, and resilience well beyond the life of a single project. Leveraging advances in satellite data,  geospatial foundation models, and compute capabilities, CroME is designed  to make high-quality and timely agricultural insights more affordable and scalable to produce and sustain as a public good. By supporting technical expertise within national institutions and producing regularly updated, well documented datasets as a public good, CroME moves away from one-off demonstration projects towards a national operational system that serves public and private sectors alike.  

CroME is designed to establish a shared, reliable public-good data foundation that integrates Earth observation data, advances in AI geospatial modeling and foundation models, and systematically collected ground data. The resulting datasets will support agricultural estimates and statistics, market and trade insights, early-warning systems, and advisory and financial services. The system is built around open models, modular architectures, and platform- agnostic design, ensuring flexibility, long-term national sovereignty, and independence from any single platform.

A core principle of the initiative is national ownership and sustainability. The system is being co-developed with Kenyan public and private sector partners to ensure it aligns with national priorities, fits operational needs, builds on past investment and initiatives and strengthens local capacity. By embedding technical expertise within national institutions and releasing outputs on a regular schedule, with documented accuracy and as a public good, the project aims to reduce reliance on one-off analyses and enable continuous improvement over time.

CroME is also designed as a model that can be replicated across diverse geographies. Beginning in Kenya, the initiative is intended to scale to other countries in the wider region—demonstrating how open, operational agricultural data systems can be built, sustained, and expanded through strong partnerships and shared standards.

This initiative is carried out in close collaboration with the Food and Agriculture Organization of the United Nations (FAO) through its ESTELLA project, ensuring coherence with existing systems and alignment with ongoing initiatives and national priorities. Together, the partners aim to establish a scalable blueprint for open agricultural data systems that can strengthen food security, market transparency, and resilience, providing a regularly updated public good that public and private sector alike can build on and integrate into operational decision support systems and services.  

Coordinated jointly by NASA Harvest and the Microsoft AI for Good Lab, CroME will be be implemented and owned by Kenyan government institutions and partners, with technical support from NASA Harvest’s  international network of partners including University of Maryland, University of Strasbourg, Arizona State University, Virginia Tech, Monash University, VITO, Allen Institute for AI (AI2), the International Institute for Applied Systems Analysis, EAGC and FAO, as well as with UNDP AI Hub for Sustainable Development, the Italian Space Agency, Italy’s Ministry of Enterprises and Made in Italy, and regional stakeholders working across the agricultural value chain.

CroME is jointly funded by the Gates Foundation and the Microsoft AI for Good Lab.

Next
Next

Where to Draw the Line: How Earth Observation Data Can Help Design High-Performing Agricultural Insurance