AI Mapping for Small-Scale Farm Transformation
Developing the first open-source AI maps to track yields and sustainable agricultural practices on small-scale farms in Africa.
Small-scale farmers are integral to Africa’s economy and food security, supplying around 80% of the continent’s food. They are also on the front lines of both climate change and land degradation, making their jobs increasingly difficult. Yet diagnosing these challenges and designing effective solutions at the farm level remains difficult, in part due to a persistent lack of data.
Accurate, farm-level data is needed to track yields and understand what’s impacting productivity. But mapping small-scale farms (those with less than 2 hectares of land) is difficult. Methods like traditional manual mapping and visual interpretation of satellite images are expensive and difficult to scale. Many advanced solutions utilizing AI and satellite imagery have been developed and tested, but these are used primarily in developed regions with access to high quality, annotated data and computation capacity. They are largely untested in developing regions that house smaller, irregularly shaped farms with mixed crops.
To help address these data gaps, WRI is developing the first AI-enabled, open-source tool to delineate small-scale farms across Africa, allowing us to better monitor crops and track yields at scale.
The core objectives of this project are to:
- Optimize AI models to accurately map boundaries of small-scale farms (under 2 hectares) across diverse farming systems.
- Integrate field-level crop data into AI models to improve crop-type mapping and productivity monitoring.
The new tool will provide ministries, agencies and partners with accurate, field-level data on small-scale farms in Africa for the first time. These insights will support 300,000 farmers, strengthen land-use policy, advance regenerative agriculture, and guide large-scale restoration and food security efforts.
For example, WRI and partners will use this information to:
- Understand current farming methods at the field level and promote sustainable agricultural practices.
- Develop localized restoration interventions for regenerative agriculture.
- Provide farm level monitoring and evidence on the impact of regenerative agriculture practices on crop yields.
- Advance actions to reduce greenhouse gas emissions in small-scale farms by supporting a transition to organic inputs such as bio-fertilizer and bio-pesticides.
- Build farm level evidence on where WRI and community-supported restoration champions will apply localized interventions — like agro-forestry, crop diversification, crop rotation, water conservation, adoption of organic fertilizers and farm restoration — to improve soil health and yields in small-scale farming systems.
If you are interested in early product access and testing the usability of the small-scale farms, get in touch with us.
Related Projects
Land & Carbon Lab
Launch PlatformLaunch Platform Visit ProjectConvened by World Resources Institute and the Bezos Earth Fund to develop breakthroughs in geospatial monitoring that power solutions for sustainable landscapes worldwide.
Part of Forests