AI for Land Productivity & Crop Yields Modeling

Overview

In Uganda's Lake Kyoga basin, smallholder farmers are watching their harvests become unpredictable and the tools to fight back have never existed, until now. The AI4D-supported project led by Makerere University is deploying artificial intelligence and machine learning to model land productivity and crop yields under shifting climate conditions, giving farmers, policymakers, and researchers a data-driven window into Uganda's agricultural future.

Uganda's smallholder farmers who produce the majority of the country's food face a mounting crisis. Erratic rainfall, soil degradation, and rising temperatures are upending centuries of farming knowledge, yet there are almost no accessible, locally-grounded tools to forecast how different land management practices will perform as the climate changes. Without this intelligence, adaptation is guesswork.

The project is building machine learning-based crop yield prediction models calibrated specifically to the Kyoga basin, drawing on both satellite data and direct household interviews with farmers to ensure that local realities shape the science. Crucially, the project is simultaneously building the human infrastructure to sustain this work: MSc students at Makerere University are being trained in AI applications for agriculture and climate adaptation seeding a generation of African researchers capable of leading this work long after the project ends.

The impact is already taking shape across two dimensions. Farmers' lived experiences gathered through community engagement are being embedded into models that can assess how different farm-based land management strategies hold up under climate stress. And by training African researchers to lead AI-driven agricultural analysis, the project is building durable, continent-rooted capacity to tackle food security challenges at scale. In a region where climate change threatens millions of livelihoods, this is AI that grows where it's planted.