---

Can AI technologies increase farmer’s resilience to climate change? Impact evaluation of Croppie

PI and co-PIs: Marcela Ibanez Diaz (University of Göttingen, Germany); Christian Bunn (CGIAR, Colombia); Romain Gaoutron (CGIAR, Colombia); Juan Carlos Muñoz-Mora (CGIAR, Colombia)

Funding amount: $150,000

Project overview: Coffee smallholders in Colombia are suffering the consequences of climate change. Rising temperatures and more variable weather can reduce yields and increase pest and disease pressure. Timely, reliable, and affordable yield forecasts could help households to plan and cope with income shocks. However, traditional field-based counting methods are labor-intensive and expensive. This project addresses that bottleneck by supporting the adoption and rigorous evaluation of Croppie. This machine-vision tool uses smartphone photos to estimate coffee yields by automatically counting fruit in the field. Working with coffee cooperatives and municipalities, the team will conduct a cluster-randomized controlled trial to assess whether improved yield information translates into enhanced risk management. Beyond immediate benefits for participating households, the project will generate a publicly shareable dataset linking georeferenced field images, yield estimates, and survey outcomes, enabling further research on climate impacts on coffee productivity. Overall, the project will provide new evidence on the links between climate variability and agricultural productivity and support scaling the tool to additional regions, value-chain actors (e.g., cooperatives and financial institutions), and potentially to other crops where low-cost yield forecasting can strengthen climate resilience.

Agriculture & Food Climate Finance & Economics Computer Vision & Remote Sensing