Enhanced Agriculture Datasets for Remote Crop Monitoring
The lack of labeled crop datasets represents a significant challenge for crop analytics. Pula Advisors and partner organizations will create and augment labeled remote sensing and field datasets during the growing season and at harvest time in Kenya and Zambia. The project will apply machine learning models to assess risk and gain insight into the impact of weather, agroecological conditions, and agronomic behaviors of farm yields and overall productivity. Additionally, the project will build national yield maps for the main crops in the two countries by collecting data points for 4,000 individual smallholder farmer households. Data will be made available via a public API and will be accessible through Github.
Kenya and Zambia
Lulu Saida, Pula Advisors
Enhancing Agricultural Datasets for Report Crop Monitoring
This project improved traditional crop experiment methodology to increase the accuracy of yield prediction models. Read More about the approach and outcomes of cereal crops in Zambia.
Visit the Bird’s Eye dashboard to access 4,000 data points and metadata covering crop health, field boundaries, and crop type classification collected by Pula.
February 2023