Generating Affordable, Scalable Agriculture Data through Mobile-Based Advertising, Field Digitization and Farm Record Keeping

Agriculture development strategies of many developing countries and institutions require cost-effective tools and methodologies for data collection and analysis of field-level agricultural production to drive investment and decision making. These datasets start with accurate field size estimates, which are the foundation of yield assessments. Optical and radar satellite remote sensing datasets, combined with analytical models, can provide an accurate picture of crop health and deliver detailed recommendations for better crop management. However, gathering low cost, high quality training data and understanding the context in which decisions are made is necessary to localize these models. This project developed an entirely remote, mobile-based data gathering system which used human-centered design to create a mobile application that can gather field boundary and crop type information in Kenya. The project involved redesigning a crop monitoring tool to be easily usable by smallholder growers in Kenya, demonstrating that we could scale it to thousands of potential users with no face-to-face engagement via an advertising campaign, and determining the accuracy of the resulting field boundaries. We were registered 2553 farmers and mapped 358 fields without any face-to-face interaction during a three-month campaign. The cost of this engagement was $1.04 per registered user. This system demonstrates the potential to deliver crop analytics at scale in Africa, enabling an improvement in the ability of digital tools to deliver a state of the art, data-driven view of agricultural activity on the continent.