Beany: An App to Provide Tailored Advice to Farmers to Improve Coffee Quality
Low-income countries participate in global value chains typically as suppliers of primary commodities, while value is added through processing in richer countries. This pattern keeps income low, and is mirrored within countries: smallholder farmers, who constitute the majority of the world's poor, produce low-quality goods sold locally at low prices, while processing is handled by exporters. A key reason, among several contributing factors, is the trade-off between quality control and value creation: as goods are processed, their quality becomes harder to observe, decreasing the price they command.
This tradeoff is evident in the coffee sector: while cherry quality is easy to assess, the quality of processed green beans is more opaque. In an ongoing project, the researchers developed a smartphone app that uses computer vision to analyze photos of green beans, verify quality and recommend purchases to our partner buyers, helping farmers move up the value chain. That is, it allows for provision of incentives for quality enhancement to individual farmers.
However, even when price incentives for quality exist and farmers have theoretical knowledge of good practices, they may struggle to meet standards: intra- or extra-household labor delegation might undermine their ability to determine whether they have followed the pre- and post harvest best practices, or they may struggle with adaptation to climatic variability. In such contexts, personalized, data-driven feedback could play a critical role. Thus, enhancing the system to infer causal relationships between specific defects and agronomic or processing practices would significantly increase its potential.
In this project, the researchers will train their existing technology to link defects to 6 key practices that could cause them, and to integrate tailored advice to farmers. This could then enable a broader evaluation of whether AI-enabled feedback mechanisms can enhance farmer decision-making, improve incomes, and promote more equitable participation in agricultural value chains.