PAN-CAFE: Participatory AI Networking - Climate Advice for Future farming practice and Environments in Costa Rica

Project Details

Description

General introduction

How can we build a more efficient two-way communication system for the Costa Rican coffee sector that provides producers in eight biogeographic regions with sophisticated real-time climate forecasting models and associated optimal cultivation and variety choices? This project specifically aims at revitalizing the CRCAFE smartphone application and website. The aim is to collect and analyze the data from (small) coffee farmers (their expertise and results) in a more streamlined way, so that they can then receive recommendations that increase the resilience of their farm and the sector as a whole. In Costa Rica, thousands of small farmers depend on coffee cultivation for their livelihood. After 200 years of cultivation, the country is known for its high-quality coffee, the result of the right match between varieties on the one hand and local soil, local microclimate and specific management and processing practices on the other. That system is being disrupted by climate change. More extreme weather causes flash floods, landslides, forest fires. The coffee rust epidemic is damaging varieties and altitudes that were previously unaffected (as the plants suffer from drought and heat stress). A number of coffee farmers are abandoning coffee activities and/or are being lured to deforest higher areas and start planting there. Several management practices have recently been shown to have questionable ecological or economic soundness.


Research approach

The researchers help to promote accurate and rapid communication around improved coffee varieties. A framework for data collection and evaluation, and for better information sharing will be set up, using artificial intelligence. On the output side, recommendations on the use of specific coffee varieties and practices are tailored to the cultivation context. Expert feedback is collected to assess the added value of the recommendations through extensive workshops. The project uses a broad participatory approach: consulting with farmers, extension experts and other actors along the value chain, on the co-design, implementation and roll-out of the new applications in CRCAFE. The improved functionalities in the system are taught through multi-variety demonstration farms in seven biogeoclimatic regions.


Relevance/Valorisation

The project is expected to contribute to the capacity building of the national partners. The implementation of AI-LLM systems offers the possibility to capture expertise in a self-learning system where recommendations can evolve with future developments. In line with the cooperative spirit of the Costa Rican coffee sector, with producers, collectors, processors and exporters aligned with the ambition to produce the highest quality coffee, the participatory approach seems to be a good idea: the new applications in CRCAFE immediately gain a broader support and a sense of co-authorship. The revitalization of the CRCAFE smartphone application and website makes that users receive recommendations that actually bring more economic and social stability to the coffee farms.


Funding provider(s)
Departement Omgeving
AcronymPAN-CAFE
StatusActive
Effective start/end date1/09/2431/08/27

Funding

  • Vlaamse Overheid

Data Management Plan flag for FRIS

  • DMP not present

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.