Introduction: YieldPredict provides a multi-crop yield forecast service for farms and agriculture organisations (cooperatives, farm associations, agrifood companies, etc.) at the European level. It is an extension of a use case developed within the Andalusian Technological Demonstrator for the Agri- Food sector, generated for the project “Towards the creation of the Federated Data Space of the Andalusian Agri-Food Sector (EDAAn)” funded by the Spanish RIA National call (PITEDS24-2024- SEDIA). The main objective of YieldPredict is to offer the Data Space as a Service (DSaaS) to the Agriculture of Data partnership, thus enabling an ecosystem to support the engagement of other use cases and scaling-up it to serve the sector in the Europe-wide context. Methodologies: In this service, Earth Observation data, historical production records, meteorological, phenological, and organisational data are integrated, at parcel or farm level, into a trusted and reliable data sharing ecosystem to be modelled into an artificial intelligence (AI) prediction service, with sufficient maturity to provide users with a series of technological capabilities for decision support. IA predictions are generated from multivariable machine learning algorithms (SARIMAX, Random Forest and other ensembles) which are trained to reach high percentage of model confidence, based on standard metrics (R2, RMSE) and past production records, so they can deliver future predictions. Output: Crop yield forecasts for the coming agricultural year, generated based on aggregated data from individual plots or farms, which have a high degree of reliability and a low margin of error. These forecasts provide essential information for decision-making regarding the upcoming period Impact: Farmers and agri-food organizations have access to a service that allows them to know (with high confidence) their crop yields in advance, enabling them to make informed decisions regarding fertilizer use, hiring, labour, fuel and energy saving. What I can offer: New users will be able to test YieldPredict with new scenarios in different countries across different crops, such as almonds, fruit trees, pistachios, and other woody crops. AI predictive models generated in different scenarios can be validated, thereby obtaining new, enriched models with greater generalization power. What I need: New users to sign up and join the service, so they can participate in testing, add new crops and geographic regions, provide feed-back and help refine AI models and test features under different conditions and scenarios.

Source: Annual AgData Use Case Summit 2026 booklet.