The FERTWINS project aims to explore predictive models by leveraging emerging technologies like Digital Twins and Explainable AI (XAI). Employing these technologies for the acquisition, transfer, and storage of relevant data will enable decision-making based on real-time data sourced from satellite imagery, sensors, and weather stations. This approach is intended to enhance fertigation management and reduce the usage of phytosanitary products, thus accelerating the digital transition and minimizing environmental impact. FERTWINS seeks to improve information flow, significantly boosting the productivity of the agri-food chain.

The project has technical objectives, including the characterization and optimization of crops and fertigation using agronomic models and ontologies. This encompasses studying models to predict evapotranspiration, investigating geospatial information modeling techniques, and analyzing agronomic models for fungal diseases in woody crops, with a focus on the potential of Digital Twins and XAI for data transmission.

Furthermore, the project will research and design an environment based on intelligent agents and Digital Twins for crop monitoring, phenotype evolution simulation based on internal and external agents, and fertigation control using low-power wireless technologies. This involves information fusion from heterogeneous data sources, designing networks for monitoring and control in smart agriculture scenarios, and creating an environment based on Virtual Agent Organizations and Digital Twins.

Lastly, FERTWINS will delve into XAI-based models for predicting evapotranspiration, crop yields, and fungal diseases, as well as cognitive systems for optimizing fertigation. This includes designing crop evapotranspiration predictions from IoT and satellite imaging, Explorable Ensemble Learning models for predicting yields of woody crops and designing a cognitive system for autonomous decision-making in fertigation and phytosanitary applications in woody crops.