Author(s): Esteban Gutierrez; Daniel Ruiz-Beamonte; Manuel Cozar; Jorge Aznar; Ignacio Latre; Eduardo Garcia; Alejandro Gonzalez; David Zambrana-Vasquez; Yassine Rqiq Moukhtari; Cesar Suela Cedenilla; Aurora Garcia Jimenez; Dorra Jouini
Linked Author(s):
Keywords: Artificial intelligence; Deficit irrigation; Machine learning; Precision agriculture; Vineyards; Water management
Abstract: Efficient water management is a critical challenge in viticulture under increasing climate variability and water scarcity. Regulated deficit irrigation (RDI) offers a promising strategy to reduce water consumption while maintaining grape quality; however, its implementation requires accurate and timely information on vine water status and environmental conditions. This study presents the ViñAI tool, a decision-support system that integrates artificial intelligence (AI) models with multi-source environmental data to optimize irrigation scheduling in vineyards. The system combines open-access meteorological data, satellite-derived evapotranspiration, and optional in-field sensor measurements to generate data-driven irrigation recommendations. Machine learning models were evaluated using a standardized framework, with Extreme Gradient Boosting (XGBoost) selected as the best-performing algorithm. Results indicate that AI-assisted irrigation strategies can significantly improve water-use efficiency, reduce energy consumption, and enhance economic performance. The proposed methodology supports scalable deployment in sensor-limited environments, contributing to sustainable and climate-resilient vineyard management.
Year: 2026