AI tools in Agri DSS pipeline - the case of irrigated sugar beet

Precision and Smart irrigation are based on Decision Support Systems, allowing growers a sustainable use of the water resource and uniform and high level yield quality, a fundamental aspects in sugar beet supply chain. Irrigation scheduling in the case of sugar beet is a critical issue, because of sensitivity of the sucrose yield to non optimal water availability. For this reason, starting from an analysis of the supply chain actors, literature has been analysed to identify available tools already used in water management and in particular the use of AI. They emerge several Machine Learning approaches, already used in several crops can be used in sugar beet irrigation scheduling, and some already in the sugar beet. They mainly include fuzzy logics for recipe application, supervised learning to estimate crop evapotraspiration and recursive neural nets to estimate soil and plant water status. The analysis also envisaged the possibility to adopt other techniques already applied in precision agriculture, as LLM to include growers knowledge in managing to prevent conditions favourable to diseases.