Title: CEO

Association/Company: Soonapse Srl

> Professional profile

Speech session
Climatic conditions and agricultural progress, protagonist agriculture

Speech
From weather forecasting to predictive water balance calculation in AI systems for Smart Irrigation

Abstract
AI systems for producing predictive irrigation advice are based on data collected in the field in real time from sensors. However, to generate forecasts of the water balance on subsequent days, they must consider a variant that is itself predictive: weather forecasts.
This variant boasts 80%-90% reliability for the first day, gradually decreasing to about 40% on the fifth day.
In the talk, the methods used to reduce the impact of weather forecast deviations in the AI model that produces the irrigation advice will be explained, keeping the reliability of the water balance forecast as close to 100% as possible.