Abstract: An irrigation water optimization system based on predictive calculation of water potential of soil through web/cloud is provided. A field data collection system includes a local weather station and a soil data detection device for each area a prediction is to be obtained. Sensors of water potential in the soil detect efforts made by the crop in using available water. A neural network provides the necessary irrigation predictions based on acquired data and evapotranspiration calculated by appropriate equations. Predictions specifically refer to concerned land and allow saving water.