Adavnced automation skills are transforming many sectors of human endeavour, including transportation, the environment, business, and agriculture. Many countries utilise an excessive quantity of already limited fresh water resources. This work demonstrates how to apply ML algorithms in an automation network based irrigation structure to predict soil moisture in order to optimise irrigation water use. Field data from installed sensors (humidity,moisture, temperature ,radiation) and virtual climate prediction information are used to forecast future soil moisture. The efficiency of numerous ML approaches for predicting soil moisture is investigated, and the GBRT results are encouraging in both accuracy and prediction. The approaches proposed can be an important area of research for maximising irrigation water usage.