In general. automated farming systems make decisions based on static models built from the properties of the plant. in the contrast, irrigation decisions in our suggested method are dynamically changing environmental conditions. the model"s learning process reveals the mathematical links between the environmental factors employed in the determining the irrigation habit and gradually improves its learning techniques as irrigation data accumulates int the model. to analyze overall system overall system performance, we constructed a test environment for the sensor edge, mobile client, and decision service in the cloud.