An IoT-based predictive irrigation scheduling framework for precision soil moisture control in greenhouses

Efficient irrigation management is essential for sustainable crop production, yet conventional IoT-based systems often struggle to coordinate soil moisture prediction and control timing. This study proposes a predictive precision irrigation strategy that integrates evaluation (Δteval) and prediction (Δtpred) intervals to improve soil water regulation. A key innovation is the establishment of a proportional relationship, Δtpred = ½Δteval, which enables accurate anticipation of soil moisture depletion and coordinated irrigation responses. The framework was validated through theoretical analysis, simulations, and two greenhouse experiments using pot cultivation with bare soil, cabbage (Brassica oleracea var. capitata), and tomato (Solanum lycopersicum L.) plants. Theoretical and simulation results demonstrated that appropriate interval configurations maintain soil water content near predefined targets across different depletion rates. While shorter evaluation intervals improved control precision, longer intervals maintained comparable target levels with fewer irrigation activations. In the first experiment, three interval combinations were tested to regulate a target soil moisture level equivalent to 29% volumetric water content (VWC). Measured sensor values remained close to the setpoint under all configurations, with minor deviations among treatments. The second experiment applied a uniform interval setting (Δteval = 30 min; Δtpred = 15 min) while assigning sequential VWC targets ranging from 10% to 30% across multiple pots. Distinct moisture levels were successfully maintained according to prescribed setpoints, and stable regulation was achieved even under varying environmental conditions. Overall, the results demonstrate that optimizing the coordination between evaluation and prediction intervals provides a practical and robust framework for precision soil moisture control in IoT-enabled irrigation systems, with potential applicability across diverse crops and production environments.