ASABE Technical Library - Abstract

Click on “Download PDF” for the PDF version or on the title for the HTML version.
If you are not an ASABE member or if your employer has not arranged for access to the full-text, view access and subscription options

AI-Driven IoT Architecture for Sustainable Solar Irrigation Pump Systems: Toward Energy-Aware and Water-Efficient Smart Agriculture

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  2026 ASABE Annual International Meeting  2600821.(doi:10.13031/aim.202600821)
Authors:   Muhammad Ashik-E-Rabbani, Md. Rasel Al Mamun, Md. Samiul Basir, Mofakkharul Islam Tonoy, Mohammad Al Fayed Sama Opi, Md. Bashirul Islam, Md. Mashurul Haque
Keywords:   AIoT, Precision agriculture, Random Forest, Smart irrigation, Solar energy.

Abstract. Agriculture is the largest consumer of global freshwater, and irrigation efficiency remains critical for food security in water- and energy-constrained regions such as Bangladesh. This study developed and evaluated an AIoT-enabled, solar-powered irrigation system that integrates real-time environmental sensing with machine learning for automated irrigation decision-making. Five solar-powered substations, each equipped with a capacitive soil moisture sensor and a DHT22 temperature-humidity sensor connected to an ESP8266 microcontroller, transmitted field data via the MQTT protocol to a solar-powered Raspberry Pi 5 main station. A Random Forest classifier trained on soil moisture, temperature, and humidity data predicted irrigation requirements and triggered a relay-controlled water pump accordingly. A Flask-based web application with an SQLite3 database and NGINX server enabled real-time monitoring and manual override. The main and substation battery systems provided 27.25 h and 24.42 h of power backup, with charging times of 4.85 h and 4.75 h, respectively. The Random Forest model achieved a precision of 99.6%, recall of 99.1%, F1-score of 99.4%, and ROC-AUC of 99.3%, with soil moisture identified as the most influential feature (79.1%). Field trials showed the system reliably switched irrigation ON at 46% ± 3% and OFF at 90% ± 2% soil moisture. These results demonstrate the potential of combining IoT, AI, and solar energy for scalable, low-cost precision irrigation in off-grid agricultural settings.

(Download PDF)    (Export to EndNotes)