AI-POWERED SMART IRRIGATION SYSTEM FOR SMALLHOLDER FARMERS

This paper presents an integrated Artificial Intelligence (AI) system designed to optimise water use and enhance crop productivity for smallholder farmers in Malawi. The proposed system addresses persistent challenges in rain-fed agriculture, including inefficient water application, high labour demand, delayed response to soil conditions, and limited access to affordable technology. By combining Internet of Things (IoT) sensing, threshold-based machine learning control logic, and cloud-based remote monitoring, the system ensures precise irrigation, reduced labour, and improved climate resilience. The architecture integrates an ESP32 WROOM-32 microcontroller with a capacitive soil moisture sensor, a DS18B20 soil temperature sensor, and a DHT11 air temperature and humidity sensor. Irrigation is initiated when soil moisture falls below 30 %, halted when it exceeds 85 %, and extended by one minute when soil temperature surpasses 30 °C. Sensor data is transmitted via Wi-Fi to the Blynk IoT cloud platform, providing farmers with real-time monitoring and manual pump override through a mobile and web dashboard. Total hardware cost is approximately MWK 313 000—below many commercial alternatives. Experimental evaluation and design validation demonstrate improved water-use efficiency, reduced labour demands, and enhanced crop productivity potential compared to conventional manual irrigation practices. The integration of low-cost sensors with intelligent threshold logic not only strengthens resource conservation but also fosters climate resilience among smallholder farmers. The threshold-based control logic (irrigation starts below 30 % soil moisture, stops above 85 %, and extends by one minute when soil temperature exceeds 30 °C) eliminated both over- and under-irrigation. Continuous capacitive-soil-moisture, DS18B20 soil-temperature and DHT11 air-temperature/humidity sensing replaced farmer guesswork and fixed schedules, delivering water only when needed In conclusion, the proposed AI-driven framework offers a comprehensive, scalable, and locally adaptable solution for modern smallholder agriculture, contributing to safer food systems and the Malawi Vision 2063 pillars of agricultural productivity, as well as United Nations Sustainable Development Goals 2 (Zero Hunger) and 13 (Climate Action).