Monitoring of Soil Humidity and Temperature using IoT and AI for Remote Management

Abstract

Efficient environmental monitoring and irrigation management are critical challenges in tropical oil palm plantations due to high humidity, temperature variability, and large cultivation areas. This study presents an Internet of Things (IoT)–based monitoring and decision-support system integrated with a lightweight linear regression model to optimize plantation management. This study presents an IoT-based monitoring and decision-support system for tropical oil palm plantations that integrates calibrated environmental sensors with a lightweight linear regression model, enabling real-time irrigation management. Deployed over three months on a 50-hectare plantation, the system achieved a 98.7% data transmission success rate and strong predictive performance (R² = 0.89, MSE = 0.45), delivering below-threshold humidity notifications with 92–94% accuracy and an average latency of 4.3 seconds via an Android application. Field results demonstrate that data-driven irrigation reduced water usage by 23%, increased fresh fruit bunch (FFB) yield by 12%, and lowered manual inspection and labor costs, confirming the system’s effectiveness, scalability, and suitability for sustainable plantation management in tropical environments.