A Comprehensive IoT-Based Automation System for Enhanced Productivity and Sustainability for Advancing Farming Efficiency

Abstract

The main objective of this research is to investigate the application of an IoT System coupled with machine learning models. The objective is using sensor data in real-time to determine the amount of water irrigation needed for a farm. The research uses a real dataset of different farming conditions such as temperature, humidity, and water level and conducts several pre-processing methods to the data to improve the quality of the input data. These methods include data cleaning, feature selection, normalization, and time-series analysis. Four machine learning methods are used to train and test the dataset. The results demonstrate that the Artificial Neural Network model is the most effective model in predicting results. It provides the best precision, recall, F1 score, and AUC-ROC curve which all amount to 98.5%, 97.5%, 98.0%, and 98.9% respectively. The model is also used in real-time for predicting irrigating pump operations from sensors. The research also presents its contribution as it integrates many sensors with renewable energy, solar panels, to support sustainability. The research has also assisted framers in obtaining valuable data to less the amount of water required in farm irrigation and helps with improving different farming techniques.