Enabling Water Management System for Agriculture Using a Low Cost Approach

• Iago Magalhães De Mesquita UFC
• Sarah Frota Alves IFCE
• Rhuan Silva Nunes UFC
• Leonardo Tabosa Albuquerque IFCE
• Francisco Aldinei Perreira Aragão IFCE
• Larissa Ferreira Rodrigues Moreira UFV
• Iális Cavalcante De Paula Júnior UFC

Abstract

Agriculture is one of Brazil's primary sources of income and has grown steadily over the years. Despite this continuous growth, the sector faces persistent challenges related to water supply and climate change, which are expected to intensify over time, exacerbating farmers' problems. This paper presents the development of a low-cost system that leverages Internet of Things (IoT) and Machine Learning (ML) technologies to address these challenges. The system forecasts soil moisture for up to 24 hours in advance, enabling the estimation of water required for the irrigation of fruit crops. The experimental results highlight the effectiveness of integrating IoT and ML for precise water management, with the Long Short-Term Memory (LSTM) algorithm demonstrating the best performance, achieving a Mean Squared Error (MSE) of 11.9487, a Root Mean Squared Error (RMSE) of 3.4567, and an R2 Score of 99%. The proposed approach improves irrigation efficiency and provides a scalable and cost-effective solution that empowers farmers to boost agricultural productivity and sustainability.

Keywords: Agriculture, Machine Learning, Internet of Things, Low cost, Water management system