Internet of things and machine learning based smart and intelligent irrigation system

Abstract

Water is wasted significantly in traditional irrigation systems. Not only is an intelligent irrigation system required to optimize water use, but it is also required to increase crop yield. The Internet of Things (IoT) and Machine Learning (ML) have enabled the development of intelligent systems capable of achieving these goals with minimal human intervention. This paper proposes an IoT-enabled and ML-trained irrigation system to optimize water usage while requiring minimal user intervention. IoT devices are used to collect soil and environmental data. In real time, this data is sent to and stored on a cloud server. From historical field data collected at the agricultural research site over a ten-year period, ML algorithms are used to generate a model. This model uses IoT sensor data to make real-time recommendations about the state of an agricultural field, such as the need for watering. Both simulation and prototype implementations are used to compare the performance of the proposed system to similar previous works. In addition to the features made available to users via a cloud platform called Thing Speak, the proposed system made better use of resources such as water. Our system reduced Garlic's Crop Water Requirement (CWR) by 6.45% and 6.72%, respectively, during the Initial and Development stages. The system can also predict the type of crop that should be planted in the current year based on the data collected. Longer-term agricultural field data would provide more insight into the area if it was analyzed with more performance evaluation parameters.