A review on a machine learning approach of an intelligent irrigation monitoring system with edge computing and the internet of things

Abstract

Abstract Water consumption during irrigation has been a much-researched area in agricultural activities, and due to the frugal nature of different practiced irrigation systems, quite a sufficient amount of water is wasted. As a result, intelligent systems have been designed to integrate water-saving techniques and climatic data collection to improve irrigation. An innovative decision-making system was developed that used Ontology to make 50% of the decision while sensor values make the remaining 50%. Collectively, the system bases its decision on a KNN machine learning algorithm for irrigation scheduling. It also uses two different database servers, an edge and an IoT server, along with a GSM module to reduce the burden of the data transmission while also reducing the latency rate. With this method, the sensors could trace and analyze the data within the network using the edge server before transferring it to the IoT server for future watering requirements. The water-saving technique ensured that the crops obtained the required amount of water to ensure crop growth and prevent the soil from reaching its wilting point. Furthermore, the reduced irrigation water also limits the potential runoff events. The results were displayed using an android application.