IoT Assisted Farming using ML techniques

Abstract

IoT has brought a new vision of smartness into every technology. Smart agriculture is one of the emerging applications of IoT which can add to the economic growth of any country. Irrigation is an important part of agriculture and automated irrigation helps the farmer to monitor the requirement of water remotely. The objective of this work is to develop a system to monitor the water requirement of the crops accurately and irrigation prediction with the assistance of Machine Learning Algorithms. In this paper, a NodeMCU based Smart Farming System is designed using Blynk and ThingSpeak cloud for an automated irrigation system. Statistics of different parameters such as temperature, rain, moisture content in air as well as soil and motion are collected using ThingSpeak Cloud using the NodeMCU. A Set of Machine Learning (ML) models like Decision Tree, K-Nearest Neighbour,Random Forest and Logistic Regression models has being utilized to analyze the data to predict irrigation with high accuracy. Among all the models it is observed that the Logistic Regression Model gives an accuracy of 99.69 %,a Precision of 98.95%,and 100 % Recall.