Introduction to Machine Learning and Internet of Things for Management in Agriculture

Abstract

Present day agriculture requires innovative approach and practices to help farmers to increase the production efficiency with simultaneous reduction in required amount of natural resources. Machine learning (ML) tools and internet of things (IoT) have great scope to modernise the agriculture system to smart one. In this paper, we discusses fundamental of machine learning tools and IoTs along wth their potential applications to make smart agricultural production system. Few commonly used machine learning tools are multiple linear regression (MLR), random forest regression (RFR), support vector machine (SVM), artificial neural network (ANN) etc. Using these machine learning tools robust predictive models are developed using large database on past experiences, which is otherwise not feasible with conventional computation algorithms. Possible ML based applications in agriculture are digital soil maps, predictive models for yield prediction, identification of pest and disease infestation through image analysis, identification of weeds through hyperspectral image analysis, species recognition through digital image processing of leaves and floral structures, soil moisture prediction through modelling water balance components etc. IoT is an embedded system with sensors, software, actuators, electronics and computer to connect and exchange data. Few of the most commonly used IoT applications in agriculture are IoT based operations of agricultural machineries, smart irrigation system, IoT smart greenhouse etc. One of the big challenge of wide application of IoTs in agriculture is interference across networksespecially with the IoT devices using the unlicensed spectrum, such as ZigBee, Wi-Fi, Sigfox, and LoRa. However, in recent times with digital India mission in the country and with requirement of increasing water and other input use efficiency in agriculture, ML and IoT based smart agriculture system has a great scope to improve agricultural productivity. This not only will fetch higher income to farmers but will make the agricultural production system as a commercial and profitable venture specifically to the educated youth of the country.