Smart Irrigation System using Zigbee Technology and Machine Learning Techniques

Agriculture plays a massive role in India's economy with nearly 60% of its land still under cultivation. In such scenario, managing the scarce water is a critical issue in this era. Therefore, a smart irrigation mechanism is necessary to improve the crop productivity. Smart irrigation can be achieved by combining the potential of Internet of Things (IoT) with machine learning algorithm(s), which supports automation and fast forward enhancements in the agricultural field. In this paper, design and implementation of a Smart Irrigation System (SIS) that automates the irrigation process in an ideal and accurate manner using machine learning techniques (decision making) is detailed. In the first module (sensor), various parameters such as the moisture content of soil, temperature and humidity of air are obtained from the field and transmitted to a local base station through Zigbee technology (IEEE 802.15.4).Second module is the base station, where Raspberry pi is used for computation and learning. That is, a classification model is built using the training set and predicts the crop water requirement (ETcrop) for the new parameters. In addition, the sensor module is solar powered, thus reducing power consumption and enhancing the system's robustness.