Automatic Crop Irrigation System using Internet of Things and Machine Learning

Excessive irrigation causes adverse effects on the physical, chemical, and biological properties of soil, resulting in a decrease in its fertility. Therefore, the appropriate selection of irrigation methods is very important. Among these, furrow irrigation is found to be the most suitable alternative because of its high efficiency and low water consumption. The present work deals with the design, development, and simulation studies of a low-cost, fully automated, crop-independent, IoT-based furrow irrigation system. The system is designed by placing an array of wireless sensor nodes to investigate an adequate quantity of furrow flow rate and optimize it as per the requirements using a cloud computing platform. The designed system is advantageous in terms of saving water, benefiting crops, and subsiding soil degradation. The moisture sensors were deployed in the crop field with the help of a signal conditioning circuit interfaced with a microcontroller-integrated Wi-Fi module (ESP8266-NodeMCU). The said module was also connected to a solenoid valve to regulate the inflow rate and the Thing Speak cloud computing platform. The WinSRFR 5.1.1 simulation tool was used to estimate the adequate amount of flow rate suitable for the crop and soil type. The cloud platform was used to acquire, store, and process data to forecast the ideal soil moisture levels for future crop development and soil health. The studies were carried out at a depth of 100 mm and by varying furrow length (50 to 150 m) in steps of 50m for four different soil types, namely coarse sand soil, sandy loam soil, silt loam soil, and black cotton-plowed soil, which are commonly observed in India. The simulation work in this paper indicates the effectiveness of this system that how we can reduce the water loss by low-cost, fully automated, crop-independent, IoT-based furrow irrigation system.