Design and Implementation of Autonomous Irrigation System Using IoT and Artificial Intelligence

Abstract

Agriculture plays a vital role in Nigeria’s economy contributing about 23 percent to the total GDP which makes it a key activity after oil. To improve crop productivity, optimum use of scarce water resources and the ability to predict rainfall is crucial in this era. The adoption of emerging technologies to automate irrigation processes is necessary to improve crop productivity. An intelligent irrigation system can be achieved through the synergy between Internet of Things (IoT) and Machine Learning (ML) algorithms, which support automation and improved efficiency in the agricultural field. In this proposed system, sensors will be deployed to collect air temperature, soil moisture, and air humidity. These raw data will be transmitted to NodeMCU to know if the soil is dry by comparing it to a trained set of data in the NodeMCU. The NodeMCU will then obtain data from the Opensource Google weather API (Application Programming Interface), OpenWeatherMap using JSON (JavaScript Object Notation) data to know if there will be rainfall in less than an hour or not. Base on the data collected from the weather API by the NodeMCU, it will then determine whether to trigger the pump for irrigation or not. If the soil is dry and rain will likely fall in less than an hour, the pump remains OFF and allow the rain to water the field, otherwise, the pump is triggered ON and a calculated amount of water is pumped into the soil. The real-time monitoring report of this process will be sent to the cloud where the farmer can access to monitor the condition of the farmland.