IoT based Irrigation and Fertilization framework for Smart Agriculture

Abstract

In recent years, Internet of Things (IoT) has played a significant impact in transforming agricultural domain by virtue of which the concept of Smart Agriculture (SA) has emerged. This concept has helped in revolutionizing agricultural practices by automating, analyzing, and computing of various agricultural parameters. There is no denying in the fact that food demand has increased at an unprecedented pace due to accelerated population growth all over the world. Therefore, it is imperative to harness emerging technology like IoT in agriculture sector to ensure sustainable food production. In order to understand the significant role of IoT in agriculture, a comprehensive literature review has been performed based on the framed research questions. Various research articles have been identified using the inclusion exclusion criteria from renowned electronic databases. Research articles were categorized on the basis of six different phases of agriculture's life cycle, i.e., soil properties and sowing, irrigation and fertilization, growth and monitoring of plants, crop protection from various diseases and weeds, harvesting, and stubble management. In addition, some of the most prominent research articles published over the past few years in IoT domain have been discussed to understand the pragmatic approach of this term. Also various web portals and mobile applications catering the agricultural domain have been presented within the study. On the basis of extensive literature review, research gaps have been identified and objectives have been framed. To achieve the framed objectives a framework namely, Minimal Distribution of Water \& Fertilization based Decision Support System MDW&FbDSS has been proposed. The framework consists of two different hardware devices, i.e., Libelium’s Waspmote Plug and Sense device and NPK sensor mounted over Arduino Uno. Using first hardware device, dataset was collected on three different crops namely,Rice (paddy), Wheat, and Capsicum on various parameters, i.e., Soil moisture (Sm), Soil temperature (St), Atmospheric pressure (Ap), Humidity (H), and Luminosity (L). Whereas the second hardware device detected the existing presence of Nitrogen (N), Phosphorus (P), and Potassium (K) within the soil. The captured data on various parameters have been saved on cloud server for analysis.\newline For irrigation assessment, Time Series Analysis of collected datasets has been performed using Auto-Regressive Integrated Moving Average Model (ARIMA) and Long Short-Term Memory (LSTM) techniques. The proposed framework has been evaluated using different parameters, i.e., Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE). It has been observed that the results of ARIMA model performed better than LSTM for weekly forecasting. For fertigation assessment, the existing presence of various fertilizers, i.e., N, P, and K within the fields has been estimated. The proposed framework reflected accuracy, and it can be concluded that the existing framework can be used for other applications, such as irrigation scheduling, target fertilizer spray, etc. To digitize the process of agriculture, two applications, i.e., web portal www.smartfasal.in and mobile application Farm analyzer have been designed. These applications allow farmers to monitor and control agricultural activities remotely. The web application enables farmer to view real-timerecordings for different parameters captured on various crops. Whereas Farm analyzer provides the user a Graphical User Interface (GUI) to view real-time readings captured by various agricultural sensors, along with weather api readings. In addition, precise estimation of water for irrigation requirement can be determined in advance by selecting the desired crop and entering the area field. Lastly, the mobile application provides an interface to remotely turn on/off irrigation value for the fields. The work presented in this study has huge potential in transforming contemporary agricultural practices by enhancing yield productivity and conserving natural resources to maximum extent.