Design and Development of IoT Based Intelligent Agriculture Management System in Greenhouse Environment 
C. Kishore Kumar1, V. Venkatesh2
1C.Kishore Kumar, Department of Embedded Systems, SASTRA Deemed to be University, Thanjavur (Tamil Nadu), India.
2V.Venkatesh, Department of CSE, SASTRA University, Thanjavur (Tamil Nadu), India.
Manuscript received on 25 August 2019 | Revised Manuscript received on 01 September 2019 | Manuscript Published on 14 September 2019 | PP: 47-52 | Volume-8 Issue-5S3, July 2019 | Retrieval Number: E10130785S319/19©BEIESP | DOI: 10.35940/ijeat.E1013.0785S319
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: The introduction of Internet of Things (IoT) has a significant impact on shaping the communication and internetworking landscapes. The upcoming IoT researches are linked with design of standards and open architectures still requiring a global attention before deployment. The main objective is to design and develop a framework on Internet of Things (IoT) for precision agriculture using Machine learning techniques, where it surges the efficiency in farming by minimizing the loss of water and studying the fertility of the field. Libelium Smart Agriculture is used to connect to the IoT which uses Waspmote module. Waspmote is the plug and sense platform which is programmed using Waspmote IDE configured to connect with the available Local Area Network (LAN). With the help of Machine learning techniques like Classification And Regression Technique (CART) and Linear Support Vector Machine (SVM), the amount of water required by the crops can be estimated. In this paper, various regression such as stochastic gradient decent and boosted tree regression techniques are compared and results were obtained. Although each model applied in this paper performed well in predicting whether the crop needs to be irrigated, the optimal prediction accuracies were acquired by Boosted Tree Regression (BTC). It is compared by the fold numbers, Root Mean Squared Error (RMSE) and coefficient of Determination (CoD). The accuracy of the boosted tree regression came out to be 91.93% and the stochastic gradient descent prediction model delivered 62.95% accuracy. The amount of water required for the irrigation is then sent to appropriate actuator like solenoid valve and motor can be turned on for that particular period of time. Calibration test results and Measurements are represented to enhance the accuracy and success rates of Precision Agriculture (PA).
Keywords: Decision Support System (DSS), Libelium, Machine Learning, Smart Agriculture.
Scope of the Article: Learning Software Design Engineering