IoT-Enabled Smart Robotic System for Greenhouse Management using Deep Learning Model with STS Approach

Abstract

A significant component of any country's Gross Domestic Product is made up of farming and agriculture. Utilizing IoT in agriculture and farming methods is essential as the global population is projected to reach around 9.6 billion by 2050. To meet such high demand, an improvisation and optimization of the current farming technologies is the need of the hour. Numerous researchers developed different application specific system for agriculture but less attention was paid towards critical aspects such as intelligence, modularity and human centric design. There is lacuna in existing developed system in the utilization of advanced technologies to their full potential. The agricultural sector wants autonomous systems that are smarter and more effective. Therefore, this research paper introduced the smart solution as an intelligent modular autonomous system with human-centric design approach for agricultural application. The developed system able to detect plant disease with more than 96% accuracy with the help of novel deep learning model designed with sharpening to Smoothening approach. The disease detection and classification results has been verified through confusion matrix method of evaluation. An intelligent robotic system has been developed to detect plant diseases using novel deep learning model and perform multiple functions like greenhouse monitoring, pesticide sprinkling etc. The robotic system has control over internet through web control system so that farmer can monitor greenhouse and control robot activity from remote place. This smart farming solution able to make farmer life simpler and perform difficult task like plant disease detection and pesticide sprinkling easily.