Smart classroom for better pedagogical process and effective energy utilisation

Smart classrooms are in essence a technology based enhanced classrooms. Several researches focusing on the subject of acoustical comfort in lecture halls and auditoriums have devised ways to achieve high quality acoustic conditions and eventually increase students’ learning outcomes. Bad acoustic quality with bad speech intelligibility would have a detrimental effect on students’ takeaway from the classroom. Not only classrooms, but bad noise levels in any environment is detrimental to the people who are part of the surroundings and may not work or comprehend to their best of abilities in such environments. The physical classroom occupies the student’s work space, hence there is a lot of potential for power saving when the volume of students present in classes is low or nil. Smart classroom automation does open up a scope for delivering appliance automation via automatic controls. Developing an intelligent control system based on microcontrollers and arduino will adhere to the concept of energy saving and providing a comfortable environment for effective teaching and learning. We can devise an algorithm that would consist of a blob detection module, a people counting module and a tracking module, with the modules’ output being used as an input to the decision making module which would control the room’s environment in an intelligent way. The proposed model would be a cost effective one, as most sensors cost a small amount, the cost of the model would be estimated around 1700 per kit attached to the main switch. The CNN-based model demonstrates superior performance with a training accuracy of 0.9202, validation accuracy of 0.9002, and test accuracy of 0.8976 for recognizing English alphabets and digits, outperforming other traditional machine learning models.