Classroom Automation Using RSSI

The surging energy costs has urged the need to minimize power consumption, especially in large education institutions where most of the power is consumed in providing well lit and air-conditioned classrooms, seminar halls, laboratories and corridors. A lot of power can be saved if we could control electrical appliances through automation, completely removing the possibility of a power loss, that is found to occur in cases where (i) classrooms are unoccupied, yet running with power, or (ii) occupancy of even a small part of the classroom involves all appliances to be turned on, due to a centralized power system. To overcome such cases of power losses, this paper pro-poses an efficient, cost effective and low-cost architecture model that consists of an occupancy monitoring system (subsystem I) integrated with a localized power system. To create such a low-cost system, the usage of Node MCU, a cheaply available microcontroller, as a wireless sensor network(WSN) node is experimented and tested for performing localization using the RSSI. The efficiency of NodeMCV as a sensor node, is measured by the accuracy of detection, which uses a mean algorithm, over a specified period of time. Additionally, to contribute to the automation of a classroom, a second subsystem is integrated with this model, which provides control over the door lock of every classroom. It also provides the means to indicate the status of occupancy in a room, to its clients, the teachers, and allots a free room if requested through an SMS message. Through repeated trials of experimentation, the proposed occupancy monitoring design is found to be successful in detecting the presence of a person(s) in Line of Sight(LOS), when limited up to 3 meters of detection region, with an accuracy of 95%(1 error every 20 min). Beyond this regions, the accuracy dips arbitrarily below an appreciable level due to multipath propagation. Secondly, results show that subsystem II functions smoothly with a maximum latency of 23s and an average latency of 14s before receiving the corresponding response action. Hence the proposed system is found to be a possible solution to save power and automate education institutions over certain limitations.