IOT Based – Automated Indoor Air Quality and LPG Leak Detection Control System using Support Vector Machine

Abstract

The rapid increase of population every year causes a great distress in the environment. Air pollution is becoming more imminent due to emissions from transportation, urbanization, and energy consumption. Hence, harmful elements contaminate the outdoor air, which then affects the health of individuals. These contaminants are the CH4, CO, CO2, and particulate matter, which also affect the indoor air quality. The indoor air quality was said to be twice to five times more dangerous than the outdoor air due to harmful substances entering and being trapped. LPG leakage is also considered as an indoor health risk. The study therefore aims to create a device to not only to monitor the different parameters of indoor air quality (IAQ), but also to notify the user about the potential risks in the specific location. The control system utilizes wireless transmission of data and a mobile application to notify the user. The sensor system also utilized a supervised machine learning using Support Vector Machines which helped the system to try and predict new data as accurate as possible. This study was able to accurately monitor the IAQ of a given location notify the user. Careful reading of sensors revealed percent differences ranging from 0.8% to 4% compared to an industrial grade air quality detector. The researchers also found a variety of opportunities for improvement within the design of the prototype.