Self Configurable Air Pollution Monitoring System Using IoT and Data Mining Techniques

Abstract

Air pollution is a perilous threat to living organisms and the whole ecosystem. The purpose of this paper is to develop a self-configurable air pollution monitoring system which can monitor and predict air pollution by applying Internet of Things (IoT) and data mining technologies. Self-configurability of the device is the ability to regulate the frequency of monitoring based on pollutant predictions. Monitoring is done using the system developed in paper [1] which collects concentration of pollutants such as Carbon monoxide, harmful gases, dust level, meteorological parameters such as temperature along with GPS location. This paper deals with using the monitored data for prediction along with humidity information. Data mining technique, Regression is used to predict the level of pollutant. These predicted values in turn decides the mode of operation of the device. The monitored data send to ThingSpeak are further analyzed using MATLAB. Map of the location is updated using red and green markers based on the level of pollution. These data along with predicted pollutant levels in ThingSpeak can be viewed by the public.