Design and Implementation of an IoT-Based Indoor Air Quality Monitor with Ensemble and Unscented Kalman Filter Forecasting

Abstract

The increase in air pollution resulting from intensified human activity in urban areas poses a serious threat to public health and environmental sustainability. In the long term, this situation also has the potential to reduce productivity, as evidenced by the spread of respiratory diseases affecting humans. Exposure to air pollution occurs not only outdoors but also indoors. This can happened due to ignorance of hygiene or exposure to outdoor air. To address this issue, a device is required that can accurately and affordably detect changes in indoor air quality. This research developed an affordable Internet of Things (IoT) based monitoring system to address these issues. By utilizing the ESP32 microcontroller along with DHT11 and MQ135 sensors, this device is capable of measuring ambient temperature, humidity, and carbon monoxide (CO) levels. The measurement results obtained by the device are then visualized using the real-time application ThingSpeak to identify changes in air quality over time. In addition, this research also employs analytical and forecasting approaches to the air quality measurement data utilizing the Ensemble Kalman Filter (EnKF) and Unscented Kalman Filter (UKF) methods. In addition, this research also applied analytical and short-term forecasting approaches to the concentration of CO, which is considered one of the air pollutants, using the Ensemble Kalman Filter (EnKF) and Unscented Kalman Filter (UKF) methods. The simulation results indicate that the EnKF method achieved the best forecasting error (RMSE) value of 0.19.