Analysis and pattern identification on smart sensors data

Abstract

This work exemplifies the use of a data analysis technique applied to indoor air quality data obtained in a laboratory. The environment data is acquired with a wireless sensor system, NSensor. The sensing system, developed at the Faculty of Engineering, University of Porto (FEUP), is used for indoor environment monitoring, with the capability to store, in a remotely accessed database, air quality parameters such as temperature, relative humidity, pressure, illuminance, carbon dioxide and volatile organic components. For the current study, it was selected the data from temperature and relative humidity, and a period of ten months was considered. The data analysis uses Fourier transforms to identify patterns on the acquired data. For the temperature data, five main patterns were possible to identify. This work explores the potential of using data analyses techniques for big data on the field of indoor air quality evaluation. To make use of this data, further developments must be carried out so that would be possible to go from the monitoring and identification to the phase of controlling the indoor environment.