Visual Analytics for the Reduction of Air Pollution on Real-Time Data Derived from WSN

Nowadays, the contaminated and poor air quality that is inhaled by the city population in industrialized cities around the world has led to one of the main causes of premature death due to respiratory diseases. Therefore, the improvement of air quality becomes a priority. In this sense, the current study aimed to design and implement a visual analytics tool, in order to process large data sets, which have been generated by wireless sensor networks (WSN), which measured different environmental pollutants in real time. Hereby, the phases of the CRISP-DM methodology have been applied as a reference to guide the process. In the data preparation phase, programs have been implemented using Python. Then, the results have been stored in collections within a MongoDB database. Furthermore, for the modeling and visual exploration of the data, the Tableau tool has been used. The evaluation of the results allowed to demonstrate certain behavior of air pollutants around the city, such as the increased air pollution levels during daylight hours. Similarly, we discovered that the presence of particulate material PM10 and PM2.5 is directly related to the increase of the Air Quality Index for the city of Quito (IQCA). This leads to the conclusion that our analysis may be useful as a support tool in the decision-making of public policies for the reduction of air pollution.