ABSTRACT
The integration of monitoring technologies in the last decades has been a key factor in the development of new ways to track air pollutants and supplementing the network of traditional monitoring systems. In this regard, the appearance of affordable and accurate sensor devices to monitor air quality has made possible to obtain relevant data about the state of the air, and moreover, eminent institutions are interested in promoting the use of novel and more affordable tools for air pollution, such as the United States Environmental Protection Agency and European institutions, within a new approach to environmental surveillance, known as Next Generation Compliance and Enforcement technologies. On other hand, in order to get more reliable measurements, the use of machine learning to support adjustment or calibration process has been used in some studies to improve the performance of monitoring devices. On this paper, led by a group of specialists of the Chilean Superintendence of Environment (henceforth, SMA from its Spanish initials), a first approach case study related to the convenience of the usage of low-cost devices in environmental enforcement will be presented. The study was made in the Metropolitan Region of Santiago and considers the spatial distribution of different particulate matter sensors in the region. Some aspects regarding communication and technical issues are presented as well as the main findings about their performance. Results illustrate that low-cost sensors, aided by machine learning algorithms, could provide a reliable enough general screening of particulate matter within a large city, constituting a valuable decision-making tool for environmental oversight, as well as a powerful preventive and deterrent approach for compliance.
