Worldwide, various air quality monitoring systems collect a huge amount of relevant data for the local air quality. The aggregation and further analysis of such data allow us to better understand air quality-related problems, to find anomalies and to identify new trends and correlations. In order to exploit the full potential of data, the analysis should utilize integrated datasets from multiple sources. However, most of the current air quality platforms support data acquisition only from a single source. At the same time, the diversity of air quality data and the specifics of time series data itself, bring additional challenges related to data pre-processing, analytics and visualization. The development of more comprehensive services, which fully utilize the variability of air quality data, requires a harmonized, multi-channel data acquisition and its seamless aggregation. This paper addresses these challenges and needs by proposing an air quality platform for monitoring, collecting and aggregating data from multiple sources and in different formats. The platform is able to compare measurements of air quality parameters from different data sources. Furthermore, it supports monitoring of sensors themselves, validation of their measurements and ignoring those in which data anomalies are spotted. The paper presents a detailed overview of the platform design, including its multi-layered architecture, data model, functional design and logical packages. The key implementation points are also introduced, such as the three data collectors realized for the integration of external APIs, provided by, and The extensive validation of the platform that is also presented in the paper employs a vast amount of diverse data collected from two cities through over 1.2 million measurements, including 3.5 million data points obtained by 880 sensors from 500 locations.


Journal of Energy Sources, Part A: Recovery, Utilization, and Environmental Effects. Publisher Taylor & Francis