A multidimensional human-centric framework for environmental intelligence: Air pollution and noise in smart cities

Abstract

For the important problem of increasing levels of air pollution and noise in urban and rural areas, we propose a holistic, multi-dimensional approach to gather, monitor and analyze heterogeneous data sources of air pollutants and noise indicators, into an integrated, intelligent computational system. Although several interesting approaches have been developed for monitoring pollution and noise, however the challenge remains for even more detailed, precise, large scale monitoring.To overcome the limitations of current systems, we envision an integrated approach to human-centric environmental intelligence, bringing together modern IoT technology and the human factor. In particular, our approach emphasizes selected behavioural and health aspects, and the complementary use of sensing technology with citizen engagement and crowdsourcing methods. The proposed system will collect diverse data from heterogeneous sources, such as mobile and static wireless sensor networks, crowdsourcing, citizen questionnaires and social media analytics, to continuously combine objective estimations with subjective perception of air quality and noise. With the use of advanced AI and Deep Learning algorithms, our system will be able to estimate air pollutants concentration and noise levels in micro-scale with adequate precision over large urban-scale environments. Furthermore, tracking of behavioral and psychological users’ input, as well as personal exposure to pollution, will allow studying the impact of air quality and noise on the users’ daily habits and the interplay of ambient conditions with behavioural factors, towards an active engagement of citizens in a hybrid techno-social manner. A reference architecture for the realization of this human-centric environmental intelligence approach is presented. Also, a planned implementation at the city of Patras, Greece is discussed. To the best of our knowledge, this is one of the first holistic, multifaceted approaches to a surveillance system for air quality and noise in urban areas.