Virtual Sensors: Toward High-Resolution Air Pollution Monitoring using AI and IoT

Abstract

The present article contributes a research vision for virtual sensing that combines Artificial Intelligence (AI) and Internet of Things (IoT) to increase the coverage of air quality information. Virtual sensors take advantage of correlations between different pollutants to estimate the concentrations of pollutants for which no affordable sensors are available. We cover key requirements and challenges, reflecting on the current state-of-the-art and identifying key research challenges. We also demonstrate the potential and feasibility of virtual sensing through experiments conducted with data from Helsinki, Finland, which show how standard <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{PM}_{2.5}$</tex> and temperature measurements can be used to provide reliable estimates of CO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> and black carbon concentrations. We also discuss potential applications that can benefit from the implementation of virtual air pollution sensors and establish a research roadmap for the path forward.