Internet of Things Enabled ML for Air Quality Assessment: Systematic Review

A severe and pervasive environmental problem that affects the entire planet is Air Pollution (AP). Numerous researchers have focused on these issues while keeping human health in mind. One of the best methods to educate people about major health issues and safeguard human health from air pollution is through information about air quality predictions. Air pollution is one of the most difficult environmental issues, and it affects many major cities. Real-time monitoring of pollution data can help local officials assess the present state of the city’s traffic and reach well-informed decisions. In order to correctly estimate the pollutant concentrations, it is necessary to have an early system for monitoring and measuring the amount of AP using the Air Quality Index. Incorporating Internet of Things (IoT)-based devices may significantly change the AQ forecast dynamically, hence enhancing the AQ forecast. The accuracy and cost of the Both the AP prediction that is mentioned and the evaluation of it using various known methodologies are quite low. While AP prediction is still permitted in some sectors, machine learning (ML) algorithm development is expanding swiftly and looking into almost all fields and applications. This paper highlights numerous ML research projects related to AP monitoring and prediction using IoT sensor data in the context of diverse cities. This paper also summarizes historical and current data based on AQ prediction models, methods, and methodologies, examines current research methodology, the advantages and disadvantages of AQ prediction, as well as the challenges related to real-time AQ monitoring and prediction.