IoT based monitoring of air quality and traffic using regression analysis

Dynamic traffic management (DTM) systems are used to reduce the negative externalities of congestion, such as air pollution in urban areas. They require and environmental monitoring infrastructures. In this paper we present a prototype low-cost Internet Things (IoT) system for flow Air Quality Index (AQI). The computation flows is based on processing video compressed domain. Only using motion vectors input, computed real-time over an embedded architecture. An estimation AQI supported by machine learning regression techniques, different feature data obtained from IoT device. These automatic techniques overcome need complex calibration other limitations devices making needed measurements pollutant gases actual AQI. experimentation with cities representing scenarios variety climate conditions, allows validating proposed As regressors, Linear Regression (LR), Gaussian Process (GPR) Random Forest (RF) compared performance metrics R2, MSE, MAE MRE resulting relevant improvement estimations our proposal.