Artificial Intelligence-Based Monitoring and Forecasting of Urban Air Pollution in Smart Cities

Abstract

Urban air pollution greatly threatens both the health of people and the natural environment, especially in fast-growing smart cities. The study looks at how AI helps trace and predict air pollution using data from urban sensors. Researchers applied four algorithms—Decision Tree, Random Forest, Support Vector Machine and Artificial Neural Network—to determine levels of pollutants PM2.5, PM10, NO₂ and CO. The models we made rely on observations and measurements from monitoring stations that are part of the city’s air quality network. Experiments showed that AI models predicted pollution exceptionally well. The best accuracy, according to the models, is found in ANN at 94.8%, RF at 92.5%, SVM at 88.3% and DT at 85.6%. The results show that many AI tools, including deep learning, assist urban planners and those responsible for early warning systems. Ongoing testing and comparisons with past work show that AI builds better results. The research greatly benefits smart cities by enabling clear management of environmental issues with the use of smart technology.