AI Based Air Quality Monitoring and Pollution Prediction: A Review of IoT Enabled Systems for Sustainable Environmental Management

Air pollution has emerged as a major environmental and public health concern, necessitating effective and continuous monitoring and accurate prediction of pollutant concentrations. The integration of the Internet of Things (IoT), Artificial Intelligence (AI), and Machine Learning (ML) offers promising solutions for real-time air quality monitoring, pollution forecasting, and proactive environmental management. This study examines the application of IoT-enabled sensor networks and AI-based predictive models in industrial and smart city environments through a secondary, desk-based review of three recent studies. The review analyses sensor-based data collection, machine learning algorithms, predictive performance, practical applications, and implementation challenges. The findings indicate that Long Short-Term Memory (LSTM) networks demonstrate strong predictive performance for time-dependent pollution data, while Random Forest models provide competitive results for specific pollutants, particularly particulate matter (PM2.5). The reviewed industrial application highlights the potential of AI-driven forecasting to support automated ventilation and early intervention, whereas the smart city application demonstrates the usefulness of predictive analytics in urban air quality management and evidence-based policymaking. Despite these benefits, challenges related to data quality, sensor calibration, scalability, system integration, and model interpretability remain significant. The study recommends improved sensor standardisation, hybrid modelling approaches, and Explainable AI techniques to enhance the reliability and practical adoption of these systems. It concludes that integrating IoT and AI can strengthen pollution control, protect public health, and support sustainable environmental management.