AI-Powered Air Quality Prediction Using IoT Sensor Networks: A Case Study from Armenia

Armenian State University of Economics, Yerevan, Armenia

^* Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

Artificial Intelligence (AI) and the Internet of Things (IoT) are transforming environmental monitoring by enabling intelligent, real-time analysis and prediction of air quality. Air pollution remains a major environmental challenge that significantly impacts public health, urban sustainability, and economic development. This study presents an AI-powered air quality prediction framework that integrates real-time Internet of Things (IoT) sensor data with advanced machine learning and deep learning models to forecast concentrations of key pollutants, including PM2.5, PM10, NO₂, and CO. A distributed network of low-cost IoT sensors was deployed across urban areas of Armenia, enabling continuous monitoring and high-resolution data collection. The collected datasets underwent rigorous pre-processing, including missing value imputation, normalization, and feature engineering, to improve the quality and predictive power of the models. Multiple machine learning algorithms—including Random Forest, Gradient Boosting, Long Short-Term Memory (LSTM) networks, and Temporal Convolutional Networks (TCN)—were trained and evaluated using standard performance metrics. Results indicate that deep learning models, particularly LSTM networks, outperform traditional algorithms in short-term forecasting accuracy. The proposed system delivers precise, location-specific air quality predictions that can assist public health authorities, environmental protection agencies, and smart city management platforms in designing proactive pollution mitigation measures and optimizing urban planning strategies. By integrating Internet of Things (IoT) infrastructure with Artificial Intelligence (AI)-based predictive analytics, the proposed framework provides a scalable intelligent decision-support platform for real-time air quality forecasting. The developed architecture demonstrates the effectiveness of deep learning models for processing heterogeneous sensor data and supports deployment within smart city infrastructures, digital monitoring systems, and intelligent environmental information platforms.