IoT-Enabled Digital Twin for Multi-Environment Air Quality and Health Risk Assessment

Abstract

Air pollution in industrial regions poses significant risks to human health, necessitating advanced monitoring and analytical frameworks. This paper presents an IoT-enabled digital twin system for real-time comparative analysis of air quality between industrial and rural environments. The proposed system integrates multi-sensor data acquisition using an ESP32 micro-controller with cloud-based data transmission through the ThingSpeak platform, enabling continuous environmental monitoring. A multi-sensor framework directly measures CO, TVOC, eCO2, temperature and humidity, while relative NO2 and SO2 indices are estimated using calibration equations. The collected data is processed using a machine learning-based digital twin model, developed using a Random Forest classifier, to distinguish between industrial and rural pollution conditions. To quantify the impact of pollution exposure, a composite pollution index is formulated using weighted contributions of different gases. This index is further utilized to estimate a Relative Health Impact Index (RHI). The digital twin synchronizes real-time environmental data and enables comparative analysis between industrial and rural environments. Experimental results from two 30-day field deployments demonstrate substantially higher pollution levels and relative health impact in industrial regions than in rural areas. The proposed system provides an effective and scalable solution for real-time environmental monitoring, predictive analysis, and decision support in air quality management. Environmental data were collected from both industrial and rural deployments over 30 consecutive days.