Real-Time Air Quality Monitoring And Forecasting System Using Fog Computing Technology

Abstract

The issue of air pollution has emerged as a significant concern in the 21st century due to its detrimental impact on the natural environment and human well-being. AIoT, which integrates artificial intelligence (AI) and Internet of Things (IoT) technology, is a highly efficient approach for assessing, analyzing, and forecasting air quality. This has a significant impact on addressing the issue of environmental pollution. Despite extensive research, numerous problems persist in implementing this AIoT technology. In this work, we suggest employing fog computing technology to actively monitor and forecast real-time indicators of environmental contamination to address current issues. The components of our proposed system use the STM32F429ZIT6 microcontroller, WiFi module ESP8266 NodeMCU, and other affordable sensors. These sensors gather data on airborne pollutants, specifically PM2.5, CO2, CO, UV index, temperature, and humidity. The gathered data is sent to ThingsBoard, an open-source fog computing system running on a Raspberry Pi 4-embedded computer. Machine Learning models and algorithms, such as Long Short-term memory (LSTM), Linear Regression (LR), Gradient Boosting (GB), and Extreme-Gradient Boosting (XGB) are employed for air quality forecasting. After thoroughly comparing the model and the standard regression technique, it has been determined that the Long-short Term Memory (LSTM) algorithm exhibits the lowest error and the most accurate prediction outcomes. The system’s monitoring and predicting outcomes will furnish researchers and the community with data to facilitate decision-making, intervention, and resolution of prevailing environmental pollution issues.