Abstract

The current urban pollution levels and the less extensive coverage of the air quality indexes by government stations have resulted in the establishment of networks for the monitoring of air contaminants at a low cost, as described in this paper. The paper discusses the application of the Air Quality Monitoring Network that is designed using AIoT technology involving the use of an ESP32 microcontroller, sensors for measuring the presence of gases and particulate matter in the air, and humidity and temperature sensors, as well as a simple interface for alerts. During the twenty-day operation of the network in the field, more than 1,967 hourly measurements of AQI indices have been captured, which were later analyzed and used for building a model for short-term prediction of AQI values. The analysis of the daily average data showed that there has been a significant rise in AQI value up to approximately 224 as compared to the subsequent decline of AQI down to approximately 93, while the hourly average data showed the reduction of AQI valued at nearly 175-188. The data obtained show that more than 60% of readings of air quality were poor while just 2.7% of all data constitute severe air quality indices. The model predicts. The findings show that the combination of a sensor node costing less than ₹3,200 and an easy time series technique makes it possible to generate actionable forecasts about pollution on the hyperlocal level to support scarcely available regulatory monitoring systems.

Keywords

How to Cite This Article

Rights