Real-Time Intelligent Monitoring of Outdoor Air Quality in an Urban Environment Using IoT and Machine Learning Algorithms

The monitoring and prediction of air quality (AQ) is key to minimizing the negative impact of air pollution, as it enables the implementation of corrective measures. An IoT-based multi-purpose monitoring system has therefore been designed. To develop a reliable remote system, this study addresses three challenges: (1) design of a low-cost compact, robust, multi-sensor system, (2) model validation over several months to ensure accurate detection, and (3) the application of machine learning (ML) techniques to classify and predict AQ. The developed system demonstrates a significant cost reduction for regular monitoring, including effective data management under harsh environmental conditions. The prototype integrates pollutant sensors, as well as the detection of liquified petroleum gas, humidity, and temperature. A dataset with more than 30,000 entries per month (data recorded approximately every minute) was saved on the platform. Results identified the three highest pollution categories, highlighting the urgency of addressing AQ in densely populated regions. The ML algorithms allowed us to predict AQ trends with 99.97% accuracy. To summarize, by reducing monitoring costs and enabling large-scale data management, this system offers an effective solution for real-time environmental monitoring. It also highlights the potential of artificial intelligence-based AQ predictions in supporting public health initiatives. This is particularly interesting for developing countries, where pollution control is limited. Future research will develop the models to include data from different environments and seasons, exploring its integration into mobile apps and cloud platforms for real-time monitoring.