Real-time AIoT platform for monitoring and prediction of air quality in Southwestern Morocco

Abstract

Urbanization and industrialization have led to a significant increase in air pollution, posing a severe environmental and public health threat. Accurate forecasting of air quality is crucial for policymakers to implement effective interventions. This study presents a novel AIoT platform specifically designed for PM2.5 monitoring in Southwestern Morocco. The platform utilizes low-cost sensors to collect air quality data, transmitted via WiFi/3G for analysis and prediction on a central server. We focused on identifying optimal features for PM2.5 prediction using Minimum Redundancy Maximum Relevance (mRMR) and LightGBM Recursive Feature Elimination (LightGBM-RFE) techniques. Furthermore, Bayesian optimization was employed to fine-tune hyperparameters of popular machine learning models for the most accurate PM2.5 concentration forecasts. Model performance was evaluated using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and the coefficient of determination (R2). Our results demonstrate that the LightGBM model achieved superior performance in PM2.5 prediction, with a significant reduction in RMSE compared to other evaluated models. This study highlights the potential of AIoT platforms coupled with advanced feature selection and hyperparameter optimization for effective air quality monitoring and forecasting.

Fig 1. Workflow diagram of the IoT monitoring system.

Fig 2. Schematic diagram of the wireless environmental monitoring device integrating PMS5003, DHT22, and MQ135 sensors with NodeMCU V3 (ESP8266).

Fig 3. Location of the experimental facility in Ait Melloul.

Map data (C) OpenStreetMap contributors, under the Open Database License (ODbL). This map was created using the folium library in Python.

Fig 4. Temporal fluctuations of pollution and meteorological parameters.

The data were observed at station 1 during the initial days of October.

Fig 5. Wind rose of the wind speed and direction in Ait Melloul.

Map data (C) OpenStreetMap contributors, under the Open Database License (ODbL). This map was created using the folium library in Python.

Fig 6. Hourly Concentration of PM_2.5 in Ait Melloul.

The figure depicts the hourly concentrations of PM_2.5 at two monitoring sites: S_RH (blue) and S_ZI (red).

Fig 7. Average hourly concentration of PM_2.5 at two monitoring stations in Ait Melloul.

Fig 8. Correlation between PM_2.5 levels and environmental parameters at two monitoring sites.

Fig 9. Boxplot of hourly mean air pollution levels in S1 and S2.

Fig 10. Comparison of imputed data: Actual vs. Imputed dataset in S1 and S2.

Fig 11. Observed and predicted PM_2.5 concentrations.

Fig 12. Fit curve of PM_2.5 real value and predicted value using the LightGBM model.

Fig 13. Feature ranking using the SHAP values for the LightGBM model.

Fig 14. Feature importance plot derived from the LightGBM model.