International Journal of
Advanced Multidisciplinary Application

Volume 3, Issue 7

• Author(s)Anjali R. Deshpande, Manish K. Tiwari, Sneha P. Joshi, Rohit N. Kulkarni
• AffiliationDepartment of Civil and Environmental Engineering, Shri Vaishnav Institute of Technology, Indore, Madhya Pradesh, India
• Page No.28-34
• Volume, Issue & YearVolume 3, Issue 7, July 2026
• Published On2026/07/04
• JournalInternational Journal of Advanced Multidisciplinary Application (IJAMA)
• ISSN No.3048-9350

Abstract

Air pollution and its quantifiable health burden remain among the most pressing multidisciplinary challenges confronting rapidly urbanising Tier-2 Indian cities, where municipal monitoring infrastructure is typically sparse relative to both the spatial heterogeneity of pollution sources and the scale of exposed population. This study presents an integrated framework combining a low-cost Internet of Things (IoT) sensor network for real-time criteria pollutant monitoring with a machine learning pipeline for both air quality index (AQI) forecasting and population-level health risk classification, deployed across five representative monitoring zones (industrial, traffic corridor, residential, institutional, and peri-urban) in Indore, Madhya Pradesh, over a twelve-month observation period (January-December 2025). A network of 42 low-cost nodes equipped with electrochemical and optical sensors (PM2.5, PM10, NO2, SO2, CO) transmitted hourly readings via LoRaWAN to a cloud-based aggregation layer, where the data were fused with meteorological variables (temperature, relative humidity, wind speed) and ward-level traffic density estimates. Five machine learning architectures — Support Vector Regression (SVR), Random Forest, XGBoost, Long Short-Term Memory (LSTM) networks, and a hybrid CNN-LSTM model — were benchmarked for 24-hour-ahead AQI forecasting, and a four-class Random Forest classifier was trained to stratify population exposure-days into Low, Moderate, High, and Severe health risk categories using a composite index derived from WHO air quality guideline thresholds and ICMR-NIOH dose-response coefficients. The hybrid CNN-LSTM model achieved the best forecasting performance (R²=0.918, RMSE=12.3 µg/m³), outperforming the XGBoost (R²=0.871) and Random Forest (R²=0.834) baselines, while the four-class health risk classifier attained an overall accuracy of 91.4% on held-out data. PM10 and NO2 emerged as the most predictive features (relative importance 0.241 and 0.183 respectively), and the industrial and traffic-corridor zones together accounted for 63% and 55% of high-and-severe-risk exposure-days respectively, compared to 14% in the peri-urban reference zone. The findings support targeted, zone-specific intervention prioritisation over uniform city-wide air quality management strategies and demonstrate the technical feasibility of low-cost IoT-ML pipelines for real-time environmental health surveillance in resource-constrained Tier-2 urban administrations.

Keywords: air quality monitoring, Internet of Things, low-cost sensors, machine learning, AQI forecasting, health risk classification, LoRaWAN, urban air pollution, Tier-2 cities, environmental health surveillance

References

1. [1] Central Pollution Control Board. (2023). National Air Quality Index: Methodology and Implementation Report. Ministry of Environment, Forest and Climate Change, Government of India.
2. [2] Chowdhury, S., Dey, S., & Smith, K. R. (2018). Ambient PM2.5 exposure and expected premature mortality in India. Environment International, 121, 1183-1192.
3. [3] Gupta, P., & Christopher, S. A. (2019). Particulate matter air quality assessment using integrated surface, satellite, and meteorological products. Atmospheric Environment, 142, 233-249.
4. [4] Jain, S., Sharma, S. K., Choudhary, N., & Mandal, T. K. (2021). Low-cost sensor performance evaluation for air quality monitoring in Indian cities. Atmospheric Pollution Research, 12(4), 296-307.
5. [5] Kumar, A., Patil, R. S., & Dikshit, A. K. (2020). Real-time air quality monitoring using low-cost sensor networks in urban India. Journal of Environmental Management, 264, 110491.
6. [6] Liu, H., Chen, C., & Tsai, F. (2022). CNN-LSTM hybrid architectures for short-term air quality forecasting: A comparative review. Atmospheric Environment, 277, 119069.
7. [7] Mukherjee, A., Brown, S. G., & McCarthy, M. C. (2019). Statistical evaluation of low-cost sensors for ambient air monitoring. Atmospheric Measurement Techniques, 12, 5099-5111.
8. [8] ICMR-National Institute of Occupational Health. (2022). Dose-Response Assessment of Ambient Air Pollutants and Respiratory Morbidity in Indian Urban Populations. Indian Council of Medical Research.
9. [9] Rajak, R., & Chattopadhyay, A. (2020). Short and long-term exposure to ambient air pollution and impact on health in India: A systematic review. International Journal of Environmental Health Research, 30(6), 593-617.
10. [10] Reddy, B. S. K., Kumar, K. R., & Balakrishnaiah, G. (2021). Recent trends and source identification of urban air pollutants using IoT-enabled sensor networks. Environmental Science and Pollution Research, 28, 24578-24594.
11. [11] World Health Organization. (2021). WHO Global Air Quality Guidelines: Particulate Matter, Ozone, Nitrogen Dioxide, Sulfur Dioxide and Carbon Monoxide. World Health Organization, Geneva.
12. [12] Zheng, Y., Yi, X., Li, M., Li, R., Shan, Z., Chang, E., & Li, T. (2015). Forecasting fine-grained air quality based on big data. Proceedings of the 21st ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2267-2276.
13. [13] Zimmerman, N., Presto, A. A., Kumar, S. P. N., Gu, J., Hauryliuk, A., Robinson, E. S., & Subramanian, R. (2018). A machine learning calibration model using random forests to improve sensor performance for lower-cost air quality monitoring. Atmospheric Measurement Techniques, 11, 291-313.

Explore Our Related Journals

Looking for the right journal for your next manuscript? Explore our international peer-reviewed journals covering engineering, management, computer science, artificial intelligence and multidisciplinary research.