Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM68780
Volume:
10 |
11 |
Nov 2026
Pages:
1-13
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.
Computer Science And Engineering
D. Y. Patil Technical Campus Talsande, Kolhapur
Civil Engineering
D. Y. Patil Technical Campus Talsande, Kolhapur
Electrical Engineering
D. Y. Patil Technical Campus Talsande, Kolhapur
Computer Science And Engineering
D. Y. Patil Technical Campus Talsande, Kolhapur
Computer Science Engineering And Data Science
D. Y. Patil Technical Campus Talsande, Kolhapur
Computer Science Engineering And Data Science
D. Y. Patil Technical Campus Talsande, Kolhapur
Air pollution monitoring is constrained by a persistent tension between the accuracy of reference-grade instruments and the spatial and temporal coverage required to describe neighbourhood-scale exposure. Internet of Things (IoT) networks can increase measurement density, while artificial intelligence (AI) methods can calibrate sensor signals, combine heterogeneous observations, map pollutant fields and forecast concentrations. This narrative review examines how these functions connect, rather than treating sensor deployment and prediction as separate technical problems. Recent field studies show that low-cost sensors can support local monitoring when they are calibrated and evaluated for the intended task; they do not establish that inexpensive devices are interchangeable with regulatory instruments. Machine-learning calibration results vary by pollutant, sensor, meteorology and location, and calibration transfer beyond co-location sites remains a central concern. Deep learning can represent temporal and spatial dependencies, but model complexity alone does not guarantee better forecasts or reliable transfer. The strongest practical direction is therefore a quality-managed pipeline that preserves provenance, validates models across time and sites, reports uncertainty, and communicates whether an output is a measurement, calibrated estimate, or forecast. Research priorities include external validation in undermonitored settings, drift-aware calibration, comparable evaluation protocols, interpretable models, and decision-focused communication. The reviewed evidence supports AIenabled IoT as a complement to established monitoring infrastructure, not a substitute for reference measurements or regulatory oversight.
KEYWORDS
AI and IoT can strengthen air-pollution monitoring by joining distributed sensing with calibration, spatial inference, forecasting and accessible information services. The reviewed studies establish that machine learning can improve low-cost sensor estimates and that spatiotemporal models can produce useful local forecasts in particular deployments. They also show that performance varies with pollutant, sensor, location, validation design and forecast horizon. The strongest conclusion is therefore conditional: these systems can complement sparse reference networks when measurements are calibrated, model transfer is tested, uncertainty and data quality are visible, and outputs are matched to a defined decision. Future progress depends less on adopting the newest architecture than on reproducible field evaluation, long-term maintenance, transparent reporting and responsible communication. AI-enabled sensor networks should extend the evidence available to communities and planners while preserving the distinct role of reference measurements and regulatory monitoring.
[1] S. R. Iyer et al., “Modeling fine-grained spatio-temporal pollution maps with lowcost sensors,” npj Climate and Atmospheric Science, 2022, doi: 10.1038/s41612-02200293-z. [2] Y. Li et al., “From air quality sensors to sensor networks: Things we need to learn,” Sensors and Actuators B Chemical, 2021, doi: 10.1016/j.snb.2021.130958. [3] P. deSouza et al., “Calibrating networks of low-cost air quality sensors,” Atmospheric measurement techniques, 2022, doi: 10.5194/amt-15-6309-2022. [4] C. Heffernan, K. Koehler, D. R. Gentner, R. D. Peng, and A. Datta, “Unified calibration and spatial mapping of fine particulate matter data from multiple low-cost air pollution sensor networks in baltimore, maryland,” Environmetrics, 2026, doi: 10.1002/env.70100. [5] T. Veiga et al., “From a low-cost air quality sensor network to decision support services: Steps towards data calibration and service development,” Sensors, 2021, doi: 10.3390/s21093190. [6] I. Vajs, D. Drajic, and Z. Cica, “Data-driven machine learning calibration propagation in a hybrid sensor network for air quality monitoring,” Sensors, 2023, doi: 10.3390/s23052815. [7] S. Felici-Castell, J. Segura-Garcia, J. Solano, R. Fayos-Jordan, A. Soriano-Asensi, and J. Alcaraz-Calero, “AI-IoT low-cost pollution-monitoring sensor network to assist citizens with respiratory problems,” Italian National Conference on Sensors, 2023, doi: 10.3390/s23239585. [8] P. Adong, E. Bainomugisha, D. Okure, and R. Sserunjogi, “Applying machine learning for large scale field calibration of low-cost PM 2 .5 and PM 10 air pollution sensors,” Applied AI Letters, 2022, doi: 10.1002/ail2.76. [9] Z. Feng, L. Zheng, B. Ren, D. Liu, J. Huang, and N. Xue, “Feasibility of low-cost particulate matter sensors for long-term environmental monitoring: Field evaluation and calibration.” Science of the Total Environment, 2024, doi: 10.1016/j.scitotenv.2024.174089. [10] A. Patton et al., “Non-linear probabilistic calibration of low-cost environmental air pollution sensor networks for neighborhood level spatiotemporal exposure assessment,” Journal of Exposure Science and Environmental Epidemiology, 2022, doi: 10.1038/s41370-022-00493-y.
Publication Date
11/6/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
1
Total views since publication
PDF Downloads
6
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...