EUVIMEDEUVIMED entdecken
Crossref

Real-Time Air Quality Monitoring and Smart Filtration System Using Edge-Based Machine Learning

Surbhi Vaidya, Shreya Chakraborty, Samiksha Kumbhalkar, Shravani Rane, Ruchita Khirodkar, Prof Anil Bavaskar

INTERNATIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT · 21.04.2026 · 10 (04) · 1-9

DOI: 10.55041/ijsrem60770 ↗

Kurz gefasst

Fachartikel in der Zeitschrift INTERNATIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT, online erschienen am 21.04.2026. Verfasst von Surbhi Vaidya und 5 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

- Indoor air pollution poses a severe and growing public health risk, with household environments frequently experiencing elevated concentrations of volatile organic compounds, combustible gases, and smoke due to cooking, cleaning activities, and inadequate ventilation. Conventional air purifiers operate on fixed schedules without awareness of actual pollutant concentrations, resulting in energy wastage during safe periods and insufficient purification during pollution events. This paper presents a real-time air quality monitoring and smart filtration system that integrates MQ2, MQ4, and MQ135 gas sensors with an ESP8266/ESP32 microcontroller performing edge-based machine learning inference for intelligent, adaptive filtration control. The system converts raw sensor readings to PPM concentrations and calculates a composite Air Quality Index (AQI). Five machine learning techniques are applied: Linear Regression for AQI prediction, Decision Tree for smart filter speed control, Isolation Forest for anomaly detection, ML-based sensor calibration correction, and time-based pattern prediction. Results are displayed locally on an LCD and simultaneously uploaded to Firebase Realtime Database feeding a web dashboard with five live real-time graphs. The HEPA H13 filter is controlled automatically with variable fan speed (0–100%) based on AQI classification and ML recommendations. Experimental validation confirms AQI prediction R² above 0.87, Decision Tree filter control accuracy above 90%, and average system response time under 3.5 seconds — all achieved at a total hardware cost of approximately Rs. 3,000. Key Words: Air Quality Index, MQ2, MQ4, MQ135, ESP32, ESP8266, Machine Learning, Linear Regression, Decision Tree, Isolation Forest, Firebase, HEPA Filter, Edge Computing, IoT, Smart Filtration.

Zitieren

APA

Vaidya, S., Chakraborty, S., Kumbhalkar, S., Rane, S., Khirodkar, R., & Bavaskar, P. A. (2026). Real-Time Air Quality Monitoring and Smart Filtration System Using Edge-Based Machine Learning. INTERNATIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT, 10(04), 1-9. https://doi.org/10.55041/ijsrem60770

Vancouver

Vaidya S, Chakraborty S, Kumbhalkar S, Rane S, Khirodkar R, Bavaskar PA. Real-Time Air Quality Monitoring and Smart Filtration System Using Edge-Based Machine Learning. INTERNATIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT. 2026;10(04):1-9. doi:10.55041/ijsrem60770

BibTeX
@article{Vaidya2026,
  author = {Vaidya, Surbhi and Chakraborty, Shreya and Kumbhalkar, Samiksha and Rane, Shravani and Khirodkar, Ruchita and Bavaskar, Prof Anil},
  title = {Real-Time Air Quality Monitoring and Smart Filtration System Using Edge-Based Machine Learning},
  journal = {INTERNATIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT},
  year = {2026},
  volume = {10},
  number = {04},
  pages = {1--9},
  doi = {10.55041/ijsrem60770}
}
Kostenlos starten →