AI-Enhanced Fault-Aware LPG Leak Monitoring with Environmental Compensation and Network-Independent Local Alarming

Zainab marid alzamili

Abstract


Low cost LPG alarms are frequently demonstrated with a short prototype test that does not measure environmental cross-sensitivity, sensor drift, transfer between alarms, nuisance alarms or network failure. In this study, the authors outline a reference architecture for Arduino/MQ-6 that is designed for fault awareness; local alarming is separated from remote communication, temperature and humidity are considered explicit covariates, and the quality of alarms is assessed on an event basis. The current reproducible simulation consists of 12 development sessions and 36 held-out sessions on 6 virtual devices (two unseen) with 216 target events, and 18 simulated monitoring hours. Environmental compensation (B2) generated 205/216 (94.91%; 95% session-bootstrap interval 91.20-97.69%) new alarms per hour of the session and 4.19 false alarms per hour of background. The nuisance-alarm reduction for the false alarm as compared to the filtered uncorrected comparator was not definitive and the nuisance alarm level was raised with a raw-signal override. Four AI algorithms (Random Forest, RBF-SVM, gradient-boosted trees, compact multilayer perceptron) are included as secondary decision models, with leak-signal, temporal, environmental, and validity features. They are to be evaluated separately from the results that were already obtained based on rules, and they should be evaluated based on device/session-level splits to avoid leakage. The paper further reviews the latest research conducted in 2022–2026 and presents a detailed synthesis of the research and a physical testing protocol for real LPG testing.

Keywords


calibration transfer; embedded AI; environmental compensation; gradient boosting; LPG leak detection; MOS gas sensor; MQ-6; neural network; Random Forest; sensor drift; SVM

Full Text:

PDF

References


D. R. Pamungkas, A. Setiawan, and A. Widiyanto, “Design of Arduino-based LPG Gas Leak Detection Tool using MQ-6 Sensor,” Innovation, Technology, and Entrepreneurship Journal, Vol. 1, No. 1, pp. 21–27, 2024, DOI: 10.31603/itej.10687.

S. W. I. Made, A. A. N. Gunawan, P. I. Ketut, and S. I. Ketut, “Design and Manufacture of LPG Gas Leak Detection based on Arduino Uno using MQ-2 Sensor,” Asian Journal of Research and Reviews in Physics, Vol. 8, No. 1, pp. 27–32, 2024, DOI: 10.9734/ajr2p/2024/v8i1155.

A. N. Abdullah et al., “Correction Model for Metal Oxide Sensor Drift Caused by Ambient Temperature and Humidity,” Sensors, Vol. 22, No. 9, 3301, 2022, DOI: 10.3390/s22093301.

R. Rivera-Martinez et al., “Using Metal Oxide Gas Sensors to Estimate the Emission Rates and Locations of Methane Leaks in an Industrial Site: Assessment with Controlled Methane Releases,” Atmospheric Measurement Techniques, Vol. 17, pp. 4257–4290, 2024, DOI: 10.5194/amt-17-4257-2024.

J. Wörner, J. Eimler, and M. Pein-Hackelbusch, “Long-Term Drift Behavior in Metal Oxide Gas Sensor Arrays: A One-Year Dataset from an Electronic Nose,” Scientific Data, Vol. 12, 1628, 2025, DOI: 10.1038/s41597-025-05993-8.

B. Jena, S. K. Pradhan, R. Jha, S. Goel, and R. Sharma, “LPG Gas Leakage Detection System using IoT,” Materials Today: Proceedings, Vol. 74, pt. 4, pp. 795–800, 2023, DOI: 10.1016/j.matpr.2022.11.172.

X. Dong and S. Han, “Real-Time Correction and Long-Term Drift Compensation in MOS Gas Sensor Arrays using Iterative Random Forests and Incremental Domain-Adversarial Networks,” Micromachines, Vol. 16, No. 9, 991, 2025, DOI: 10.3390/mi16090991.

J. Lin and X. Zhan, “Sensor-Drift Compensation in Electronic-Nose-based Gas Recognition using Knowledge Distillation,” Informatics, Vol. 13, No. 1, 15, 2026, DOI: 10.3390/informatics13010015.

J. Schauer, J. Morsch, D. Arendes, A. Schütze, and C. Bur, “Interpretable Calibration Transfer and Drift Compensation for MOS Gas Sensors in Complex Gas Mixtures,” Sensors, Vol. 26, No. 14, 4595, 2026, DOI: 10.3390/s26144595.

Q. Yang et al., “MDFE-Net: A Meta-Learning Driven Dual-Branch Feature Extraction Network for E-Nose Sensor Drift Adaptation,” ACS Sensors, Vol. 11, No. 5, 2026, DOI: 10.1021/acssensors.6c00538.

R. Li, Z. Li, B. A. Kofi, J. Sun, Y. He, and M. Jiao, “A Comprehensive Review of Algorithms for Drift Compensation in Metal Oxide Semiconductor Gas Sensor Arrays,” Chemosensors, Vol. 14, No. 6, 143, 2026, DOI: 10.3390/chemosensors14060143.

K. Jiang et al., “Gas Sensor Drift Compensation using Semi-Supervised Ensemble Classifiers with Multi-Level Features and Center Loss,” ACS Sensors, Vol. 10, No. 4, pp. 2906–2918, 2025, DOI: 10.1021/acssensors.4c03655.

Zhengzhou Winsen Electronics Technology Co., Ltd., “MQ-6 MOS Flammable Gas Sensor,” product documentation, accessed Oct. 1, 2026.

Arduino, “Analog Input Pins,” Arduino Documentation, accessed Oct. 1, 2026.

Espressif Systems, “ESP8266 Hardware Design Guidelines,” ver. 2.8, accessed Oct. 1, 2026.




DOI: https://doi.org/10.32520/stmsi.v15i11.7013

Article Metrics

Abstract view : 0 times
PDF - 0 times

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.