AIOT-Enabled Autonomous Pre-Ignition Safety Control for Oxygen-Enriched Clinical Spaces

Abstract

Oxygen-rich places like ICUs and operating rooms are tricky when it comes to electrical safety. There’s just more danger—fires and explosions become a real threat if something goes wrong with the wiring. Higher oxygen levels make it way easier for sparks to start a fire, especially if you’ve got things like leakage currents, worn-out insulation, or equipment acting up. Most of the time, you protect against electrical faults with things like fuses and circuit breakers. They do their job by cutting the power immediately when they spot a problem. That works in regular settings. In oxygen-filled environments, though, yanking the plug in one quick move can actually make things worse. You get arcs and sparks just when you don’t want them, and there’s a real risk of starting a fire. Plus, cutting the power suddenly could stop critical medical equipment right when people need it most. The idea here is to do better. This work follows forward a software-based system made for high oxygen. Instead of hardware, it contiously watches how the current and keeps tabs on oxygen levels using series data. By breaking the current into short windows, it pulls out useful features—like the root-mean-square current, signs of leakage (that’s the DC offset), and how quickly the current is changing. Then it uses machine learning models—like sequence networks and boosted classifiers—that are trained mostly on normal data. This helps spot unusual or dangerous patterns as soon as they pop up. Once the system notices something off, it combines the anomaly score with the oxygen level to come up with a real-time risk index. Here, isolation isn’t just about slamming everything off at once. Instead, it means separating out the suspicious load in a careful, staged way. Depending on how bad the fault is and how much oxygen is around, the system might wait a bit or isolate things in steps. The goal is to isolate the risk of sparks while keeping important equipment running. Tests in simulation show that this predictive, oxygen-aware approach keeps things safer and more reliable in sensitive clinical settings. It’s a smarter, more flexible way to handle electrical faults when the stakes are high.

Introduction

Electrical faults in oxygen-enriched clinical environments such as ICUs, operation theaters, and oxygen supply areas pose a serious fire hazard because higher oxygen levels reduce ignition energy and increase flame propagation. Traditional protection systems, including circuit breakers and relays, rely on fixed current thresholds and do not consider fault evolution over time or environmental conditions. As a result, they may either trip unnecessarily or fail to respond appropriately, potentially increasing ignition risks.

To address these limitations, the proposed study introduces a context-aware adaptive electrical isolation system that combines sliding-window current analysis with oxygen-weighted ignition risk estimation. The system continuously monitors electrical current and oxygen concentration using sensors, processes the data through a microcontroller, and calculates a Fault Intensity Score (FIS) based on signal variance and rate of change. Oxygen levels are normalized and incorporated into a risk model to determine overall system risk.

The architecture consists of several modules: electrical sensing, sliding-window processing, fault intensity evaluation, oxygen risk estimation, risk fusion and decision-making, control and isolation, cooldown and recovery, and alert indication. Based on the calculated risk index, the system operates in three modes: Normal Monitoring, Delayed Isolation, and Immediate Isolation. Unlike conventional systems, isolation is adaptive and reversible, allowing automatic power restoration once stable conditions are detected.

The algorithm continuously collects current and oxygen data, extracts statistical features, computes the FIS, and combines it with oxygen sensitivity to generate a risk value. Decision thresholds determine whether the system continues normal operation, delays isolation, or disconnects power immediately. After isolation, the system enters a cooldown phase and restores power automatically when conditions return to safe levels.

Experimental evaluation using 50 simulated test cases demonstrated that the proposed system effectively distinguishes between stable, warning, and critical conditions. Low current fluctuations resulted in normal operation, while higher signal variations and rapid current changes produced warning or critical states. The results confirm that integrating temporal electrical analysis with oxygen-aware risk assessment provides a more reliable and safer protection mechanism for oxygen-rich clinical environments than traditional threshold-based systems.

Conclusion

This paper proposed a context-aware adaptive electrical isolation framework specifically developed for oxygen-rich clinical environments. Unlike other electrical protection systems that utilize instantaneous threshold tripping techniques, this proposed method employs a sliding window analysis in combination with oxygen-aware risk evaluation to make more intelligent decisions regarding electrical isolation. This proposed system effectively identifies fault conditions using current analysis over a specified time interval. The proposed framework also includes oxygen weighting to make it more effective in terms of safety considerations due to its ability to sense oxygen levels in the environment and adjust its decisions accordingly. The proposed framework was developed using a microcontroller-based system with current sensing, ADC conversion, relay control, and oxygen simulation inputs. The proposed system was found to be effective in performing intelligent decisions regarding electrical isolation, including delaying electrical isolation, immediately disconnecting in critical fault conditions, and returning to normal operation after a cool-down period. The classification results obtained using this proposed system show its reliable operation in differentiating normal and fault conditions while ensuring system stability. The proposed framework is an effective solution in terms of cost, scalability, and practical application in improving electrical safety in oxygen-rich environments. This proposed framework is a significant improvement over traditional electrical safety techniques because it includes environmental considerations and reversible electrical isolation using a combination of time analysis and oxygen weighting

References

[1] T.-H. Cheng, C.-H. Chen, C.-H. Lin, B.-H. Sheu, C.-H. Lin and W.-P. Chen, \"Leakage Current Detector and Warning System Integrated with Electric Meter\" Electronics, Volume 12, Issue 9, DOI: https://doi.org/10.3390/electronics12092123 [2] W. R. Lee, \"A clinical study of electrical accidents\" British Journal of Industrial Medicine, Volume 18, Issue 4, pp. 260–269, DOI: https://oem.bmj.com/content/18/4/260.short [3] A. Hal?c? and S. ??leyen, \"Occupational Accidents Caused by Electricity in Turkey\" UluborluMesleki Bilimler Dergisi, Volume 5, Issue 2, DOI: https://dergipark.org.tr/en/pub/umbd/article/970512 [4] C. Maguire, \"Fires from Causes Other Than Electrical Malfunctions: Theory and Case Studies\" in Fire Investigation, N. Nic Daéid, Ed. CRC Press, DOI: https://www.taylorfrancis.com/chapters/edit/10.1201/9780203646953-7/fires-causes-electrical-malfunctions-theory-case-studies-caroline-maguire

Copyright

Copyright © 2026 M N Sowmiya, Priyadarshini P, Chitra Sree M. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.