Prism-Predictive Industrial Safety and Monitoring Using AI

Abstract

Industrial reliability is paramount for maintaining operational efficiency and safety in high-stakes manufacturing sectors. Traditional maintenance strategies often fail to predict sudden, catastrophic failures or deliberate sabotage attempts in real time. This paper presents the design and implementation of an AI-driven Predictive Industrial Safety and Monitoring (PRISM) system. Developed using the Python and Flask framework, the system integrates a Simulation Service to model the behavior of reactors, fermenters, distillation columns, and heat exchangers under both normal and sabotage conditions. A dynamic analytical approach utilizing pure machine learning and linear regression-based trend prediction is employed. Critical deviations automatically trigger voice and email alerts via Twilio and smtplib services. Experimental results show reliable anomaly detection within milliseconds and predictive warnings within a 45-second window, demonstrating the system’s effectiveness in enhancing industrial safety and resilience.

Introduction

The text presents a Predictive Industrial Safety and Monitoring (PRISM) system designed to improve reliability in modern industrial environments using AI-driven predictive maintenance and real-time fault detection. It addresses challenges in cyber-physical and IoT-enabled systems where equipment failures, downtime, and safety risks are increasing due to complex infrastructure.

Traditional maintenance methods rely on static thresholds, manual monitoring, and scheduled inspections, which often fail to detect rapid or evolving failures and delay response actions. To overcome these limitations, PRISM introduces a real-time, AI-based pipeline that processes streaming sensor data and predicts equipment degradation before failure occurs.

The system uses a simulated industrial environment (reactors, fermenters, distillation columns, and heat exchangers) to generate sensor data safely. Incoming data is stored in sliding time windows and analyzed using a Linear Regression model, which detects trends and predicts time-to-failure (TTF) based on slope analysis and confidence scoring. Based on predictions, the system classifies machine states as normal or degrading.

When a fault is detected, an automated response system immediately notifies operators via voice calls (Twilio) and email alerts, ensuring rapid intervention. A chatbot interface further supports real-time status updates and troubleshooting.

Experimental results show that the system can accurately predict gradual failures, such as pressure buildup, with low error rates and minimal latency in alerts. The architecture demonstrates fast, real-time decision-making suitable for industrial safety applications.

Conclusion

The Predictive Industrial Safety and Monitoring (PRISM) system is designed for application in high-stakes manufacturing sectors to maintain operational efficiency and enhance industrial safety. It acts as an advanced predictive maintenance tool that monitors critical industrial assets, such as chemical reactors, fermenters, distillation columns, and heat exchangers. By utilizing a dynamic linear regression approach trained on rolling historical buffers, the system successfully identifies gradual equipment degradation like creeping thermal runaway or slow pressure buildup. The integration of a lightweight Python and Flask architecture allows for seamless real-time data processing and model fitting. The automated response system, leveraging Twilio for voice calls alongside SMTP implementations for email notifications, ensures that critical personnel are proactively alerted before a physical breach occurs. Overall, the PRISM system helps reduce unplanned downtime and improves safety by enabling early detection of failures.

References

[1] Achouch, M., Dimitrova, M., Ziane, K., Sattarpanah Karganroudi, S., Dhouib, R., Ibrahim, H., & Adda, M. (2022). On Predictive Maintenance in Industry 4.0: Overview, Models, and Challenges. Applied Sciences, 12(16), 8081. doi: 10.3390/app12168081. [2] Albattah, A., & Rassam, M. A. (2022). A Correlation-Based Anomaly Detection Model for Wireless Body Area Networks Using Convolutional Long Short-Term Memory Neural Network. Sensors, 22(5), 1951. doi: 10.3390/s22051951. [3] Aminzadeh, A., Sattarpanah Karganroudi, S., Majidi, S., Dabompre, C., Azaiez, K., Mitride, C., & Sénéchal, E. (2025). A Machine Learning Implementation to Predictive Maintenance and Monitoring of Industrial Compressors. Sensors, 25(4), 1006. doi: 10.3390/s25041006. [4] Belim, M., Meireles, T., Gonçalves, G., & Pinto, R. (2024). Forecasting models analysis for predictive maintenance. Frontiers in Manufacturing Technology, 4. doi: 10.3389/fmtec.2024.1475078. [5] DeMedeiros, K., Hendawi, A., & Alvarez, M. (2023). A Survey of AI-Based Anomaly Detection in IoT and Sensor Networks. Sensors, 23(3), 1352. doi: 10.3390/s23031352. [6] Du, Y., Huang, Y., Wan, G., & He, P. (2022). Deep Learning-Based Cyber–Physical Feature Fusion for Anomaly Detection in Industrial Control Systems. Mathematics, 10(22), 4373. doi: 10.3390/math10224373. [7] Hector, I., & Panjanathan, R. (2024). Predictive maintenance in Industry 4.0: a survey of planning models and machine learning techniques. PeerJ Computer Science, 10, e2016. doi: 10.7717/peerj-cs.2016. [8] Jaramillo-Alcazar, A., Govea, J., & Villegas-Ch, W. (2023). Anomaly Detection in a Smart Industrial Machinery Plant Using IoT and Machine Learning. Sensors, 23(19), 8286. doi: 10.3390/s23198286. [9] Li, Z., Wang, K., & He, Y. (2016). Industry 4.0 - Potentials for Predictive Maintenance. Proceedings of the 6th International Workshop of Advanced Manufacturing and Automation, 8. doi: 10.2991/iwama-16.2016.8. [10] Martí, L., Sanchez-Pi, N., Molina, J., & Garcia, A. (2015). Anomaly Detection Based on Sensor Data in Petroleum Industry Applications. Sensors, 15(2), 2774-2797. doi: 10.3390/s150202774. [11] Mol?da, M., Ma?ysiak-Mrozek, B., Ding, W., Sunderam, V., & Mrozek, D. (2023). From Corrective to Predictive MaintenanceA Review of Maintenance Approaches for the Power Industry. Sensors, 23(13), 5970. doi: 10.3390/s23135970. [12] Muntean, M. V. (2024). Real-Time Detection of IoT Anomalies and Intrusion Data in Smart Cities Using Multi-Agent System. Sensors, 24(24), 7886. doi: 10.3390/s24247886. [13] Okfie, M. I. H., & Mishra, S. (2024). Anomaly Detection in IIoT Transactions using Machine Learning: A Lightweight Blockchain-based Approach. Engineering, Technology & Applied Science Research, 14, 14645-14653. doi: 10.48084/etasr.7384. [14] Raeiszadeh, M., Ebrahimzadeh, A., Glitho, R. H., Eker, J., & Mini, R. A. F. (2024). Real-Time Adaptive Anomaly Detection in Industrial IoT Environments. IEEE Transactions on Network and Service Management, 21(6), 6839-6856. doi: 10.1109/tnsm.2024.3447532. [15] Rafique, S. H., Abdallah, A., Musa, N. S., & Murugan, T. (2024). Machine Learning and Deep Learning Techniques for Internet of Things Network Anomaly Detection Current Research Trends. Sensors, 24(6), 1968. doi: 10.3390/s24061968.

Copyright

Copyright © 2026 Dr. N Vinothini, Balavishnu J, Pravin M, Rajapandi K, Vinith S. 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.