Health Assessment of Pump Stations using Time Series Anomaly Detection Deploying AI on the Industrial Edge

##plugins.themes.bootstrap3.article.main##

##plugins.themes.bootstrap3.article.sidebar##

Published Nov 5, 2024
Abhishek Murthy Babak Afshin-Pour Willem Malloy Vasileios Geroulas

Abstract

Industrial operators aim to adopt proactive asset management strategies. However, changes in workplace tenure have led to reduced operator expertise, hindering this goal. Shifting from manually detecting equipment problems, to using automated technology with predictive analysis can help address this challenge. However, legacy systems, like the Supervisory Control and Data Acquisition (SCADA), are not well-suited to scalable predictive approaches. 

SCADA is widely used to manually monitor and manage distributed physical assets. Supporting infrastructure was designed and optimized for that need. Specialized communication protocols are utilized for applications which span large geographical deployments. These protocols ensure data robustness and consistency in variable-quality network environments. However, the resulting data, while forming enterprise data pipelines, lacks granularity and has irregular time spacing, making it unsuitable for machine learning applications.

We present a hybrid cloud-to-edge health monitoring solution for assets connected to SCADA or other legacy control systems. Our solution uses a modbus-based polling system on the edge, to collect data at a much higher granularity than the adjacent SCADA system, letting us detect even subtle and acute patterns in the data. Note that no new sensors are needed, as we connect to the same registers as the existing SCADA system. The high granularity data is assessed at the edge for anomalies, using time series anomaly detection algorithms.  We then synthesize the prediction into a health index that quantifies the recency and the frequency of the detected anomalies for the asset.  The health index is then transmitted to a web-based application, where the user can configure thresholds for generating alerts based on the criticality of the asset. 

We demonstrate our solution in a case study, where the application was deployed using Schneider Electric's Customer First Digital Hub, to monitor a sewage pump station for blockages and other subtle deviations in operating patterns.

How to Cite

Murthy, A., Afshin-Pour, B., Malloy, W., & Geroulas, V. (2024). Health Assessment of Pump Stations using Time Series Anomaly Detection : Deploying AI on the Industrial Edge. Annual Conference of the PHM Society, 16(1). https://doi.org/10.36001/phmconf.2024.v16i1.3870
Abstract 48 | PDF Downloads 49

##plugins.themes.bootstrap3.article.details##

Keywords

Time Series Anomaly Detection, SCADA, Edge computing, health index, sewage pumps

References
1. Concetti, L., Mazzuto, G., Ciarapica, F. E., & Bevilacqua, M. (2023, 03). An unsupervised anomaly detection based on self-organizing map for the oil and gas sector. Applied Sciences, 13, 3725. doi: 10.3390/app13063725

2. Giro, R. A., Bernasconi, G., Giunta, G., & Cesari, S. (2021). A data-driven pipeline pressure procedure for remote monitoring of centrifugal pumps. Journal of Petroleum Science and Engineering, 205, 108845.

3. Moreno-Rodenas, A. M., Duinmeijer, A., & Clemens, F. H. (2021). Deep-learning based monitoring of fog layer dynamics in wastewater pumping stations. Water Research, 202, 117482.

4. Mosallam, A., Medjaher, K., & Zerhouni, N. (2013). Nonparametric time series modelling for industrial prognostics and health management. International Journal of Advanced Manufacturing Technology, 1-25. Scikit-Learn. (n.d.). LocalOutlierFactor. https://scikit-learn.org/stable/modules /generated/sklearn.neighbors. LocalOutlierFactor.html. (Accessed: 2024-06-16)

5. Senk, I., Tegeltija, S., & Tarjan, L. (2024). Machine learning in modern scada systems: Opportunities and challenges. In 2024 23rd international symposium infotehjahorina (infoteh) (p. 1-5).

6. Trstenjak, B., Palasek, B., & Trstenjak, J. (2019, Oct.). A decision support system for the prediction of wastewater pumping station failures based on cbr continuous learning model. Engineering, Technology amp; Applied Science Research, 9(5), 47454749.
Section
Industry Experience Papers