ANALYSIS OF SENSOR INDICATORS FOR PREDICTIVE MAINTENANCE OF HEAVY INDUSTRY PROCESS LINES
DOI:
https://doi.org/10.32689/maup.it.2025.1.3Keywords:
process line, predictive maintenance, condition-based predictive maintenance, predictive model, descriptive statistics, contextual statistics, correlation analysisAbstract
The article considers the directions of digital transformation in the maintenance of process lines of heavy industry, aimed at predicting potential equipment malfunctions before their complete functional failure. The choice of methods for detecting problematic data coming from sensors of the process line is justified, aimed at reducing the time interval between the occurrence of a problem and its detection. Approaches to ensuring optimal updates of Predictive Maintenance models are considered, which enhance the accuracy of failure prediction. The purpose of the article is to research and identify effective methods for analyzing the quality of data from sensors of process lines in heavy industry to increase the accuracy of failure prediction in predictive maintenance. Methodology. Software and hardware solutions aimed at increasing the accuracy of forecasting faults in process lines based on the RCM methodology have been analyzed. Methods of correlation analysis, descriptive and contextual statistics were used to monitor sensor indicator data. A sequence of stages for detecting problematic data in real-world production conditions in heavy industry is proposed, which provide automated generation of potential sources of faults. The scientific novelty of the research lies in the identification of methods and approaches to analyzing the performance of sensors of process lines in the conditions of Industry 4.0, aimed at increasing the accuracy of predictive maintenance models. Conclusions. Maintaining Predictive Maintenance models that serve heavy industry process lines in real production conditions requires adjusting and updating their parameters based on constant monitoring of the equipment status in real time. To ensure the detection of problematic data, the feasibility of developing an information system, implementing the analysis of sensor indicators data using descriptive statistics, contextual statistics, and correlation analysis methods. The results of implementing the developed information and analytical system, which includes a recommendation system at the stage of forming a data set for analysis, were accompanied by an increase in failure prediction accuracy and a decrease in the time between the occurrence and detection of problems leading to equipment failures. This indicates that operational tasks are being solved with higher accuracy and efficiency.
References
Болюбаш Н. М., Фінькова О. В. Система аналізу даних для прогнозного технічного обслуговування технологічних ліній. Матеріали XXVІІ Всеукраїнської науково-практичної конференції «Могилянські читання – 2024: Досвід та тенденції розвитку суспільства в Україні: глобальний, національний та регіональний аспекти». Технічні науки. Миколаїв: Вид-во ЧНУ імені П. Могили. 2024. С. 163–167. URL: https://dspace.chmnu.edu.ua/jspui/bitstream/123456789/2507/1/Технічні%20науки.pdf#page=165.
Темчур В. С., Баган Т. Г. Методи глибокого навчання моделей для прогнозного обслуговування. Вчені записки ТНУ ім. В.І. Вернадського. Серія: технічні науки. 2023. Т. 34 (73), № 6. С. 155–162. URL: https://doi.org/10.32782/2663-5941/2023.6/23.
Achouch M., Dimitrova M., Ziane K., Sattarpanah Karganroudi S., Dhouib R., Ibrahim H., Adda M. On Predictive Maintenance in Industry 4.0: Overview, Models, and Challenges. Applied Sciences. 2022. Vol. 12, No 16. P. 8081. URL: https://doi.org/10.3390/app12168081.
Baptista M., Sankararaman S., et al. Forecasting fault events for predictive maintenance using data-driven techniques and ARMA modeling. Computers & Industrial Engineering. 2018. Vol. 115. P. 41–53. URL: https://doi.org/10.1016/j.cie.2020.107060.
Blann D. Maximizing the P-F interval through condition-based maintenance. URL: https://www.maintworld.com/Applications/Maximizing-the-P-F-Interval-Through-Condition-Based-Maintenance.
Buyske S., Fagerstrom R., Ying Z. A class of weighted log-rank tests for survival data when the event is rare. Journal of the American Statistical Association. 2020. Vol. 95, No 449. P. 249–258. URL: https://doi.org/10.1080/01621459.2000.10473918.
Geisbush, J., Ariaratnam, S. T. Reliability centered maintenance (RCM): literature review of current industry state of practice. Journal of Quality in Maintenance Engineering. 2023. Vol. 29, No. 2. PP. 313–337. URL: https://doi.org/10.1108/JQME-02-2021-0018.
Hamasha M. M., Bani-Irshid A. H., Al Mashaqbeh S., et al. Strategical selection of maintenance type under different conditions. Scientific Reports. 2023. Vol. 13, P. 15560. URL: https://doi.org/10.1038/s41598-023-42751-5.
Horrell M., Reynolds L., McElhinney A. Data Science in Heavy Industry and the Internet of Things. Harvard Data Science Review. 2020. Issue 2 (2). URL: https://doi.org/10.1162/99608f92.834c6595.
Keleko A. T., Kamsu-Foguem B., Ngouna R. H., Tongne A. Artificial Intelligence and Real-Time Predictive Maintenance in Industry 4.0: a bibliometric analysis. AI and Ethics. 2022. Vol 2. P. 533–577. URL: https://doi.org/10.1007/s43681-021-00132-6.
Li L., Liu M., Shen W., Cheng G. An expert knowledge-based dynamic maintenance task assignment model using discrete stress-strength interference theory. Knowledge-Based Systems. 2017. Vol. 131. P.135–148. URL: https://doi.org/10.1016/j.knosys.2017.06.008.
Nokeri T. C. Dash Bootstrap Components. In: Web App Development and Real-Time Web Analytics with Python. Pub.: Berkeley, CA. 2022. P. 87–97. URL: https://doi.org/10.1007/978-1-4842-7783-6_6.
Oliynyk K. 5 Use Cases of Predictive Maintenance using IoT. URL: https://webbylab.com/blog/use-cases-of-iotbased-predictive-maintenance.
Ong H., Niyato D., Yuen C. Predictive Maintenance for Edge-Based Sensor Networks. 2020. URL: https://arxiv.org/pdf/2007.03313.
Patel M, Vasa J., Patel B. Predictive Maintenance: A Comprehensive Analysis and Future Outlook. Published in 2nd International Conference on Futuristic Technologies (INCOFT). 2024. URL: https://doi.org/10.1109/INCOFT60753.2023.10425122.
Rudin C. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence. 2019. Vol. 1, No 5. P. 206-215. URL: https://doi.org/10.1038/s42256-019-0048-x.
Serradilla O., Zugasti E., Zurutuza U. Deep Learning Models for Predictive Maintenance: A Survey. 2020. URL: https://doi.org/10.48550/arXiv.2010.03207.
Sharma J., Mittal M.L., Soni G. Condition-based maintenance using machine learning and role of interpretability: a review. International Journal of System Assurance Engineering and Management. Vol. 15. P. 1345–1360. URL: https://doi.org/10.1007/s13198-022-01843-7.
Sikorska J., Hodkiewicz M., Ma L. Prognostic modelling options for remaining useful life estimation by industry. Mechanical Systems and Signal Processing. 2011. Vol. 25, No 5. P. 1803-1836. URL: https://doi.org/10.1016/j.ymssp.2010.11.018.
Ucar A., Karakose M., Kırımça N. Artificial Intelligence for Predictive Maintenance Applications: Key Components, Trustworthiness, and Future Trends. Applied Sciences. 2024. Vol. 14, No 2. P. 898. URL: https://doi.org/10.3390/app14020898.
Wheeler K.R., Kurtoglu T., Poll S.D. A survey of health management user objectives related to diagnostic and prognostic metrics. In Proceedings International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. 2009. P. 1287–1298. URL: https://doi.org/10.1115/DETC2009-87073.
Yuan M., Wu Y., Lin L. Fault diagnosis and remaining useful life estimation of aero engine using LSTM neural network. In Proceedings International Conference on Aircraft Utility Systems (AUS). IEEE. 2016. P. 135–140. URL: https://doi.org/10.1109/AUS.2016.7748035.
Zhang C., Lim P., Qin A. K., Tan K. C. Multiobjective Deep Belief Networks Ensemble for Remaining Useful Life Estimation in Prognostics. IEEE Transactions on Neural Networks and Learning Systems. 2017. Vol. 28, No 10. P. 2306–2318. URL: https://doi.org/10.1109/TNNLS.2016.2582798.
