Integrating Computer Vision and Probabilistic Machine Learning for Enhanced Predictive Maintenance in Manufacturing Systems
Keywords:
Predictive Maintenance, Computer Vision, Monte Carlo Dropout, Convolutional Neural Network, Uncertainty Estimation, Calibration, Random Forest, Support Vector Machine, Industry 4.0, Smart ManufacturingAbstract
As the industry transitions to Industry 4.0, the use of predictive maintenance (PdM) in factories has become crucial for achieving high efficiency and lower costs. Despite these powerful models, the use of deep learning in PdM is hindered by concerns about their reliability, interpretability, and the uncertainty they introduce. This paper outlines a combination of computer vision and probabilistic machine learning that helps improve decision-making in predictive maintenance. Using a conversion process, time-series information from sensors is transformed into images, allowing Convolutional Neural Networks (CNNs) to understand the spatial details of the machine’s health. The challenge of model confidence is overcome by using Monte Carlo (MC) Dropout during inference in the CNN to generate multiple possible outcomes. The integration enables immediate uncertainty estimation, which plays a crucial role in critical maintenance decisions. The system proposed in this study is evaluated against two well-known interpretable models using classification metrics, ROC curves, and calibration plots. This model indicated that CNN+MC Dropout performs better with visual scenes and uncertainty detection, but traditional models are better at classifying and guiding predictions. As a result, both understanding and trusting a model, along with obtaining accurate results, are crucial. The research demonstrates that uncertainty-aware deep learning enhances our experience and trust in the system. Further studies will include work on live camera images using temporal models based on recurrent probabilistic networks.
Downloads
References
Aravind, R., C.V. Shah, and M.D. Surabhi, Machine learning applications in predictive maintenance for vehicles: case studies. International journal of engineering and computer science, 2022. 11(11).
Khan, A.I. and S. Al-Habsi, Machine learning in computer vision. Procedia Computer Science, 2020. 167: p. 1444-1451.
Seliya, N., A. Abdollah Zadeh, and T.M. Khoshgoftaar, A literature review on one-class classification and its potential applications in big data. Journal of Big Data, 2021. 8: p. 1-31.
Serradilla, O., et al., Deep learning models for predictive maintenance: a survey, comparison, challenges and prospects. Applied Intelligence, 2022. 52(10): p. 10934-10964.
Sharpe, C., et al., A comparative evaluation of supervised machine learning classification techniques for engineering design applications. Journal of Mechanical Design, 2019. 141(12): p. 121404.
Fernandes, M., J.M. Corchado, and G. Marreiros, Machine learning techniques applied to mechanical fault diagnosis and fault prognosis in the context of real industrial manufacturing use-cases: a systematic literature review. Applied Intelligence, 2022. 52(12): p. 14246-14280.
T, G., R. V, and U.K. N. L, Detecting Security Threats in Wireless Sensor Networks using Hybrid Network of CNNs and Long Short-Term Memory. International Journal of Intelligent Systems and Applications in Engineering, 2023. 12(1s): p. 704 - 722.
Malali, N., Using Machine Learning to Optimize Life Insurance Claim Triage Processes Via Anomaly Detection in Databricks: Prioritizing High-Risk Claims for Human Review. International Journal of Engineering Technology Research & Management (IJETRM), 2022. 6(06).
Rishabh Rajesh, S., Exploring the Use of Cloud-Based AI and ML for Real-Time Anomaly Detection and Predictive Maintenance in Industrial IoT Systems. International Journal of Intelligent Systems and Applications in Engineering, 2023. 11(4): p. 925 – 937.
Raschka, S., J. Patterson, and C. Nolet, Machine Learning in Python: Main Developments and Technology Trends in Data Science, Machine Learning, and Artificial Intelligence. Information, 2020. 11(4): p. 193.
Yuan, C.J., et al., Expert Analysis for Multi-criteria Human-in-the-Loop Input Selection for Predictive Maintenance Model, in Materials, Design and Manufacturing for Sustainable Environment: Select Proceedings of ICMDMSE 2022. 2022, Springer. p. 461-473.
Staartjes, V.E. and J.M. Kernbach, Foundations of Machine Learning-Based Clinical Prediction Modeling: Part III—Model Evaluation and Other Points of Significance, in Machine Learning in Clinical Neuroscience: Foundations and Applications. 2021, Springer. p. 23-31.
Aseeri, A.O., Uncertainty-aware deep learning-based cardiac arrhythmias classification model of electrocardiogram signals. Computers, 2021. 10(6): p. 82.
Lambert, B., et al., Trustworthy clinical AI solutions: a unified review of uncertainty quantification in deep learning models for medical image analysis. arXiv preprint arXiv:2210.03736, 2022.
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
All papers should be submitted electronically. All submitted manuscripts must be original work that is not under submission at another journal or under consideration for publication in another form, such as a monograph or chapter of a book. Authors of submitted papers are obligated not to submit their paper for publication elsewhere until an editorial decision is rendered on their submission. Further, authors of accepted papers are prohibited from publishing the results in other publications that appear before the paper is published in the Journal unless they receive approval for doing so from the Editor-In-Chief.
IJISAE open access articles are licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. This license lets the audience to give appropriate credit, provide a link to the license, and indicate if changes were made and if they remix, transform, or build upon the material, they must distribute contributions under the same license as the original.
