Enhancing Predictive Maintenance with Interpretable AutoML: A Case Study on Detecting Ball-Bearing Faults Using IoT Data

Gregory Davrazos, George Raftopoulos, Theodor Panagiotakopoulos, Sotiris Kotsiantis, Achilles Kameas

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    Abstract

    Predictive maintenance has emerged as a crucial approach to enhance the reliability and efficiency of industrial systems. Leveraging Internet of Things (IoT) data, predictive maintenance enables the timely detection and prevention of equipment failures, thereby minimizing downtime and maintenance costs. In this paper, we present a framework for predictive maintenance using interpretable machine learning algorithms, facilitated by Automated Machine Learning (AutoML) techniques. Focusing specifically on ball-bearing faults, a common issue in many mechanical systems, we demonstrate the effectiveness of the approach in accurately predicting impending failures.

    Original languageEnglish
    Title of host publication15th International Conference on Information, Intelligence, Systems and Applications, IISA 2024
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    ISBN (Electronic)9798350368833
    DOIs
    Publication statusPublished - 2024
    Event15th International Conference on Information, Intelligence, Systems and Applications, IISA 2024 - Chania, Greece
    Duration: 17 Jul 202420 Jul 2024

    Publication series

    Name15th International Conference on Information, Intelligence, Systems and Applications, IISA 2024

    Conference

    Conference15th International Conference on Information, Intelligence, Systems and Applications, IISA 2024
    Country/TerritoryGreece
    CityChania
    Period17/07/2420/07/24

    Keywords

    • AutoML
    • CWRU Ball Bearing Dataset
    • Internet of Things
    • Interpretable Machine Learning
    • Predictive Maintenance
    • PyCaret

    Fingerprint

    Dive into the research topics of 'Enhancing Predictive Maintenance with Interpretable AutoML: A Case Study on Detecting Ball-Bearing Faults Using IoT Data'. Together they form a unique fingerprint.

    Cite this