Augmented Ultrasonic Data for Machine Learning

    Research output: Contribution to journalArticleScientificpeer-review

    Abstract

    Flaw detection in non-destructive testing, especially in complex signals like ultrasonic data, has thus far relied heavily on the expertise and judgement of trained human inspectors. While automated systems have been used for a long time, these have mostly been limited to using simple decision automation, such as signal amplitude threshold.

    The recent advances in various machine learning algorithms have solved many similarly difficult classification problems, that have previously been considered intractable. For non-destructive testing, encouraging results have already been reported in the open literature, but the use of machine learning is still very limited in NDT applications in the field. Key issue hindering their use, is the limited availability of representative flawed data-sets to be used for training.

    In the present paper, we develop modern, very deep convolutional network to detect flaws from phased-array ultrasonic data. We make extensive use of data augmentation to enhance the initially limited raw data and to aid learning. The data augmentation utilizes virtual flaws - a technique, that has successfully been used in training human inspectors and is soon to be used in nuclear inspection qualification. The results from the machine learning classifier are compared to human performance. We show, that using sophisticated data augmentation, modern deep learning networks can be trained to achieve superhuman performance by significant margin.
    Original languageEnglish
    Number of pages9
    JournalNDT & E International
    Publication statusSubmitted - 27 Mar 2019
    MoE publication typeA1 Journal article-refereed

    Fingerprint

    machine learning
    Nondestructive examination
    Learning systems
    ultrasonics
    Ultrasonics
    Defects
    learning
    augmentation
    education
    human performance
    nondestructive tests
    defects
    qualifications
    phased arrays
    classifiers
    automation
    Learning algorithms
    availability
    inspection
    margins

    Cite this

    @article{40644821b1434f50a1c8e0e713303b89,
    title = "Augmented Ultrasonic Data for Machine Learning",
    abstract = "Flaw detection in non-destructive testing, especially in complex signals like ultrasonic data, has thus far relied heavily on the expertise and judgement of trained human inspectors. While automated systems have been used for a long time, these have mostly been limited to using simple decision automation, such as signal amplitude threshold.The recent advances in various machine learning algorithms have solved many similarly difficult classification problems, that have previously been considered intractable. For non-destructive testing, encouraging results have already been reported in the open literature, but the use of machine learning is still very limited in NDT applications in the field. Key issue hindering their use, is the limited availability of representative flawed data-sets to be used for training.In the present paper, we develop modern, very deep convolutional network to detect flaws from phased-array ultrasonic data. We make extensive use of data augmentation to enhance the initially limited raw data and to aid learning. The data augmentation utilizes virtual flaws - a technique, that has successfully been used in training human inspectors and is soon to be used in nuclear inspection qualification. The results from the machine learning classifier are compared to human performance. We show, that using sophisticated data augmentation, modern deep learning networks can be trained to achieve superhuman performance by significant margin.",
    author = "Iikka Virkkunen and Tuomas Koskinen and Oskari Jessen-Juhler and Jari Rinta-aho",
    year = "2019",
    month = "3",
    day = "27",
    language = "English",
    journal = "NDT & E International",
    issn = "0963-8695",
    publisher = "Elsevier",

    }

    Augmented Ultrasonic Data for Machine Learning. / Virkkunen, Iikka; Koskinen, Tuomas; Jessen-Juhler, Oskari; Rinta-aho, Jari.

    In: NDT & E International, 27.03.2019.

    Research output: Contribution to journalArticleScientificpeer-review

    TY - JOUR

    T1 - Augmented Ultrasonic Data for Machine Learning

    AU - Virkkunen, Iikka

    AU - Koskinen, Tuomas

    AU - Jessen-Juhler, Oskari

    AU - Rinta-aho, Jari

    PY - 2019/3/27

    Y1 - 2019/3/27

    N2 - Flaw detection in non-destructive testing, especially in complex signals like ultrasonic data, has thus far relied heavily on the expertise and judgement of trained human inspectors. While automated systems have been used for a long time, these have mostly been limited to using simple decision automation, such as signal amplitude threshold.The recent advances in various machine learning algorithms have solved many similarly difficult classification problems, that have previously been considered intractable. For non-destructive testing, encouraging results have already been reported in the open literature, but the use of machine learning is still very limited in NDT applications in the field. Key issue hindering their use, is the limited availability of representative flawed data-sets to be used for training.In the present paper, we develop modern, very deep convolutional network to detect flaws from phased-array ultrasonic data. We make extensive use of data augmentation to enhance the initially limited raw data and to aid learning. The data augmentation utilizes virtual flaws - a technique, that has successfully been used in training human inspectors and is soon to be used in nuclear inspection qualification. The results from the machine learning classifier are compared to human performance. We show, that using sophisticated data augmentation, modern deep learning networks can be trained to achieve superhuman performance by significant margin.

    AB - Flaw detection in non-destructive testing, especially in complex signals like ultrasonic data, has thus far relied heavily on the expertise and judgement of trained human inspectors. While automated systems have been used for a long time, these have mostly been limited to using simple decision automation, such as signal amplitude threshold.The recent advances in various machine learning algorithms have solved many similarly difficult classification problems, that have previously been considered intractable. For non-destructive testing, encouraging results have already been reported in the open literature, but the use of machine learning is still very limited in NDT applications in the field. Key issue hindering their use, is the limited availability of representative flawed data-sets to be used for training.In the present paper, we develop modern, very deep convolutional network to detect flaws from phased-array ultrasonic data. We make extensive use of data augmentation to enhance the initially limited raw data and to aid learning. The data augmentation utilizes virtual flaws - a technique, that has successfully been used in training human inspectors and is soon to be used in nuclear inspection qualification. The results from the machine learning classifier are compared to human performance. We show, that using sophisticated data augmentation, modern deep learning networks can be trained to achieve superhuman performance by significant margin.

    M3 - Article

    JO - NDT & E International

    JF - NDT & E International

    SN - 0963-8695

    ER -