Automated intelligent system for sound signalling device quality assurance

T. Maniak, Chrisina Jayne, Rahat Iqbal, Faiyaz Doctor

    Research output: Contribution to journalArticle

    15 Citations (Scopus)

    Abstract

    This paper presents a novel approach to the detection and recognition of faulty audio signalling devices as part of an automated industrial manufacturing quality assurance process. The proposed system outperforms other well-established automated systems based on mel-frequency cepstrum coefficients (MFCC) and multi-layer perceptron (MLP). It uses both unlabelled sound data and labelled historical data acquired from human experts in detecting faulty signalling devices. The unlabelled data is used to train a deep neural network generative model to create multiple levels of hierarchical feature extractors which are used to train an MLP classifier, with the intent to model the human reasoning and judging processes in respect to sound classification. This paper presents the results of real world experiments based on data pertaining to the audio signalling quality assurance process for car instrument cluster manufacturing. These results show that the proposed system is able to successfully classify speakers into two groups: “Good” and “No good” depending on the part quality. The proposed system proves to be capable enough to eliminate the need for a manual inspection within the manufacturing process and is shown to be able to diagnose a fault with a high degree of accuracy. This work can be extended to other areas of automotive inspection where there is a need for a robust solution to sound detection and where an output signal is represented by a complex and changing frequency spectrum even with significant environmental noise
    Original languageEnglish
    Pages (from-to)600-611
    JournalInformation Sciences
    Volume10
    Early online date14 Oct 2014
    DOIs
    Publication statusPublished - 10 Feb 2015

    Fingerprint

    Quality Assurance
    Intelligent systems
    Intelligent Systems
    Quality assurance
    Acoustic waves
    Multilayer neural networks
    Manufacturing
    Inspection
    Perceptron
    Multilayer
    Cepstrum
    Extractor
    Generative Models
    Frequency Spectrum
    Historical Data
    Classifiers
    Railroad cars
    Network Model
    Fault
    Eliminate

    Bibliographical note

    This paper is not available on the repository

    Keywords

    • Non-speech sound recognition
    • Audio signal processing
    • Artificial neural networks
    • Auto-encoder
    • Manufacturing
    • Automation

    Cite this

    Automated intelligent system for sound signalling device quality assurance. / Maniak, T.; Jayne, Chrisina; Iqbal, Rahat; Doctor, Faiyaz.

    In: Information Sciences, Vol. 10, 10.02.2015, p. 600-611.

    Research output: Contribution to journalArticle

    Maniak, T. ; Jayne, Chrisina ; Iqbal, Rahat ; Doctor, Faiyaz. / Automated intelligent system for sound signalling device quality assurance. In: Information Sciences. 2015 ; Vol. 10. pp. 600-611.
    @article{8a5c3941e1db44409b9d1afc1be3cede,
    title = "Automated intelligent system for sound signalling device quality assurance",
    abstract = "This paper presents a novel approach to the detection and recognition of faulty audio signalling devices as part of an automated industrial manufacturing quality assurance process. The proposed system outperforms other well-established automated systems based on mel-frequency cepstrum coefficients (MFCC) and multi-layer perceptron (MLP). It uses both unlabelled sound data and labelled historical data acquired from human experts in detecting faulty signalling devices. The unlabelled data is used to train a deep neural network generative model to create multiple levels of hierarchical feature extractors which are used to train an MLP classifier, with the intent to model the human reasoning and judging processes in respect to sound classification. This paper presents the results of real world experiments based on data pertaining to the audio signalling quality assurance process for car instrument cluster manufacturing. These results show that the proposed system is able to successfully classify speakers into two groups: “Good” and “No good” depending on the part quality. The proposed system proves to be capable enough to eliminate the need for a manual inspection within the manufacturing process and is shown to be able to diagnose a fault with a high degree of accuracy. This work can be extended to other areas of automotive inspection where there is a need for a robust solution to sound detection and where an output signal is represented by a complex and changing frequency spectrum even with significant environmental noise",
    keywords = "Non-speech sound recognition, Audio signal processing, Artificial neural networks, Auto-encoder, Manufacturing, Automation",
    author = "T. Maniak and Chrisina Jayne and Rahat Iqbal and Faiyaz Doctor",
    note = "This paper is not available on the repository",
    year = "2015",
    month = "2",
    day = "10",
    doi = "10.1016/j.ins.2014.09.042",
    language = "English",
    volume = "10",
    pages = "600--611",
    journal = "Information Sciences",
    issn = "0020-0255",
    publisher = "Elsevier",

    }

    TY - JOUR

    T1 - Automated intelligent system for sound signalling device quality assurance

    AU - Maniak, T.

    AU - Jayne, Chrisina

    AU - Iqbal, Rahat

    AU - Doctor, Faiyaz

    N1 - This paper is not available on the repository

    PY - 2015/2/10

    Y1 - 2015/2/10

    N2 - This paper presents a novel approach to the detection and recognition of faulty audio signalling devices as part of an automated industrial manufacturing quality assurance process. The proposed system outperforms other well-established automated systems based on mel-frequency cepstrum coefficients (MFCC) and multi-layer perceptron (MLP). It uses both unlabelled sound data and labelled historical data acquired from human experts in detecting faulty signalling devices. The unlabelled data is used to train a deep neural network generative model to create multiple levels of hierarchical feature extractors which are used to train an MLP classifier, with the intent to model the human reasoning and judging processes in respect to sound classification. This paper presents the results of real world experiments based on data pertaining to the audio signalling quality assurance process for car instrument cluster manufacturing. These results show that the proposed system is able to successfully classify speakers into two groups: “Good” and “No good” depending on the part quality. The proposed system proves to be capable enough to eliminate the need for a manual inspection within the manufacturing process and is shown to be able to diagnose a fault with a high degree of accuracy. This work can be extended to other areas of automotive inspection where there is a need for a robust solution to sound detection and where an output signal is represented by a complex and changing frequency spectrum even with significant environmental noise

    AB - This paper presents a novel approach to the detection and recognition of faulty audio signalling devices as part of an automated industrial manufacturing quality assurance process. The proposed system outperforms other well-established automated systems based on mel-frequency cepstrum coefficients (MFCC) and multi-layer perceptron (MLP). It uses both unlabelled sound data and labelled historical data acquired from human experts in detecting faulty signalling devices. The unlabelled data is used to train a deep neural network generative model to create multiple levels of hierarchical feature extractors which are used to train an MLP classifier, with the intent to model the human reasoning and judging processes in respect to sound classification. This paper presents the results of real world experiments based on data pertaining to the audio signalling quality assurance process for car instrument cluster manufacturing. These results show that the proposed system is able to successfully classify speakers into two groups: “Good” and “No good” depending on the part quality. The proposed system proves to be capable enough to eliminate the need for a manual inspection within the manufacturing process and is shown to be able to diagnose a fault with a high degree of accuracy. This work can be extended to other areas of automotive inspection where there is a need for a robust solution to sound detection and where an output signal is represented by a complex and changing frequency spectrum even with significant environmental noise

    KW - Non-speech sound recognition

    KW - Audio signal processing

    KW - Artificial neural networks

    KW - Auto-encoder

    KW - Manufacturing

    KW - Automation

    U2 - 10.1016/j.ins.2014.09.042

    DO - 10.1016/j.ins.2014.09.042

    M3 - Article

    VL - 10

    SP - 600

    EP - 611

    JO - Information Sciences

    JF - Information Sciences

    SN - 0020-0255

    ER -