Edge-Based RNN Anomaly Detection Platform in Machine Tools

Chia Yu Lin; Chih Ping Weng; Li-Chun Wang; Hong-Han Shuai; Wen Peng Tseng

doi:10.1080/23080477.2019.1578921

Edge-Based RNN Anomaly Detection Platform in Machine Tools

Chia Yu Lin, Chih Ping Weng, Li-Chun Wang^*, Hong-Han Shuai, Wen Peng Tseng

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

6 Scopus citations

Abstract

With the rapid advances in machine learning algorithms and sensing technologies, machine prognostics and health management (PHM) via data-driven approaches has become a trend in sophisticated machine tool industry. The run-to-failure data are necessary for data-driven approaches. However, the average life of the machine is two to three years, the time of collecting data is extended. It is a big challenge to collect run-to-failure data and build a PHM model. Therefore, we propose an Edge-based RNN Anomaly Detection Platform (ERADP). ERADP builds the model based on healthy data and notify anomalies two hours in advance. The true alarm rate is up to 100%. Besides, ERADP can accelerate the training time almost 120 times faster than the traditional model.

Original language	English
Pages (from-to)	139-146
Number of pages	8
Journal	Smart Science
Volume	7
Issue number	2
DOIs	https://doi.org/10.1080/23080477.2019.1578921
State	Published - 3 Apr 2019

Keywords

Prognostics and health management (PHM)
anomaly detection (AD)
recurrent neural network (RNN)

Access to Document

10.1080/23080477.2019.1578921

Cite this

@article{78378381ca9d4b7e82ae857d601b4670,

title = "Edge-Based RNN Anomaly Detection Platform in Machine Tools",

abstract = "With the rapid advances in machine learning algorithms and sensing technologies, machine prognostics and health management (PHM) via data-driven approaches has become a trend in sophisticated machine tool industry. The run-to-failure data are necessary for data-driven approaches. However, the average life of the machine is two to three years, the time of collecting data is extended. It is a big challenge to collect run-to-failure data and build a PHM model. Therefore, we propose an Edge-based RNN Anomaly Detection Platform (ERADP). ERADP builds the model based on healthy data and notify anomalies two hours in advance. The true alarm rate is up to 100%. Besides, ERADP can accelerate the training time almost 120 times faster than the traditional model.",

keywords = "Prognostics and health management (PHM), anomaly detection (AD), recurrent neural network (RNN)",

author = "Lin, {Chia Yu} and Weng, {Chih Ping} and Li-Chun Wang and Hong-Han Shuai and Tseng, {Wen Peng}",

year = "2019",

month = apr,

day = "3",

doi = "10.1080/23080477.2019.1578921",

language = "English",

volume = "7",

pages = "139--146",

journal = "Smart Science",

issn = "2308-0477",

publisher = "Taylor and Francis Ltd.",

number = "2",

}

TY - JOUR

T1 - Edge-Based RNN Anomaly Detection Platform in Machine Tools

AU - Lin, Chia Yu

AU - Weng, Chih Ping

AU - Wang, Li-Chun

AU - Shuai, Hong-Han

AU - Tseng, Wen Peng

PY - 2019/4/3

Y1 - 2019/4/3

N2 - With the rapid advances in machine learning algorithms and sensing technologies, machine prognostics and health management (PHM) via data-driven approaches has become a trend in sophisticated machine tool industry. The run-to-failure data are necessary for data-driven approaches. However, the average life of the machine is two to three years, the time of collecting data is extended. It is a big challenge to collect run-to-failure data and build a PHM model. Therefore, we propose an Edge-based RNN Anomaly Detection Platform (ERADP). ERADP builds the model based on healthy data and notify anomalies two hours in advance. The true alarm rate is up to 100%. Besides, ERADP can accelerate the training time almost 120 times faster than the traditional model.

AB - With the rapid advances in machine learning algorithms and sensing technologies, machine prognostics and health management (PHM) via data-driven approaches has become a trend in sophisticated machine tool industry. The run-to-failure data are necessary for data-driven approaches. However, the average life of the machine is two to three years, the time of collecting data is extended. It is a big challenge to collect run-to-failure data and build a PHM model. Therefore, we propose an Edge-based RNN Anomaly Detection Platform (ERADP). ERADP builds the model based on healthy data and notify anomalies two hours in advance. The true alarm rate is up to 100%. Besides, ERADP can accelerate the training time almost 120 times faster than the traditional model.

KW - Prognostics and health management (PHM)

KW - anomaly detection (AD)

KW - recurrent neural network (RNN)

UR - http://www.scopus.com/inward/record.url?scp=85066116329&partnerID=8YFLogxK

U2 - 10.1080/23080477.2019.1578921

DO - 10.1080/23080477.2019.1578921

M3 - Article

AN - SCOPUS:85066116329

SN - 2308-0477

VL - 7

SP - 139

EP - 146

JO - Smart Science

JF - Smart Science

IS - 2

ER -

Edge-Based RNN Anomaly Detection Platform in Machine Tools

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this