Anomaly detection for injection molding using probabilistic deep learning

Vili Ketonen; Jan Olaf Blech

doi:10.1109/ICPS49255.2021.9468190

Anomaly detection for injection molding using probabilistic deep learning

Vili Ketonen, Jan Olaf Blech

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

12 Citations (Scopus)

Abstract

Data collection from industrial devices is becoming more popular with the advent of technologies and trends such as the Industrial Internet of Things (IIoT) and Industry 4.0. We propose a deep learning-based approach to detect anomalies in real-time from multivariate time series data and interpret the detected anomalies' root causes. We apply the method to real-world industrial data collected from two injection molding machines. Additionally, we evaluate the method using artificially generated multivariate time series data. We compare the performance of the method to five baseline algorithms from the literature. Our results indicate that the method can detect anomalies in the injection molding machine data and interpret the root causes of the detected anomalies with high performance. Similarly, the method works well on the artificially generated multivariate time series data, demonstrating that the method is also applicable to other multivariate time series data problems.

Original language	English
Title of host publication	Proceedings - 2021 4th IEEE International Conference on Industrial Cyber-Physical Systems, ICPS 2021
Publisher	IEEE
Pages	70-77
Number of pages	8
ISBN (Electronic)	978-1-7281-6207-2
ISBN (Print)	978-1-6654-3045-6
DOIs	https://doi.org/10.1109/ICPS49255.2021.9468190
Publication status	Published - 10 May 2021
MoE publication type	A4 Conference publication
Event	International Conference on Industrial Cyber-Physical Systems - Virtual, Online Duration: 10 May 2021 → 13 May 2021 Conference number: 4

Conference

Conference	International Conference on Industrial Cyber-Physical Systems
Abbreviated title	ICPS
City	Virtual, Online
Period	10/05/2021 → 13/05/2021

Funding

This work was done during the Reboot (www.rebootiotfactory.fi), funded by Business Finland. † Deceased.

Keywords

AI
Anomaly detection
Deep learning
Injection molding
Machine learning
Multivariate time series

Access to Document

10.1109/ICPS49255.2021.9468190

Cite this

@inproceedings{29ab30ca204a44e899e2b47bdd3d0dfb,

title = "Anomaly detection for injection molding using probabilistic deep learning",

abstract = "Data collection from industrial devices is becoming more popular with the advent of technologies and trends such as the Industrial Internet of Things (IIoT) and Industry 4.0. We propose a deep learning-based approach to detect anomalies in real-time from multivariate time series data and interpret the detected anomalies' root causes. We apply the method to real-world industrial data collected from two injection molding machines. Additionally, we evaluate the method using artificially generated multivariate time series data. We compare the performance of the method to five baseline algorithms from the literature. Our results indicate that the method can detect anomalies in the injection molding machine data and interpret the root causes of the detected anomalies with high performance. Similarly, the method works well on the artificially generated multivariate time series data, demonstrating that the method is also applicable to other multivariate time series data problems.",

keywords = "AI, Anomaly detection, Deep learning, Injection molding, Machine learning, Multivariate time series",

author = "Vili Ketonen and Blech, \{Jan Olaf\}",

note = "Funding Information: This work was done during the Reboot (www.rebootiotfactory.fi), funded by Business Finland. † Deceased. Publisher Copyright: {\textcopyright} 2021 IEEE.; International Conference on Industrial Cyber-Physical Systems, ICPS ; Conference date: 10-05-2021 Through 13-05-2021",

year = "2021",

month = may,

day = "10",

doi = "10.1109/ICPS49255.2021.9468190",

language = "English",

isbn = "978-1-6654-3045-6",

pages = "70--77",

booktitle = "Proceedings - 2021 4th IEEE International Conference on Industrial Cyber-Physical Systems, ICPS 2021",

publisher = "IEEE",

address = "United States",

}

Ketonen, V & Blech, JO 2021, Anomaly detection for injection molding using probabilistic deep learning. in Proceedings - 2021 4th IEEE International Conference on Industrial Cyber-Physical Systems, ICPS 2021., 9468190, IEEE, pp. 70-77, International Conference on Industrial Cyber-Physical Systems, Virtual, Online, 10/05/2021. https://doi.org/10.1109/ICPS49255.2021.9468190

Anomaly detection for injection molding using probabilistic deep learning. / Ketonen, Vili; Blech, Jan Olaf.
Proceedings - 2021 4th IEEE International Conference on Industrial Cyber-Physical Systems, ICPS 2021. IEEE, 2021. p. 70-77 9468190.

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - Anomaly detection for injection molding using probabilistic deep learning

AU - Ketonen, Vili

AU - Blech, Jan Olaf

N1 - Conference code: 4

PY - 2021/5/10

Y1 - 2021/5/10

N2 - Data collection from industrial devices is becoming more popular with the advent of technologies and trends such as the Industrial Internet of Things (IIoT) and Industry 4.0. We propose a deep learning-based approach to detect anomalies in real-time from multivariate time series data and interpret the detected anomalies' root causes. We apply the method to real-world industrial data collected from two injection molding machines. Additionally, we evaluate the method using artificially generated multivariate time series data. We compare the performance of the method to five baseline algorithms from the literature. Our results indicate that the method can detect anomalies in the injection molding machine data and interpret the root causes of the detected anomalies with high performance. Similarly, the method works well on the artificially generated multivariate time series data, demonstrating that the method is also applicable to other multivariate time series data problems.

AB - Data collection from industrial devices is becoming more popular with the advent of technologies and trends such as the Industrial Internet of Things (IIoT) and Industry 4.0. We propose a deep learning-based approach to detect anomalies in real-time from multivariate time series data and interpret the detected anomalies' root causes. We apply the method to real-world industrial data collected from two injection molding machines. Additionally, we evaluate the method using artificially generated multivariate time series data. We compare the performance of the method to five baseline algorithms from the literature. Our results indicate that the method can detect anomalies in the injection molding machine data and interpret the root causes of the detected anomalies with high performance. Similarly, the method works well on the artificially generated multivariate time series data, demonstrating that the method is also applicable to other multivariate time series data problems.

KW - AI

KW - Anomaly detection

KW - Deep learning

KW - Injection molding

KW - Machine learning

KW - Multivariate time series

UR - https://www.scopus.com/pages/publications/85112373727

U2 - 10.1109/ICPS49255.2021.9468190

DO - 10.1109/ICPS49255.2021.9468190

M3 - Conference article in proceedings

AN - SCOPUS:85112373727

SN - 978-1-6654-3045-6

SP - 70

EP - 77

BT - Proceedings - 2021 4th IEEE International Conference on Industrial Cyber-Physical Systems, ICPS 2021

PB - IEEE

T2 - International Conference on Industrial Cyber-Physical Systems

Y2 - 10 May 2021 through 13 May 2021

ER -

Anomaly detection for injection molding using probabilistic deep learning

Abstract

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this