Machine-Learning Models on the Edge to reduce Data Volume in Wide-Area Networks between various Production Sites

Iris Weis; Birgit Vogel-Heuser; Patrick Holstein; Emanuel Trunzer

doi:10.1109/IECON43393.2020.9254984

Machine-Learning Models on the Edge to reduce Data Volume in Wide-Area Networks between various Production Sites

Iris Weis, Birgit Vogel-Heuser, Patrick Holstein, Emanuel Trunzer

Chair of Automation and Information Systems

Technical University of Munich

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Scopus citations

Abstract

The availability of vast amounts of data in automated production systems reveals the potential for data-driven improvements. Jointly using this data across different sites or even across different companies will further increase the validity of data-driven models. However, the throughput in wide area networks is limited, limiting the large-scale transmission of data. Therefore, this paper proposes a data reduction approach to reduce network load based on regression and time series models directly on the shop floor. The machine-learning models are used to predict the signals of the automated production system to prevent the transmission of extensive raw data. It is shown that the approach reduces the network load significantly while still ensuring the fulfillment of the real-time control tasks of the programmable logic controller at any time. Thereby, the reduction of the data is dependent on the error of the reconstructed data that can be tolerated.

Original language	English
Title of host publication	Proceedings - IECON 2020
Subtitle of host publication	46th Annual Conference of the IEEE Industrial Electronics Society
Publisher	IEEE Computer Society
Pages	3831-3835
Number of pages	5
ISBN (Electronic)	9781728154145
DOIs	https://doi.org/10.1109/IECON43393.2020.9254984
State	Published - 18 Oct 2020
Event	46th Annual Conference of the IEEE Industrial Electronics Society, IECON 2020 - Virtual, Singapore, Singapore Duration: 19 Oct 2020 → 21 Oct 2020

Publication series

Name	IECON Proceedings (Industrial Electronics Conference)
Volume	2020-October

Conference

Conference	46th Annual Conference of the IEEE Industrial Electronics Society, IECON 2020
Country/Territory	Singapore
City	Virtual, Singapore
Period	19/10/20 → 21/10/20

Keywords

automated production systems
data acquisition on the edge
data reduction
machine learning
network load
time series prediction
wide-area networks

Access to Document

10.1109/IECON43393.2020.9254984

Cite this

Weis, I., Vogel-Heuser, B., Holstein, P., & Trunzer, E. (2020). Machine-Learning Models on the Edge to reduce Data Volume in Wide-Area Networks between various Production Sites. In Proceedings - IECON 2020: 46th Annual Conference of the IEEE Industrial Electronics Society (pp. 3831-3835). Article 9254984 (IECON Proceedings (Industrial Electronics Conference); Vol. 2020-October). IEEE Computer Society. https://doi.org/10.1109/IECON43393.2020.9254984

Weis, Iris ; Vogel-Heuser, Birgit ; Holstein, Patrick et al. / Machine-Learning Models on the Edge to reduce Data Volume in Wide-Area Networks between various Production Sites. Proceedings - IECON 2020: 46th Annual Conference of the IEEE Industrial Electronics Society. IEEE Computer Society, 2020. pp. 3831-3835 (IECON Proceedings (Industrial Electronics Conference)).

@inproceedings{f4beca7b8be84539b2327e2dcc6bb183,

title = "Machine-Learning Models on the Edge to reduce Data Volume in Wide-Area Networks between various Production Sites",

abstract = "The availability of vast amounts of data in automated production systems reveals the potential for data-driven improvements. Jointly using this data across different sites or even across different companies will further increase the validity of data-driven models. However, the throughput in wide area networks is limited, limiting the large-scale transmission of data. Therefore, this paper proposes a data reduction approach to reduce network load based on regression and time series models directly on the shop floor. The machine-learning models are used to predict the signals of the automated production system to prevent the transmission of extensive raw data. It is shown that the approach reduces the network load significantly while still ensuring the fulfillment of the real-time control tasks of the programmable logic controller at any time. Thereby, the reduction of the data is dependent on the error of the reconstructed data that can be tolerated.",

keywords = "automated production systems, data acquisition on the edge, data reduction, machine learning, network load, time series prediction, wide-area networks",

author = "Iris Weis and Birgit Vogel-Heuser and Patrick Holstein and Emanuel Trunzer",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 46th Annual Conference of the IEEE Industrial Electronics Society, IECON 2020 ; Conference date: 19-10-2020 Through 21-10-2020",

year = "2020",

month = oct,

day = "18",

doi = "10.1109/IECON43393.2020.9254984",

language = "English",

series = "IECON Proceedings (Industrial Electronics Conference)",

publisher = "IEEE Computer Society",

pages = "3831--3835",

booktitle = "Proceedings - IECON 2020",

}

Weis, I, Vogel-Heuser, B, Holstein, P & Trunzer, E 2020, Machine-Learning Models on the Edge to reduce Data Volume in Wide-Area Networks between various Production Sites. in Proceedings - IECON 2020: 46th Annual Conference of the IEEE Industrial Electronics Society., 9254984, IECON Proceedings (Industrial Electronics Conference), vol. 2020-October, IEEE Computer Society, pp. 3831-3835, 46th Annual Conference of the IEEE Industrial Electronics Society, IECON 2020, Virtual, Singapore, Singapore, 19/10/20. https://doi.org/10.1109/IECON43393.2020.9254984

Machine-Learning Models on the Edge to reduce Data Volume in Wide-Area Networks between various Production Sites. / Weis, Iris; Vogel-Heuser, Birgit; Holstein, Patrick et al.
Proceedings - IECON 2020: 46th Annual Conference of the IEEE Industrial Electronics Society. IEEE Computer Society, 2020. p. 3831-3835 9254984 (IECON Proceedings (Industrial Electronics Conference); Vol. 2020-October).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Machine-Learning Models on the Edge to reduce Data Volume in Wide-Area Networks between various Production Sites

AU - Weis, Iris

AU - Vogel-Heuser, Birgit

AU - Holstein, Patrick

AU - Trunzer, Emanuel

PY - 2020/10/18

Y1 - 2020/10/18

N2 - The availability of vast amounts of data in automated production systems reveals the potential for data-driven improvements. Jointly using this data across different sites or even across different companies will further increase the validity of data-driven models. However, the throughput in wide area networks is limited, limiting the large-scale transmission of data. Therefore, this paper proposes a data reduction approach to reduce network load based on regression and time series models directly on the shop floor. The machine-learning models are used to predict the signals of the automated production system to prevent the transmission of extensive raw data. It is shown that the approach reduces the network load significantly while still ensuring the fulfillment of the real-time control tasks of the programmable logic controller at any time. Thereby, the reduction of the data is dependent on the error of the reconstructed data that can be tolerated.

AB - The availability of vast amounts of data in automated production systems reveals the potential for data-driven improvements. Jointly using this data across different sites or even across different companies will further increase the validity of data-driven models. However, the throughput in wide area networks is limited, limiting the large-scale transmission of data. Therefore, this paper proposes a data reduction approach to reduce network load based on regression and time series models directly on the shop floor. The machine-learning models are used to predict the signals of the automated production system to prevent the transmission of extensive raw data. It is shown that the approach reduces the network load significantly while still ensuring the fulfillment of the real-time control tasks of the programmable logic controller at any time. Thereby, the reduction of the data is dependent on the error of the reconstructed data that can be tolerated.

KW - automated production systems

KW - data acquisition on the edge

KW - data reduction

KW - machine learning

KW - network load

KW - time series prediction

KW - wide-area networks

UR - http://www.scopus.com/inward/record.url?scp=85097740391&partnerID=8YFLogxK

U2 - 10.1109/IECON43393.2020.9254984

DO - 10.1109/IECON43393.2020.9254984

M3 - Conference contribution

AN - SCOPUS:85097740391

T3 - IECON Proceedings (Industrial Electronics Conference)

SP - 3831

EP - 3835

BT - Proceedings - IECON 2020

PB - IEEE Computer Society

T2 - 46th Annual Conference of the IEEE Industrial Electronics Society, IECON 2020

Y2 - 19 October 2020 through 21 October 2020

ER -

Weis I, Vogel-Heuser B, Holstein P, Trunzer E. Machine-Learning Models on the Edge to reduce Data Volume in Wide-Area Networks between various Production Sites. In Proceedings - IECON 2020: 46th Annual Conference of the IEEE Industrial Electronics Society. IEEE Computer Society. 2020. p. 3831-3835. 9254984. (IECON Proceedings (Industrial Electronics Conference)). doi: 10.1109/IECON43393.2020.9254984

Machine-Learning Models on the Edge to reduce Data Volume in Wide-Area Networks between various Production Sites

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this