A WGAN Approach to Synthetic TBM Data Generation

Paul Johannes Unterlaß; Georg H. Erharter; Alla Sapronova; Thomas Marcher

doi:10.1007/978-3-031-20241-4_1

A WGAN Approach to Synthetic TBM Data Generation

Paul Johannes Unterlaß^*, Georg H. Erharter, Alla Sapronova, Thomas Marcher

^*Corresponding author for this work

Institute of Rock Mechanics and Tunnelling (2200)

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

Abstract

In this work we propose a generative adversarial network (GAN) based approach of generating synthetic geotechnical data for further applications in research and education. Geotechnical data generated by GANs shows similar characteristics as the original data, but still presents unique samples with no connection to the technical content of the original data. The data can therefore be made available publicly without any legal issues.

A WGAN (Wasserstein GAN) algorithm is used to generate synthetic tunnel boring machine (TBM) operational data based on real data from a major European tunnel construction site. The demands on the synthetic TBM data are of a dualistic nature: on the one hand, the data has to be sufficiently dissimilar to the original data, so that it does not create confidentiality issues (demand for originality). On the other hand, it has to show the same patterns and follow the same rules as the original data, so that it can be used as if it were real TBM data (demand for conformity). The WGAN model describes how a synthetic dataset is generated, in terms of a probabilistic model based on real data. By sampling from this model, we are able to generate new, unique synthetic and realistic TBM data.

We show that the demands for originality and conformity of the newly generated data are fulfilled.

Original language	English
Title of host publication	Trends on Construction in the Digital Era - Proceedings of ISIC 2022
Editors	António Gomes Correia, Miguel Azenha, Paulo J.S. Cruz, Paulo Novais, Paulo Pereira
Publisher	Springer, Cham
Pages	3-19
Number of pages	17
Volume	306
ISBN (Electronic)	978-3-031-20241-4
ISBN (Print)	978-3-031-20240-7
DOIs	https://doi.org/10.1007/978-3-031-20241-4_1
Publication status	Published - 2023
Event	International Society for Intelligent Construction 2022 Conference: Trends on Construction n the Post-Digital Era: ISIC 2022 - Avenida D. Afonso Henriques, 701, Guimarães, Portugal Duration: 6 Sept 2022 → 9 Sept 2022 Conference number: 3 https://icisic2022.com/

Publication series

Name	Lecture Notes in Civil Engineering
Volume	306 LNCE
ISSN (Print)	2366-2557
ISSN (Electronic)	2366-2565

Conference

Conference	International Society for Intelligent Construction 2022 Conference: Trends on Construction n the Post-Digital Era
Abbreviated title	ISIC 2022
Country/Territory	Portugal
City	Guimarães
Period	6/09/22 → 9/09/22
Internet address	https://icisic2022.com/

Keywords

Synthetic data generation
Generative adversarial networks
Machine learning
TBM operational data
Tunnelling
Tunneling

ASJC Scopus subject areas

Civil and Structural Engineering

Access to Document

10.1007/978-3-031-20241-4_1

Cite this

Unterlaß, P. J., Erharter, G. H., Sapronova, A., & Marcher, T. (2023). A WGAN Approach to Synthetic TBM Data Generation. In A. Gomes Correia, M. Azenha, P. J. S. Cruz, P. Novais, & P. Pereira (Eds.), Trends on Construction in the Digital Era - Proceedings of ISIC 2022 (Vol. 306, pp. 3-19). (Lecture Notes in Civil Engineering; Vol. 306 LNCE). Springer, Cham. https://doi.org/10.1007/978-3-031-20241-4_1

A WGAN Approach to Synthetic TBM Data Generation. / Unterlaß, Paul Johannes; Erharter, Georg H.; Sapronova, Alla et al.
Trends on Construction in the Digital Era - Proceedings of ISIC 2022. ed. / António Gomes Correia; Miguel Azenha; Paulo J.S. Cruz; Paulo Novais; Paulo Pereira. Vol. 306 Springer, Cham, 2023. p. 3-19 (Lecture Notes in Civil Engineering; Vol. 306 LNCE).

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

Unterlaß, PJ, Erharter, GH, Sapronova, A & Marcher, T 2023, A WGAN Approach to Synthetic TBM Data Generation. in A Gomes Correia, M Azenha, PJS Cruz, P Novais & P Pereira (eds), Trends on Construction in the Digital Era - Proceedings of ISIC 2022. vol. 306, Lecture Notes in Civil Engineering, vol. 306 LNCE, Springer, Cham, pp. 3-19, International Society for Intelligent Construction 2022 Conference: Trends on Construction n the Post-Digital Era, Guimarães, Portugal, 6/09/22. https://doi.org/10.1007/978-3-031-20241-4_1

@inproceedings{4461d319529b467fb4c8b055f7c4a55d,

title = "A WGAN Approach to Synthetic TBM Data Generation",

abstract = "In this work we propose a generative adversarial network (GAN) based approach of generating synthetic geotechnical data for further applications in research and education. Geotechnical data generated by GANs shows similar characteristics as the original data, but still presents unique samples with no connection to the technical content of the original data. The data can therefore be made available publicly without any legal issues.A WGAN (Wasserstein GAN) algorithm is used to generate synthetic tunnel boring machine (TBM) operational data based on real data from a major European tunnel construction site. The demands on the synthetic TBM data are of a dualistic nature: on the one hand, the data has to be sufficiently dissimilar to the original data, so that it does not create confidentiality issues (demand for originality). On the other hand, it has to show the same patterns and follow the same rules as the original data, so that it can be used as if it were real TBM data (demand for conformity). The WGAN model describes how a synthetic dataset is generated, in terms of a probabilistic model based on real data. By sampling from this model, we are able to generate new, unique synthetic and realistic TBM data.We show that the demands for originality and conformity of the newly generated data are fulfilled.",

keywords = "Synthetic data generation, Generative adversarial networks, Machine learning, TBM operational data, Tunnelling, Tunneling",

author = "Unterla{\ss}, {Paul Johannes} and Erharter, {Georg H.} and Alla Sapronova and Thomas Marcher",

year = "2023",

doi = "10.1007/978-3-031-20241-4_1",

language = "English",

isbn = "978-3-031-20240-7",

volume = "306",

series = "Lecture Notes in Civil Engineering",

publisher = "Springer, Cham",

pages = "3--19",

editor = "{Gomes Correia}, Ant{\'o}nio and Miguel Azenha and Cruz, {Paulo J.S.} and Paulo Novais and Paulo Pereira",

booktitle = "Trends on Construction in the Digital Era - Proceedings of ISIC 2022",

note = "International Society for Intelligent Construction 2022 Conference: Trends on Construction n the Post-Digital Era : ISIC 2022, ISIC 2022 ; Conference date: 06-09-2022 Through 09-09-2022",

url = "https://icisic2022.com/",

}

TY - GEN

T1 - A WGAN Approach to Synthetic TBM Data Generation

AU - Unterlaß, Paul Johannes

AU - Erharter, Georg H.

AU - Sapronova, Alla

AU - Marcher, Thomas

N1 - Conference code: 3

PY - 2023

Y1 - 2023

N2 - In this work we propose a generative adversarial network (GAN) based approach of generating synthetic geotechnical data for further applications in research and education. Geotechnical data generated by GANs shows similar characteristics as the original data, but still presents unique samples with no connection to the technical content of the original data. The data can therefore be made available publicly without any legal issues.A WGAN (Wasserstein GAN) algorithm is used to generate synthetic tunnel boring machine (TBM) operational data based on real data from a major European tunnel construction site. The demands on the synthetic TBM data are of a dualistic nature: on the one hand, the data has to be sufficiently dissimilar to the original data, so that it does not create confidentiality issues (demand for originality). On the other hand, it has to show the same patterns and follow the same rules as the original data, so that it can be used as if it were real TBM data (demand for conformity). The WGAN model describes how a synthetic dataset is generated, in terms of a probabilistic model based on real data. By sampling from this model, we are able to generate new, unique synthetic and realistic TBM data.We show that the demands for originality and conformity of the newly generated data are fulfilled.

AB - In this work we propose a generative adversarial network (GAN) based approach of generating synthetic geotechnical data for further applications in research and education. Geotechnical data generated by GANs shows similar characteristics as the original data, but still presents unique samples with no connection to the technical content of the original data. The data can therefore be made available publicly without any legal issues.A WGAN (Wasserstein GAN) algorithm is used to generate synthetic tunnel boring machine (TBM) operational data based on real data from a major European tunnel construction site. The demands on the synthetic TBM data are of a dualistic nature: on the one hand, the data has to be sufficiently dissimilar to the original data, so that it does not create confidentiality issues (demand for originality). On the other hand, it has to show the same patterns and follow the same rules as the original data, so that it can be used as if it were real TBM data (demand for conformity). The WGAN model describes how a synthetic dataset is generated, in terms of a probabilistic model based on real data. By sampling from this model, we are able to generate new, unique synthetic and realistic TBM data.We show that the demands for originality and conformity of the newly generated data are fulfilled.

KW - Synthetic data generation

KW - Generative adversarial networks

KW - Machine learning

KW - TBM operational data

KW - Tunnelling

KW - Tunneling

UR - http://www.scopus.com/inward/record.url?scp=85145022139&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-20241-4_1

DO - 10.1007/978-3-031-20241-4_1

M3 - Conference paper

SN - 978-3-031-20240-7

VL - 306

T3 - Lecture Notes in Civil Engineering

SP - 3

EP - 19

BT - Trends on Construction in the Digital Era - Proceedings of ISIC 2022

A2 - Gomes Correia, António

A2 - Azenha, Miguel

A2 - Cruz, Paulo J.S.

A2 - Novais, Paulo

A2 - Pereira, Paulo

PB - Springer, Cham

T2 - International Society for Intelligent Construction 2022 Conference: Trends on Construction n the Post-Digital Era

Y2 - 6 September 2022 through 9 September 2022

ER -

A WGAN Approach to Synthetic TBM Data Generation

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

International Society for Intelligent Construction 2022 Conference: Trends on Construction n the Post-Digital Era