Enhance 3D point cloud accuracy through supervised machine learning for automated rolling stock maintenance: a railway sector case study

Randika K.W. Vithanage; Colin S. Harrison; Anjali K.M. DeSilva

doi:10.1109/iCCECOME.2018.8658788

Enhance 3D point cloud accuracy through supervised machine learning for automated rolling stock maintenance: a railway sector case study

Randika K.W. Vithanage, Colin S. Harrison, Anjali K.M. DeSilva

Mechanical Engineering

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

5 Citations (Scopus)

Abstract

This paper presents findings of a case study conducted to introduce industrial robots into automatic train coupler inspection of Siemens Class 380 rolling-stock. The targets are localized by coalescing RGB and time of flight (ToF) sensor data. The study examines several supervised machine learning techniques to improve the overall accuracy of 3D point clouds. A cost factor which reflects root mean square, mean absolute error and coefficient of determination is defined to evaluate the performance of the learning algorithms. The best-suited models are further validated using simulation data and selected to include in overall robotic sensing system.

Original language	English
Title of host publication	2018 International Conference on Computing, Electronics & Communications Engineering (iCCECE)
Editors	Madhi H. Miraz, Peter S. Excell, Andrew Jones, Safeeullah Soomro, Maaruf Ali
Publisher	IEEE
Pages	242-246
Number of pages	5
ISBN (Electronic)	9781538649046
ISBN (Print)	9781538649053
DOIs	https://doi.org/10.1109/iCCECOME.2018.8658788
Publication status	Published - 7 Mar 2019

Keywords

sensor fusion, railway maintenance, robotic vision, supervised machine learning

Documents and Links

10.1109/iCCECOME.2018.8658788

Cite this

Vithanage, R. K. W., Harrison, C. S., & DeSilva, A. K. M. (2019). Enhance 3D point cloud accuracy through supervised machine learning for automated rolling stock maintenance: a railway sector case study. In M. H. Miraz, P. S. Excell, A. Jones, S. Soomro, & M. Ali (Eds.), 2018 International Conference on Computing, Electronics & Communications Engineering (iCCECE) (pp. 242-246). IEEE. https://doi.org/10.1109/iCCECOME.2018.8658788

Vithanage, Randika K.W. ; Harrison, Colin S. ; DeSilva, Anjali K.M. / Enhance 3D point cloud accuracy through supervised machine learning for automated rolling stock maintenance: a railway sector case study. 2018 International Conference on Computing, Electronics & Communications Engineering (iCCECE). editor / Madhi H. Miraz ; Peter S. Excell ; Andrew Jones ; Safeeullah Soomro ; Maaruf Ali. IEEE, 2019. pp. 242-246

@inproceedings{3ffc596ee5714de798f6dbac564ee2b8,

title = "Enhance 3D point cloud accuracy through supervised machine learning for automated rolling stock maintenance: a railway sector case study",

abstract = "This paper presents findings of a case study conducted to introduce industrial robots into automatic train coupler inspection of Siemens Class 380 rolling-stock. The targets are localized by coalescing RGB and time of flight (ToF) sensor data. The study examines several supervised machine learning techniques to improve the overall accuracy of 3D point clouds. A cost factor which reflects root mean square, mean absolute error and coefficient of determination is defined to evaluate the performance of the learning algorithms. The best-suited models are further validated using simulation data and selected to include in overall robotic sensing system.",

keywords = "sensor fusion, railway maintenance, robotic vision, supervised machine learning",

author = "Vithanage, {Randika K.W.} and Harrison, {Colin S.} and DeSilva, {Anjali K.M.}",

note = "Acceptance and AAM requested. ST 12/12/19 and chaser 31/1/20 ET Copy not found in other repository ET 21/2/20 AAM: wrong version provided, requested AAM again 18/6/20 ET Acceptance date (in SAN): author confirmed month and year; used last date in month – ET 18/6/20 ",

year = "2019",

month = mar,

day = "7",

doi = "10.1109/iCCECOME.2018.8658788",

language = "English",

isbn = "9781538649053",

pages = "242--246",

editor = "Miraz, {Madhi H.} and Excell, {Peter S.} and Andrew Jones and Safeeullah Soomro and Maaruf Ali",

booktitle = "2018 International Conference on Computing, Electronics & Communications Engineering (iCCECE)",

publisher = "IEEE",

}

Vithanage, RKW, Harrison, CS & DeSilva, AKM 2019, Enhance 3D point cloud accuracy through supervised machine learning for automated rolling stock maintenance: a railway sector case study. in MH Miraz, PS Excell, A Jones, S Soomro & M Ali (eds), 2018 International Conference on Computing, Electronics & Communications Engineering (iCCECE). IEEE, pp. 242-246. https://doi.org/10.1109/iCCECOME.2018.8658788

Enhance 3D point cloud accuracy through supervised machine learning for automated rolling stock maintenance: a railway sector case study. / Vithanage, Randika K.W.; Harrison, Colin S.; DeSilva, Anjali K.M.

2018 International Conference on Computing, Electronics & Communications Engineering (iCCECE). ed. / Madhi H. Miraz; Peter S. Excell; Andrew Jones; Safeeullah Soomro; Maaruf Ali. IEEE, 2019. p. 242-246.

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

TY - GEN

T1 - Enhance 3D point cloud accuracy through supervised machine learning for automated rolling stock maintenance: a railway sector case study

AU - Vithanage, Randika K.W.

AU - Harrison, Colin S.

AU - DeSilva, Anjali K.M.

N1 - Acceptance and AAM requested. ST 12/12/19 and chaser 31/1/20 ET Copy not found in other repository ET 21/2/20 AAM: wrong version provided, requested AAM again 18/6/20 ET Acceptance date (in SAN): author confirmed month and year; used last date in month – ET 18/6/20

PY - 2019/3/7

Y1 - 2019/3/7

N2 - This paper presents findings of a case study conducted to introduce industrial robots into automatic train coupler inspection of Siemens Class 380 rolling-stock. The targets are localized by coalescing RGB and time of flight (ToF) sensor data. The study examines several supervised machine learning techniques to improve the overall accuracy of 3D point clouds. A cost factor which reflects root mean square, mean absolute error and coefficient of determination is defined to evaluate the performance of the learning algorithms. The best-suited models are further validated using simulation data and selected to include in overall robotic sensing system.

AB - This paper presents findings of a case study conducted to introduce industrial robots into automatic train coupler inspection of Siemens Class 380 rolling-stock. The targets are localized by coalescing RGB and time of flight (ToF) sensor data. The study examines several supervised machine learning techniques to improve the overall accuracy of 3D point clouds. A cost factor which reflects root mean square, mean absolute error and coefficient of determination is defined to evaluate the performance of the learning algorithms. The best-suited models are further validated using simulation data and selected to include in overall robotic sensing system.

KW - sensor fusion, railway maintenance, robotic vision, supervised machine learning

U2 - 10.1109/iCCECOME.2018.8658788

DO - 10.1109/iCCECOME.2018.8658788

M3 - Conference contribution

SN - 9781538649053

SP - 242

EP - 246

BT - 2018 International Conference on Computing, Electronics & Communications Engineering (iCCECE)

A2 - Miraz, Madhi H.

A2 - Excell, Peter S.

A2 - Jones, Andrew

A2 - Soomro, Safeeullah

A2 - Ali, Maaruf

PB - IEEE

ER -

Vithanage RKW, Harrison CS , DeSilva AKM. Enhance 3D point cloud accuracy through supervised machine learning for automated rolling stock maintenance: a railway sector case study. In Miraz MH, Excell PS, Jones A, Soomro S, Ali M, editors, 2018 International Conference on Computing, Electronics & Communications Engineering (iCCECE). IEEE. 2019. p. 242-246 doi: 10.1109/iCCECOME.2018.8658788

Enhance 3D point cloud accuracy through supervised machine learning for automated rolling stock maintenance: a railway sector case study

Abstract

Keywords

Documents and Links

Fingerprint

Cite this