Tampere University of Technology

TUTCRIS Research Portal

Profile extraction and deep autoencoder feature extraction for elevator fault detection

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Standard

Profile extraction and deep autoencoder feature extraction for elevator fault detection. / Mishra, Krishna; Krogerus, Tomi; Huhtala, Kalevi.

16th International Conference on Signal Processing and Multimedia Applications: SIGMAP 2019, 26-28 July, 2019, Prague, Czech Republic. ed. / Christian Callegari. Vol. 16 2019. ed. Prague, Czech Republic : SCITEPRESS, 2019. p. 313-320.

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Harvard

Mishra, K, Krogerus, T & Huhtala, K 2019, Profile extraction and deep autoencoder feature extraction for elevator fault detection. in C Callegari (ed.), 16th International Conference on Signal Processing and Multimedia Applications: SIGMAP 2019, 26-28 July, 2019, Prague, Czech Republic. 2019 edn, vol. 16, SCITEPRESS, Prague, Czech Republic, pp. 313-320, INTERNATIONAL CONFERENCE ON SIGNAL PROCESSING AND MULTIMEDIA APPLICATIONS, 1/01/00. https://doi.org/10.5220/0007802003130320

APA

Mishra, K., Krogerus, T., & Huhtala, K. (2019). Profile extraction and deep autoencoder feature extraction for elevator fault detection. In C. Callegari (Ed.), 16th International Conference on Signal Processing and Multimedia Applications: SIGMAP 2019, 26-28 July, 2019, Prague, Czech Republic (2019 ed., Vol. 16, pp. 313-320). Prague, Czech Republic: SCITEPRESS. https://doi.org/10.5220/0007802003130320

Vancouver

Mishra K, Krogerus T, Huhtala K. Profile extraction and deep autoencoder feature extraction for elevator fault detection. In Callegari C, editor, 16th International Conference on Signal Processing and Multimedia Applications: SIGMAP 2019, 26-28 July, 2019, Prague, Czech Republic. 2019 ed. Vol. 16. Prague, Czech Republic: SCITEPRESS. 2019. p. 313-320 https://doi.org/10.5220/0007802003130320

Author

Mishra, Krishna ; Krogerus, Tomi ; Huhtala, Kalevi. / Profile extraction and deep autoencoder feature extraction for elevator fault detection. 16th International Conference on Signal Processing and Multimedia Applications: SIGMAP 2019, 26-28 July, 2019, Prague, Czech Republic. editor / Christian Callegari. Vol. 16 2019. ed. Prague, Czech Republic : SCITEPRESS, 2019. pp. 313-320

Bibtex - Download

@inproceedings{bb0b36182dfa4678ac924491eed7ee4a,
title = "Profile extraction and deep autoencoder feature extraction for elevator fault detection",
abstract = "In this paper, we propose a new algorithm for data extraction from time series signal data, and furthermore automatic calculation of highly informative deep features to be used in fault detection. In data extraction elevator start and stop events are extracted from sensor data, and a generic deep autoencoder model is also developed for automated feature extraction from the extracted profiles. After this, extracted deep features are classified with random forest algorithm for fault detection. Sensor data are labelled as healthy and faulty based on the maintenance actions recorded. The remaining healthy data are used for validation of the model to prove its efficacy in terms of avoiding false positives. We have achieved 100{\%} accuracy in fault detection along with avoiding false positives based on new extracted deep features, which outperforms results using existing features. Existing features are also classified with random forest to compare results. Our developed algorithm provides better results due to the new deep features extracted from the dataset compared to existing features. This research will help various predictive maintenance systems to detect false alarms, which will in turn reduce unnecessary visits of service technicians to installation sites.",
author = "Krishna Mishra and Tomi Krogerus and Kalevi Huhtala",
year = "2019",
month = "7",
day = "28",
doi = "10.5220/0007802003130320",
language = "English",
isbn = "978-989-758-378-0",
volume = "16",
pages = "313--320",
editor = "Christian Callegari",
booktitle = "16th International Conference on Signal Processing and Multimedia Applications",
publisher = "SCITEPRESS",
edition = "2019",

}

RIS (suitable for import to EndNote) - Download

TY - GEN

T1 - Profile extraction and deep autoencoder feature extraction for elevator fault detection

AU - Mishra, Krishna

AU - Krogerus, Tomi

AU - Huhtala, Kalevi

PY - 2019/7/28

Y1 - 2019/7/28

N2 - In this paper, we propose a new algorithm for data extraction from time series signal data, and furthermore automatic calculation of highly informative deep features to be used in fault detection. In data extraction elevator start and stop events are extracted from sensor data, and a generic deep autoencoder model is also developed for automated feature extraction from the extracted profiles. After this, extracted deep features are classified with random forest algorithm for fault detection. Sensor data are labelled as healthy and faulty based on the maintenance actions recorded. The remaining healthy data are used for validation of the model to prove its efficacy in terms of avoiding false positives. We have achieved 100% accuracy in fault detection along with avoiding false positives based on new extracted deep features, which outperforms results using existing features. Existing features are also classified with random forest to compare results. Our developed algorithm provides better results due to the new deep features extracted from the dataset compared to existing features. This research will help various predictive maintenance systems to detect false alarms, which will in turn reduce unnecessary visits of service technicians to installation sites.

AB - In this paper, we propose a new algorithm for data extraction from time series signal data, and furthermore automatic calculation of highly informative deep features to be used in fault detection. In data extraction elevator start and stop events are extracted from sensor data, and a generic deep autoencoder model is also developed for automated feature extraction from the extracted profiles. After this, extracted deep features are classified with random forest algorithm for fault detection. Sensor data are labelled as healthy and faulty based on the maintenance actions recorded. The remaining healthy data are used for validation of the model to prove its efficacy in terms of avoiding false positives. We have achieved 100% accuracy in fault detection along with avoiding false positives based on new extracted deep features, which outperforms results using existing features. Existing features are also classified with random forest to compare results. Our developed algorithm provides better results due to the new deep features extracted from the dataset compared to existing features. This research will help various predictive maintenance systems to detect false alarms, which will in turn reduce unnecessary visits of service technicians to installation sites.

UR - https://www.scitepress.org/ProceedingsDetails.aspx?ID=0N9+1/B4ih0=&t=1

UR - http://www.wikicfp.com/cfp/servlet/event.showcfp?eventid=82467&copyownerid=45217

UR - http://www.sigmap.icete.org/?y=2019

U2 - 10.5220/0007802003130320

DO - 10.5220/0007802003130320

M3 - Conference contribution

SN - 978-989-758-378-0

VL - 16

SP - 313

EP - 320

BT - 16th International Conference on Signal Processing and Multimedia Applications

A2 - Callegari, Christian

PB - SCITEPRESS

CY - Prague, Czech Republic

ER -