Early Detection of Faillure in Conveyor Chain Systems by Wireless Sensor Node

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Authors

  • Ghada Bouattour
  • Lidu Wang
  • Sajidah Al-Hammouri
  • Jiachen Yang
  • Christian Viehweger
  • Olfa Kanoun

In the ever-growing industrial landscape, the early detection of failures in machines with high accuracy becomes more and more crucial and essential to safe and dependable operations. Based on this concept, a machine learning algorithm is investigated for early detection of failure for conveyor chain systems. The proposed approach is based on the integration of wireless sensor nodes in the conveyor chain to measure the vibrations. The collected data has been acquired in different working conditions including slight imbalances as well as early failure scenarios that do affect the plastic chain. The data was collected using a conveyor chain with a length of 2 meters and with programmable speed and movement scenarios. Furthermore, different loads and forces have been considered during the data collection to mimic real applications in the lab. The selection of features to avoid correlation between them is considered. After comparison between different machine learning algorithms, the C-SVM algorithm is selected with an accuracy of 96.5%, which guarantees high precision and selectivity to the failures.

Original languageEnglish
Title of host publicationIEEE SENSORS 2023 : Conference Proceedings
Number of pages4
Place of PublicationPiscataway
PublisherInstitute of Electrical and Electronics Engineers Inc.
Publication date2023
ISBN (print)979-8-3503-0388-9
ISBN (electronic)979-8-3503-0387-2
DOIs
Publication statusPublished - 2023
Externally publishedYes
Event2023 IEEE SENSORS, SENSORS 2023 - Vienna, Austria
Duration: 29.10.202301.11.2023
https://2023.ieee-sensorsconference.org

Bibliographical note

The authors would like to thank the German Federal Ministry for Economic Affairs and Climate Action and AIF for funding the project WearTrack within the Central Innovation Program for SMEs (ZIM). Also, the authors would like to thank the DAAD for funding the project “Promotion of Higher Education in Biomedical Engineering” with the grant number: 57612192.

Publisher Copyright:
© 2023 IEEE.

    Research areas

  • Early Failure Detection, Industry 5.0, Machine learning, Wireless Sensor Node
  • Engineering