Log-based predictive maintenance in discrete parts manufacturing

Clemens Gutschi, Nikolaus Furian, Josef Suschnigg, Dietmar Neubacher, Siegfried Vössner

Research output: Contribution to journalConference articlepeer-review

Abstract

The performance of discrete parts manufacturing systems is heavily influenced by unplanned machine breakdowns. Predictive maintenance allows for the conversation of unplanned machine breakdowns to scheduled corrective maintenance actions. We present a data-driven approach for estimating the probability of machine breakdown during specified time interval in the future. Machine learning algorithms are utilized for a specific use-case which is based on real-world data-sets including machine log messages, event logs and operational information. The paper describes applied data-mining, feature-extraction and machine learning methods and concludes with results indicating that machine failures can be reliably predicted up to 168 hours in advance.
Original languageEnglish
Pages (from-to)528-533
Number of pages6
JournalProcedia CIRP
Volume79
DOIs
Publication statusPublished - 2019
EventIntelligent Computation in Manufacturing Engineering: Innovative and Cognitive Production Technology and Systems - Island of Ischia, Ischia, Italy
Duration: 18 Jul 201820 Jul 2018
http://www.icme.unina.it/ICME%2018/ICME_14.htm

Keywords

  • predictive analytics
  • Predictive maintenance
  • random forest
  • ensemble prediction
  • Machine learning
  • Probability of failure estimation
  • maintenance
  • Maintenance Management
  • data mining
  • feature engineering
  • feature selection
  • Remaining useful life
  • Ensemble prediction
  • Random forrest

ASJC Scopus subject areas

  • Artificial Intelligence
  • Industrial and Manufacturing Engineering
  • Management Science and Operations Research
  • Control and Systems Engineering

Fields of Expertise

  • Mobility & Production

Fingerprint

Dive into the research topics of 'Log-based predictive maintenance in discrete parts manufacturing'. Together they form a unique fingerprint.

Cite this