Assigning machines to incomparable maintenance strategies with ELECTRE-SORT

Alessio Ishizaka, Philippe Nemery

Research output: Contribution to journalArticlepeer-review

189 Downloads (Pure)

Abstract

The increasing demand on productivity and quality requires machines to be constantly available for production. It is therefore crucial to develop an adequate maintenance programme. To facilitate this, several criteria need to be considered, such as: downtime, maintenance frequency, spare parts costs, bottleneck impacts, etc. In the literature, a strategy is selected for each machine with a multi-criteria decision choice method. However, before making an informed decision, each strategy needs to be tested on each machine and then their performances evaluated with a multicriteria decision method. This is time-consuming, inefficient and often unfeasible. As machines׳ performances are usually systematically collected by industries, a much more practical approach is to assign machines to a maintenance strategy. This is referred to as a sorting problem. However, this problem cannot be solved by existing multi-criteria sorting methods because maintenance strategies cannot always be completely ordered: incomparable strategies exist. Recently, a Decision Making Grid was proposed to allocate machines to incomparable strategies. However, this technique can only be applied to problems with two criteria. In this paper, we have developed ELECTRE-SORT, a new sorting method that is able to consider an unlimited number of criteria in order to assign machines to incomparable strategies. A case study illustrates that ELECTRE-SORT provides more precise and flexible maintenance strategies than the Decision Making Grid.
Original languageEnglish
Pages (from-to)45-59
JournalOmega
Volume47
Early online date27 Mar 2014
DOIs
Publication statusPublished - 1 Sep 2014

Keywords

  • Maintenance
  • Multi-criteria decision making
  • Incomparable strategies
  • Sorting; ELECTRESort

Fingerprint

Dive into the research topics of 'Assigning machines to incomparable maintenance strategies with ELECTRE-SORT'. Together they form a unique fingerprint.

Cite this