The effect of multi-sensor data on condition-based maintenance policies

Heletjé E. van Staden*, Robert N. Boute

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Industry 4.0 promises reductions in maintenance costs through access to digital technologies such as the Internet of Things, cloud computing and data analytics. Many of the promised benefits to maintenance are, however, dependent on the quality of the data obtained through sensors and related technologies. In this work, we consider the effect of access to different levels of deterioration data quality, resulting in partial information about the underlying state of the system being monitored, by means of sensors, on condition-based maintenance policies. The sensors may be either internal company sensors, or more informative external sensors of which access is obtained at a cost. We analyze the structure of the optimal policy, where the actions are either to perform maintenance, to pay for external sensor information or to continue system operation with internal sensor information only. We show that the optimal policy consists of at most four regions based on the believed deterioration state of the system. The analysis allows us to numerically investigate the decision maker's willingness to pay for more informative external sensor information with respect to the level of external sensor informativeness, when compared to that of the internal sensor, and the cost thereof.

Original languageEnglish (US)
Pages (from-to)585-600
Number of pages16
JournalEuropean Journal of Operational Research
Issue number2
StateAccepted/In press - 2020


  • Condition-based maintenance
  • Maintenance
  • Multi-sensor data
  • Partially observable Markov decision process

ASJC Scopus subject areas

  • Computer Science(all)
  • Modeling and Simulation
  • Management Science and Operations Research
  • Information Systems and Management

Fingerprint Dive into the research topics of 'The effect of multi-sensor data on condition-based maintenance policies'. Together they form a unique fingerprint.

Cite this