A semi automated model for improving naval vessel system reliability and maintenance data management

Research output: Contribution to conferencePaperpeer-review

27 Downloads (Pure)


The demanding nature of Naval operational requirements leads to rapid deterioration and decline in the reliability of ships systems and machineries. In most Navies ships built with design life of 25-30 years begin to significantly decline in performance around 7-8 years after joining service. Consequently, these leads to frequent and often prolonged downtime and huge maintenance cost. The Nigerian Navy like other navies is equally faced with this situation in a challenging manner due to the introduction of new platforms and to non-standardised data management. Therefore, a data management approach that is focused on the use of maintenance, repair, and overhaul (MRO) data is proposed. The proposed approach will build on the existing data collection and management practiced in the Nigerian Navy while identifying alternatives for both onboard and fleet level maintenance data collection and management. In this regard a platform for a predictive machinery condition monitoring approach based on failure mode and component criticality is proposed. In this research a methodology for data collection and fault labelling is presented. Diagnostic analysis using Feedforward Artificial Neural Network classification model was used for fault classification.
Original languageEnglish
Number of pages12
Publication statusPublished - 1 Apr 2022
EventRINA Autonomous Ships conference 2022 - London, United Kingdom
Duration: 31 Mar 20221 Apr 2022


ConferenceRINA Autonomous Ships conference 2022
Country/TerritoryUnited Kingdom
Internet address


  • mission critical
  • Bayesian belief network
  • failure mode effect and criticality analysis
  • artificial neural network
  • prediction model
  • data collection
  • fleet


Dive into the research topics of 'A semi automated model for improving naval vessel system reliability and maintenance data management'. Together they form a unique fingerprint.

Cite this