Investigating ship system performance degradation and failure criticality using FMECA and artificial neural networks

The goal of all maintenance methods is to eliminate failures or reduce their occurrence. Ex-tended downtime on key ships systems such as power generation plants can lead to undesirable consequences beyond economic and operational losses, especially considering naval vessels. One solution to overcome this challenge is through a system-specific analysis that identifies the most critical component and possible causes of delays be it technical or logistics. In this regard, this paper presents a methodology using FMECA approach that adopts the risk priority number differently to identify Mission Critical Components. This was supported with ANN classification using unsupervised learning to identify patterns in the data that signifies the onset of performance degradation and potential failures onboard an OPV. The study has identified some critical components and failure patterns that contribute to extended downtime based on survey and machinery maintenance reports. Recommendations were provided on preventing/mitigating the failures and how to prioritize existing ship systems maintenance