Abstract
One of the main factors that influence the performance of the global fleet is the physical condition of its vessels. Well maintained ships exhibit higher reliability, profitability and safety. Even though machinery failures are inevitable, their occurrences can be minimised, or even predicted, when predictive maintenance schemes are implemented. This paper examines how a condition based maintenance approach can be used in conjunction with modern approaches from data analytics. The methodology combines the use of a quantitative Dynamic Fault Tree Analysis (DFTA) with data clustering to identify critical ship systems. Data from critical machinery operating conditions are handled and using Fuzzy Logic (FL) risk indices are obtained, based on which maintenance actions are suggested. The use of expert judgment, Original Equipment Manufacturer (OEM) thresholds and sensorial data form the input for presented methodology.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of the 2018 Smart Ship Technology Conference |
| Place of Publication | London |
| Publisher | Royal Institution of Naval Architects |
| Number of pages | 7 |
| Publication status | Published - 23 Jan 2018 |
Keywords
- maintenance
- fuzzy sets
- monitoring
- predictive