A hybrid predictive maintenance approach for ship machinery systems: a case of main engine bearings

Abstract

A hybrid maintenance strategy is proposed, combining reliability-centered and condition-based maintenance strategies to prevent engine failures. A case study on the lubrication deficiency of a ship's main engine bearings was conducted. The reliability-centered maintenance phase is based on reliability-increasing rules, equipment facts, and a knowledge base. Then, a strategy is developed where system parameters are continuously monitored. These parameters are evaluated in the inference engine, and maintenance recommendations are provided to the ship operator. The investment cost of th developed system is discussed using a real engine failure case. The study shows that the strategy can improve maintenance on existing ships and system reliability at low costs. The methodology offers a cost-effective approach for the sustainability of the main engine, as demonstrated by a real crankshaft failure report. Unlike existing practices, it provides influential maintenance decisions based on evaluating system parameters within four operation ranges. Simulations show it detects insufficient lubrication faults at low engine loads, superior to current applications. The proposed approach suits different types of existing and newly built ships, engines like dual-fuel, and technologies such as semi-autonomous ships.