Conditioned maintenance – predictive analysis in maritime propulsion engines

Condition Based Maintenance (CBM) finds today ideal conditions for its implementation: current development of sensor networks, low-cost wireless data transmission and sensors with local processing capacity; all this coupled with the need for more efficient and effective maintenance policies. The method we propose is based on statistical processing of signals and consists in applying control charts to monitor system condition in the context of CBM, aiming the reduction of failure rate, increasing both reliability and availability of maritime equipment. Control charts, being a primary tool for control and monitoring of processes, can be used to control variables such as oil pressure and temperature. This will allow us to distinguish between common and special causes of data variation.With this methodology we intend, initially, to estimate the statistical parameters of signals, by applying the Standard Deviation Chart. In a second stage, for monitoring and modeling the behavior of equipment, we applied modified CUSUM (Cumulative Sum) and EWMA (ExponentiallyWeighted Moving Average) charts.