ABSTRACT
Condition-based maintenance (CBM) involves the scheduling of maintenance and replacement activities using information obtained from one or more condition monitoring (CM) processes over time. As with many reliability applications, we are interested in modelling the time remaining before failure, or residual life (RL) of a component, part or piece of machinery. Relevant prognostic techniques include; proportional hazards models (see Makis & Jardine (1991) and Banjevic & Jardine (2006)), proportional intensity models (see Vlok et al., (2004)), neural networks (see Zhang & Ganesan (1997)) and stochastic filters (see Wang & Christer (2000) and Wang (2002)).
