Development Of Output-Based Decision Support Maintenance Model (OBDSMM) For Production Machines

The research presented in this thesis is concerned about maintenance decisions analysis to determine when is the right time to perform maintenance, what is the appropriate type of maintenance action should be carried out and which component is required maintenance. In industry, these series of decisions analysis is essential for minimising the overall maintenance costs and avoiding the unexpected failure, where in most cases it is very expensive. In literature, many maintenance techniques were introduced and proposed towards maintenance decisions making, which all of them can be classified under time-based maintenance (TBM) and condition-based maintenance (CBM) approaches. Literature review reveals that the application of TBM is not practical in real industrial practice compared with CBM. This literature finding can be argued according to the three application criteria; data required and collection, data analysis/modelling and decision process. Although the application of CBM is more realistic in real cases (agreed by many researchers), but there are some missing part of CBM towards maintenance decisions making process. From the literature review study, it is found that most of CBM application reported only focused on the condition monitoring process rather than maintenance decisions making. Another finding is that the monitoring parameter used in most CBM research is towards physical changes signals (e.g. vibration, sound, wear, etc), where it only useful to monitor the condition/health of the targeted component. In fact, in order to make the maintenance decision, the monitoring signal to be used must represent the actual failure process of the component its self. Thus, the right maintenance decisions towards actual failure perspective of the component can be made