Maintenance is a generic term referring to a variety of actions on all kinds of assets, different life stages, forms of detoriation, available resources financial, information and knowledge. Maintenance therefore comprises a multitude aspects. It is not more than logic that there is no general model or abbreviation covering all possible aspects of maintenance. In reality there is little knowledge in maintenance departments on which models are suited for which practical problem nor which phenomena are really driving problems. Maintenance Management systems mainly store accounting information (i.e. resources, costs) on planned or curative events. The use of the tacit knowledge contained in the maintenance technicians is hardly stored or elicit.
An essential part of a maintenance knowledge in condition based maintenance is the pace of deterioration and how that can be influenced by which policy. Maintenance actions can only be optimized if they address relevant deterioration patterns and failure mechanisms. Statistical averages, like failure rate, will be inappropriate for most condition based maintenance actions. Analyzing data without knowledge of the underlying failure mechanisms lead to wrong results. Hence, structured data have to be collected using a well-defined hierarchical structuring and possible failures for each critical component. This is, however, mostly not the case, except perhaps in the airline industry. Another aspect in doing so is that under good maintenance practices there is little failure data available. If a failure occur repeatedly, one may either change the system design or its operation to prevent failures, with the result that data collection has to start again.
Most maintenance is executed following the principle of ‘the best decision given the certain problem and available information’. A better decision making process has to be balanced against the costs to collect the required data. The potential savings on sophisticated data driven decision making are mostly just too low to justify. If feasible, data driven decision making can results in a lowering of the indirect costs, from which other parts of the organization benefit, but not maintenance itself. Since this type of savings is less tangible, it is less convincing to higher management. In general, condition based maintenance often focus on the wrong type of maintenance i.e., planned revisions and overhauls.