To assist RCM practitioners answer the 7 Questions , RCM provides systematic processes. Questions 1 through 4 are systematically answered via the FMEA process. To address questions 5 through 7, we turn to the decision logic tree.

For each failure mode identified in the FMEA, we need to ask;

5. In what way does each failure matter? (what is the consequence?)
6. What systematic task can be performed proactively to prevent, eliminate, or reduce the consequences of each failure mode?
7. What must be done if a suitable proactive task cannot be found?

To answer these questions, we break it down further into the following questions:

1. Is the failure mode credible (reasonably likely)?

   The FMEA uses brainstorming techniques. As such, it will often identify failure modes that are not credible or are otherwise unproductive to pursue. For example, your asset may fail due to asteroid strike. This will stop the asset's function and has a non-zero probability. However, we are unlikely to pursue such a failure mode. So the first question is to identify non-credible failure modes or simply any failure mode that will not be taken forward - stopping the process.

2. Is the failure mode hidden?

   If we are going to pursue this failure mode, the prime question is whether the failure will be observed during normal operation. The consequence and subsequent maintenance strategy is directly related to whether the failure mode is observable.

3. Does the failure mode have a safety or environmental consequence?

   We now have to determine the consequence type. The RCM process asks the question - is the maintenance strategy economically viable? However, it does not apply a monetary value to human life or the environment. If the failure mode is either safety or environmentally critical, we must provide preventative maintenance or redesign the system to remove/mitigate the risk.

4. Does the failure mode have an operational consequence?

   If it does not have a safety or environmental consequence, we must ask if it has an operational consequence. If it does, we must take into account the operational cost of such failures when deciding if the proposed maintenance strategy is economically viable. If the failure mode does not have a safety, environmental or operational consequence, it is a non-operational failure mode, and only the direct cost of maintenance is considered.

5. Is the next most desirable maintenance strategy valid?

   Now we know if it is hidden and its consequence type, we start to work our way through the possible maintenance strategies. SAE JA1011 states that each strategy should be evaluated and, of the applicable strategies, the most cost-effective should be selected. In practice, this approach is highly inefficient. Instead, a heuristic approach is used. The possible maintenance strategies are prioritized in order of desirability. The analysis will take each strategy in turn, and if the strategy is both applicable and "economically viable" it is selected. There is no need to continue with the evaluation.

   However, not all strategies are applicable to all failure modes. For example, to perform condition-based maintenance, you need to identify poor condition before the functional failure. So a suitably large P-F interval is required to allow for a scheduled inspection task to identify potential failures.

6. What must be done if a suitable proactive task cannot be found?

   The final layer of the decision logic tree asks question 7 of the 7 Questions. The options available are dependent on the answers to the questions above, but the main two options are No Scheduled Maintenance (Run-To-Failure) and Redesign. The first strategy simply lets the asset fail, and you repair/replace when it does (e.g. a household lightbulb). The second takes you back to the drawing board - how can we design out this failure mode or at least provide an opportunity to perform a maintenance strategy from the list above.

These questions are usually provided in a decision logic tree / decision diagram. Your organization should produce a decision logic tree that is applicable to your situation - the exact question text will depend on the technologies and assets you use and how your organization is structured.

For each question, you should define clear decision criteria. For example, what are the quantifiable limits for a safety consequence? Or what calculations should be used to determine "economically viable"?

An example logic tree: