Redesign is standard output from RCM.
It comes into play when the requirements of the asset exceed it's inherent reliability. In layman's terms, if we want it to deliver, say, 15 HP and it is a 10 HP motor, then it is not fit for purpose. Period.
And there is no form of maintenance known to man that will make it do what the users require of it. That would be a modification, which although sometimes managed by maintenance, is not actually a maintenance task.
The standard way of dealing with this during the RCM process is to note it, and to refer it on for redesign. RCM itself does not cater for redesign, but will determine when one is required.
The problem here is three fold...
1st, the RCM team continues working on the analysis. Most times there is no issue with this, however if the redesign is required because of something to do with the principal function, then the asset often has no chance of meeting the user requirements. Regardless of what else is done to it.
2nd, the actual redesign task is often left to someone within the company to get on with, and often never actually occurs.
And 3rd, no-one is checking whether the redesign is actually completed, what the implications are for the remainder of the RCM analysis, including the now modified asset.
The answer is as simple as it is hard to carry out. The redesign MUST fall within the scope of the RCM project!
When this occurs there are a positive implications, in particular
a) The task can be assigned to Mechanical / Electrical / Instrumentation / Piping (Whatever) Engineers as required. The project can then be monitored, managed through as part of the overall project plan, and delivered in a guaranteed fashion.
This is the only real sure fire way to ensure the integrity of the final result, instead of relying on distinct departments to carry out their responsibilities. (When they have totally separate priorities normally)
b) The team can judge whether the redesign has fatal implications for the analysis effort. By that I mean, will the failure to meet this particular function be the cause of significant operational or safety (in particular) limitations?
If so then they can stop, put it on hold and move onto the next asset / asset system.
If not then they can continue down the path with the one they are working on.
If you are reading this thinking, "Hmmm. Seems like he thinks that the RCM initiative should be pretty substantial", then my answer to you would be "Yes I do". Particularly if it is greater than a one or two asset analysis.
The days of RCM being shorthand for populating database systems are well and truly behind us.
More and more companies are rapidly realizing that this approach DOES NOT impact greatly on asset performance. As they do they are also starting to realise that development of failure management strategies is not an important step, it is the primary responsibility of maintenance and asset managers.