What are the key programme success measures that need to be met in implementing for an enterprise wide maintenance system?

like any program the key questions are why, what and how.
assume we already know why do we need a maintenance system.
so the second step is what parts of our enterprise do need maintenance program? we define the extent of maintenance system.
define functional locations and equipment IDs.
in this point we strictly need to cover all of critical equipments and so on.
the second step returns to how by means according to the characteristics and specification of equipments in our functional locations we must provide standard operation procedures.
these are prerequisites in every Enterprise asset maintenance systems.

1Creating the tab of several failures ; types-causes-results2Etablishing a strong bridge between the legacy payroll and the human resources base3Technicians and end-users training and reskilling

If I understood the question correctly I believe one of the answers would be replaning capability.
To answer this question I believe that it’s necessary to consider two things from it, the need for performance indicators and business process management.
These two are normally inseparable and should be the result of systematic approach towards data acquisition and ETL, hence having all the necessary analytical information to build a model.
This might mean that things could get quite complex, because enterprise wide maintenance might mean a scheduling problem for24x7 rotation and/or a resource allocation problem and/or supply chain problem.
There are also other issues when building these kind of models, namelly they need a very long lead time which mean that after completion they could be outdated.
But on the other hand if one has the high quality HR to develop such a model and sufficient computer power one can obtain a cutting edge advantage, because replanning is always available and contingencies could be overcome in a swift fashion.
The following article is amongst many other succes stories and adresses severall issues that fall in the maintenance domain and probably could be adapted to the question’s subject: http://www.scienceofbetter.org/can_do/success_stories/Interfaces/1526-551X-R.pdf I hope I was of any help Thank you