A maintenance checklist that has never failed an item is not evidence of a healthy site. It is evidence that the checks are unanswerable, too long to do properly, or both. This is the commonest condition of maintenance checklists everywhere: inherited from a previous system, generic to an asset type rather than to the asset, and long enough that completing them honestly would take longer than anybody has. This page is about writing checks that can produce a no, keeping the list short enough that they get a real answer, and pruning it every year on the evidence of what it has actually found.
Every item must be able to fail
'Check the motor' cannot fail. 'No excess play in the motor shaft, no discolouration on the terminal block, running current within 10% of nameplate' can, and two technicians would answer it the same way. That is the test: specific enough to be answerable, and answerable no.
Shorter is more honest
A fifteen-item checklist done at the machine beats a fifty-item one completed in the van. Everyone involved knows which is happening, and the long list is the one that produces uniform ticks. Cut items that have never found anything and spend the time on the ones that have.
Leave space for what was found, always
Even on a check sheet, even on paper. A page of ticks records attendance; one line of description records maintenance knowledge. Without it the checklist cannot be improved because there is nothing to improve it from.
Prune annually against the findings
Once a year, look at what each item has found. Items with no findings across many cycles come off or lengthen; faults appearing between checks mean the interval is too long or the check is looking in the wrong place. Almost nobody closes this loop, and it is where a checklist stops being inherited and starts being yours.
Questions people ask about maintenance checklist
Should checklists be per asset or per type?
Per type to start, per asset where the installation differs. A dusty room, a wet room or continuous duty change what is worth checking.
How long should a check take?
As long as the items honestly need. If a fifteen-minute list is being done in four, the problem is the list rather than the technician.
Who owns the checklist?
The maintenance manager, with technician input, and it should be dated and versioned. A checklist nobody owns is one nobody can change, which is how generic items survive for years.