A preventive maintenance checklist is the most-copied document in maintenance and the least examined. Most of them are inherited, generic, and quietly wrong for the machine they are attached to, which is how a site ends up with a checklist that has never failed an item in three years. A check that never fails is not evidence that the machine is healthy; it is evidence that nobody is looking. This page is about writing checklists that produce information: specific enough to be answerable, short enough to be done properly, and built so that finding something is a normal outcome rather than an exception.
Write checks that can fail
'Check hydraulics' cannot fail; 'hydraulic oil above minimum, no visible leak at the ram seals' can. The test for every line is whether two technicians would agree on the answer, and whether the answer has ever been no. Items that cannot produce a no are decoration and lengthen a checklist people already resent.
Make it shorter than you want to
A twenty-item checklist done properly beats a sixty-item one done from memory in the van. Sixty items on a six-minute check is a form being completed rather than a machine being examined, and everybody involved knows it. Cut the items that have never found anything and put the time into the ones that have.
Leave room for what was found
The single field that turns a checklist into maintenance knowledge. A page of ticks tells you a visit happened; 'belt showing wear on the drive side' tells you whether the interval is right. Without it, the checklist can never be improved because there is nothing to improve it from.
Review the checklist against its own history
Once a year, look at what the checks actually found. Items that never fail come off or their interval lengthens; failures that keep appearing between checks mean the interval is too long or the check is looking at the wrong thing. This is the whole loop, and almost nobody closes it.
Questions people ask about preventive maintenance checklist
Should the checklist be per asset or per asset type?
Per type as a starting point, per asset where something about the installation is different. Copying one machine's checklist onto a different duty is how generic items get in.
Digital or paper?
Either, as long as findings reach the maintenance team the same day. Paper that sits in a van until Friday delays every fault it found by a week, which is usually the real argument for going digital.
Who should sign it?
The person who did it, with a name rather than initials, because a checklist is a record somebody may have to stand behind. Countersigning by a supervisor adds a delay and rarely adds information.