Planned preventive maintenance: the schedule is the whole system

Planned preventive maintenance means servicing equipment on a defined interval rather than waiting for it to stop. Almost nobody disagrees with the idea, and most implementations quietly collapse anyway — not because the maintenance is hard, but because the schedule was built for an imaginary team with more hours than yours. A plan that generates two hundred tasks a month for two engineers is not a maintenance system; it is a monthly reminder that you are behind.

What a planned preventive maintenance plan is built from

• The asset list, with enough identity to be unambiguous — a tag number, not "the compressor in the back".
• A criticality rating per asset, because you cannot service everything at the same intensity and pretending otherwise is how the important items get missed.
• The interval and its basis: calendar time, running hours, cycles or throughput. Getting the basis wrong is the most expensive error in the whole exercise.
• The task itself, specific enough that two different people do the same thing — "inspect" is not a task, "check belt tension against 12 mm deflection" is.
• The time it takes and the parts it consumes, so the plan can be checked against capacity before it goes live.
• Who is competent to do it, where competence is formally required.

Choosing the interval

Three sources, in descending order of reliability, and most plans use only the first.

1. The manufacturer’s recommendation, which is a starting point written for average duty in an average environment.
2. Your own failure history, which is better than the manual once you have two years of it — this is what the maintenance log is actually for.
3. Regulatory or insurance requirements, which are not negotiable and should be flagged separately in the plan so nobody optimises them away.
4. Adjust from calendar to usage where duty varies: a machine running one shift and one running three do not wear at the same rate, and a calendar interval quietly under-serves the second.
5. Review the intervals annually against what the failures actually were, and be willing to lengthen an interval as well as shorten it.

Size the plan to the hours you have before you turn it on. A plan that is 60% completed every month teaches the team that the schedule is advisory, and once that belief sets in, the critical tasks slip along with the trivial ones. Fewer tasks that all get done beats a complete plan that nobody finishes.

Proving it happened

• A completion record per task with the date and the person, because "we do it monthly" is not evidence to an insurer or an inspector.
• Readings captured as numbers rather than pass or fail — a bearing temperature trend predicts a failure, a tick does not.
• Findings raised as their own work rather than noted in a comment, or the defect discovered during a service is the one nobody fixes.
• Parts used, linked back to stock, so the consumption is visible before the shelf is empty.
• Deferred tasks recorded as deferred with a reason, which is the difference between a managed backlog and an invisible one.

Where the plan lives

Ettex Sheets is enough for most small and mid-sized operations: a row per asset and task, with the interval, the last completion and the next due date, sorted so this week’s work is a filter rather than a search. It sits naturally next to the asset register, which is where the identity and criticality of each item should already live. Ettex is not a CMMS — there is no work-order routing, no mobile technician app and no condition monitoring. If your plan runs to thousands of tasks across multiple sites, buy the dedicated tool; below that, the dedicated tool is usually abandoned within a year and the spreadsheet survives.