How to design preventive maintenance schedules
Requirements
- Available on: Orbit, Odyssey, Cosmos
- Roles: Owner, Admin, Manager
What a rule can and cannot do
A PM rule defines an interval — a count of days, months, or years — and a warning window, and attaches to one asset or to a whole category.
Intervals are time-based. Scheduling by hours run or distance travelled is not available, so a fleet schedule has to be expressed in months rather than miles.
Attach rules to categories, not assets
This is the decision that matters most.
A rule on a category governs every asset in it, including assets added later. A new pump filed under "Pumps" inherits the pump service schedule without anyone remembering to attach it.
A rule on a single asset covers exactly that asset, and every new one is a new job for somebody.
Design your categories so that things sharing a service schedule sit together, then attach the rule once.
Choosing intervals
Start from the manufacturer, then adjust from your own history.
The manual is a starting point, not an answer — it assumes conditions that may not be yours. After a year of service records you have better evidence than the manual does.
Adjust for how hard the asset works. A machine running two shifts needs a shorter interval than the identical machine running one.
Be honest about capacity. A schedule that generates more work than your team can absorb produces a permanent overdue list, and a permanent overdue list trains everyone to ignore alerts. That is worse than a longer interval honestly kept.
Setting the warning window
The warning window is how many days ahead you are told. It defaults to 30.
Set it to the lead time you actually need, which is usually longer than people first think:
- Work your own team does at short notice — a week or two
- Work needing a part ordered — the lead time plus a margin
- Work needing an external vendor booked — three to four weeks
A warning that arrives after the point at which you could still act is not a warning.
Attach checklists
Set a checklist template on the rule and every work order it raises carries structured inspection steps.
This is what makes the schedule productive rather than a tick-box exercise: a failed checklist item raises a corrective work order automatically, so an inspection that finds a problem produces a repair task without anyone having to remember.
Stagger them
Twelve quarterly rules all falling due on the first of the month is one impossible week and eleven quiet ones.
Spread start dates so the work arrives evenly.
Prioritise by criticality
Set asset criticality properly. It is not decorative — on Odyssey and above it multiplies SLA response windows, and it is what lets you triage an overdue backlog sensibly.
If everything is Critical, nothing is.
Review the schedules
Once a year, look at what the schedules produced:
- Rules permanently overdue — the interval or the resourcing is wrong
- Services that never find anything — the interval may be too short
- Failures between services — the interval is too long, or the checklist is missing something
- Rules on archived assets — archive the rule too
Rules are archived rather than deleted, so tidying up is safe and reversible.
Common mistakes
Attaching rules to individual assets. Every new asset then needs manual work.
Copying the manual without adjusting. You buy services you do not need, or miss ones you do.
Short warning windows on vendor-dependent work. The alert arrives too late to book anyone.
Creating schedules with no capacity to meet them. The overdue list becomes noise.
Never reviewing. A schedule set once and never revisited stops matching the equipment within a couple of years.
Related articles
Need Help?
If you have questions not covered in this article, our support team is here to help.
Contact Support