A maintenance plan is the document that defines, for each asset, which maintenance tasks will be performed, how often, by whom, with which materials and in how much time. It is the foundation of everything planning schedules and maintenance executes: without a plan, the routine becomes reaction. This guide shows how to build a plan from scratch, in 9 steps, and how to keep it alive after the first version enters the system.
Plan, schedule and weekly program: each in its place
The plan defines the content: the tasks and their trigger rules. The schedule distributes that content across the calendar, balancing the load over the weeks. The weekly program closes the agenda with the crew assigned. Mixing the three produces the classic symptom: a technically correct plan nobody can execute, because the load all landed in the same month.
Building the maintenance plan in 9 steps
1. Asset inventory and hierarchy
A complete list of what exists, organized as a tree: area, system, equipment, component. Without hierarchy, history gets lost at the wrong level. If the plant runs SAP, this tree is the module's functional locations and equipment.
2. Criticality of each asset
Not every machine deserves the same plan. The criticality matrix crosses impact (safety, production, cost, quality) with each asset's reality and splits the fleet into classes A, B and C. The class defines the investment: class A assets get a dense strategy, class C can live well with planned corrective.
3. Strategy per asset
With criticality in hand, choose each asset's intervention trigger among the types of maintenance: preventive by time or usage, predictive by condition, planned corrective for what can fail without consequence. The decision is per failure mode, not per fashion.
4. Task definition
Three sources feed the task list: the manufacturer's manual (filtered, not copied), the asset's own failure history and technical analysis (for critical assets, RCM-grade rigor pays off). Every task needs an executable description: "check pump" is not a task, "measure vibration at the drive-end bearing and compare with the 4.5 mm/s limit" is.
5. Frequencies and triggers
Interval by time (weekly, monthly, semiannual), by usage (operating hours, cycles) or by condition (measurement limit). The common mistake is inheriting the manual's frequency and never revisiting: manufacturers size for the worst usage scenario, and your context rarely is it.
6. Resources, materials and time
Each task carries estimated hours, craft, tools, materials and required safety permits. That is what lets planning and control turn plan into schedule without storeroom surprises.
7. Schedule and load balancing
Tasks distributed across the calendar, leveled against the team's real capacity. The classic format is the 52-week plan, which exposes overloaded months at a glance.
8. Loading into the system
A plan in a spreadsheet is a plan at risk: without automatic triggering, discipline depends on memory. In SAP PM, the plan becomes strategy, task list and maintenance plan, and preventive orders are born on their own; the full routine is in industrial preventive maintenance.
9. Run, measure and revise
A good plan is a living plan: plan compliance and schedule adherence measure execution (formulas in the maintenance KPIs guide), and field findings feed back into tasks and frequencies. A preventive that never finds anything is a candidate for a longer interval; a failure between preventives calls for a shorter one or condition monitoring.
The preventive maintenance plan: the core of the document
In practice, most of the plan is preventive: the fixed-interval interventions that hold the failure curve down. A solid preventive plan has, for every A and B asset: tasks with procedures, justified frequencies (not inherited), execution checklists and measurement points where applicable, so the preventive builds a history beyond mere execution. The strategy is covered in the preventive maintenance guide; executing it in the field with checklists, including offline, is what preventive maintenance software solves.
The mistakes that turn plans into paper
Four patterns repeat in plan audits: copying the manufacturer's manual whole (instant overload and disbelief), a plan without criticality (everything is priority, so nothing is), frequencies frozen for years (the copy-paste plan) and a plan without an owner, which nobody revises because revising is nobody's job.
Maintenance plan template: the minimum structure
A functional template has these columns: asset (tag and description), component, executable task, maintenance type, frequency and trigger, craft, estimated hours, materials, execution condition (running or stopped) and procedure reference. Execution is recorded on work orders, not on the plan; the downloadable work order templates complete the document kit.
Frequently Asked Questions
What should a maintenance plan contain?
For each asset: executable tasks, the maintenance type, the frequency with its trigger (time, usage or condition), craft and labor hours, materials and execution conditions. All loaded in a system that triggers the orders on its own.
What is the difference between a maintenance plan and a schedule?
The plan defines the content (tasks and trigger rules); the schedule distributes that content across the calendar, balancing the load week by week. A plan only becomes routine when it gains a schedule and a weekly program.
How do you define task frequencies?
Start from the manufacturer's manual as an initial hypothesis and adjust by history: preventives that never find defects are candidates for longer intervals; failures between preventives call for shorter ones or condition monitoring. Frequency is a living decision, not an inheritance.
Where does the maintenance plan live in SAP?
In the SAP PM module: the strategy defines the cycles, the task list carries the operations and the maintenance plan connects everything to the asset, generating the period's preventive orders automatically.
How often should the plan be revised?
A full annual revision per asset class, plus event-driven revision: every relevant failure and every empty preventive (no findings) should trigger the question "is this asset's plan right?".
