A well-run preventive maintenance program keeps the equipment that matters on a maintenance strategy suited to how critical it is and how it tends to fail. Work happens at the right frequency, so machines are neither run to failure nor serviced more often than they need. Jobs are planned with the parts and materials already identified and on hand, so a crew can complete the work in one visit. A program run this way holds availability high and keeps unplanned failures and their cost down.
That kind of program is only possible when the plant’s own records can support it. It cannot schedule maintenance on equipment that is not recorded, and it cannot plan a job without knowing what a machine is built from and which parts it takes. When maintenance falls short, the cause is usually here, in the asset and item records the program depends on, rather than in the maintenance strategy itself.
These problems show up in one of three places.
The asset registry
Problems here start with equipment that is not in the registry at all. An asset that was never registered is never scheduled, and it gets attention only when it fails. Beyond missing assets, the common issues are equipment recorded under inconsistent names and duplicate entries for the same asset. The functional-location hierarchy also often fails to match how the plant is actually arranged. When the hierarchy is wrong, work cannot be planned or reported against the right part of the plant, and the history that should show a reliability team where problems concentrate becomes unreliable.
The item master
Even where the asset registry is sound, the item master is often thin. Parts are frequently recorded without the manufacturer or supplier part numbers needed to identify and order the correct item. Bills of material are missing or are not linked to the assets that use the parts, so a planner cannot see what a given machine is built from. Specification data such as rating, size, and material is often absent, which means the correct spare cannot be confirmed without checking the equipment directly. Each of these gaps turns a routine job into a delay while someone locates the right part.
The maintenance program
Some failures show up in the program itself. A schedule does not fire because the asset or item it depends on is missing from the records. Work requests are split apart unnecessarily, for example separating the electrical and mechanical work on the same machine, which makes the schedule harder to coordinate. Jobs are planned without confirming that the materials are on hand, so the crew arrives and cannot finish the work. Each of these traces back to the records rather than to the maintenance strategy.
How a site can tell which layer is the constraint
Most of this can be established from records the site already holds, without a field campaign and without starting a software project.
- Reconcile the register for a single unit against the P&IDs and equipment lists for that unit. Equipment that appears on a drawing with no corresponding asset record is a register gap, and the items marked typical on the drawings are usually where the count has gone wrong.
- Pull the corrective work orders raised over the last two years and look at the tags they were written against. Equipment that has taken corrective work but carries no PM, or has no asset record at all, surfaces quickly in that comparison.
- Take five critical assets and attempt to order their spares using only the system, with no phone calls and no binders. Where that exercise does not arrive at a purchasable part number, the item master is the constraint.
- Sample the PMs that were deferred or rescheduled and code the reason for each one: materials, equipment availability, crew, access. The distribution across those categories shows where the friction sits.
- Review purchasing history for parts bought repeatedly against assets that carry no bill of materials. Those are item master gaps that have already been paid for more than once.
- Count the separate requests that one significant piece of equipment generated during the last outage, and compare that against what the work itself required.
- Check whether the equipment absorbing the most corrective effort is represented by a single record in each case. Duplicate and misspelled entries conceal the worst performers by dividing their history between two names.
Rebuilding the layers in the order they depend on
The asset register first. That means designing a functional location hierarchy that matches the way the plant is run, closing the gaps that a reconciliation of the existing sources identifies, removing duplicates, agreeing a naming standard, and enforcing it at the point of entry.
The item master second. Manufacturer and OEM part numbers attached, vendor cross references added, the specification fields that identify a spare correctly populated, and parts linked to their parent assets.
PM last. With a complete register and an item master linked to it, the schedule becomes a criticality-driven decision about which assets receive which tasks at which intervals, bundled so that the work can be planned and released with the materials staged against it.
The order matters more than the pace at which the work is done. A PM program built over a register with holes in it and an item master with gaps will look correct in the system and fail at the equipment.
What this means
The return a plant expects from preventive maintenance depends on the asset and item data underneath it. Weak records undercut even a well-designed program, producing missed schedules, wrong parts, wasted trips, and equipment that is less reliable than the program on paper suggests. Building that foundation correctly is unglamorous work, and it is what determines whether the rest of the program performs.
The same pattern is now appearing with AI. Companies are planning to apply AI to maintenance and reliability, and those efforts depend on the same asset and item data that preventive maintenance depends on. We cover that in a companion piece on the data foundation for AI in maintenance and reliability.
This is part of the work 2E does. The firm modernizes asset registration, item master, materials, and PM data when an operator’s existing records cannot support a system migration, automation, predictive-maintenance workflow, or AI application. There is more detail on the consulting and implementation capabilities around that work.
