Infrastructure AdvisoryTwo assets, not one
Infrastructure organisations operate physical assets. They also operate the information needed to understand them: drawings, specifications, configuration records, commissioning evidence, maintenance history, certification and warranty documentation.
The distinction looks administrative until something goes wrong. Then a more fundamental question emerges. Can the organisation demonstrate what asset it has, what has happened to it, and what rights or obligations still attach to it?
That is a question about evidence, and it is worth separating from the question of outcome. Whether a remedy or an obligation would succeed depends on its terms, the contract it sits in, the governing law and the facts. Whether it can be evidenced at all depends on the organisation’s own records — and that part is entirely within its control.
The four fields most positions turn on
A component can be visibly installed and working while important questions remain unanswered. In practice the answerable ones reduce to four.
Identity — is this the specified item, or something accepted as an alternative at the time? Date — when did it enter service, and is that date evidenced or entered? Intervention history — what has been done to it, when, and to what requirement? Party — who performed each intervention, and under what arrangement?
Those four are unglamorous and they carry most of the weight. An organisation that can produce all four for a given item can answer nearly any question that arises about it. An organisation that can produce two is reconstructing.
Granularity, and the second life of a component
Two failure modes recur, and both are structural rather than careless.
The first is granularity. An obligation usually attaches to a component. A maintenance system frequently records at the level of an assembly, a system or a location, because that is the level at which work is planned. Where those two levels differ, the record cannot answer a question posed at component level, however complete it looks.
The second is replacement history. When an item is replaced, the record typically captures what is now installed and loses what was removed — its identity, its service period and anything that attached to it. The asset register describes the present accurately and the past not at all. That is adequate for maintenance planning and inadequate for almost everything else.
Handover does not guarantee completeness
Infrastructure information crosses more organisational boundaries than the asset does. Designers create it, contractors develop it, suppliers provide product information, commissioning teams add evidence, project teams assemble it, and operators inherit it and are expected to maintain it for decades.
Every transfer is an opportunity for divergence, and the divergence is rarely deliberate. A drawing may represent design intent rather than the as-built position. Equipment may have been substituted for sound reasons never carried into the controlled record. Commissioning information may sit in a project archive rather than the asset register. Later modifications may be properly executed and poorly reflected in the information describing them.
The result is not a missing record. It is an asset and a record that gradually cease to describe the same thing, while both remain in active use.
Warranty questions expose the problem
Warranty situations make this unusually visible.
A warranty may appear straightforward until a failure occurs. The organisation may then need to establish what was supplied, when it entered service, whether it was commissioned as required, what maintenance was undertaken, what modifications were made and whether relevant requirements were met.
A missing record does not necessarily mean that an entitlement has been lost. Nor does possession of a warranty document automatically establish that a particular claim will succeed. What either position depends on is the terms, the contract, the governing law and the facts. But incomplete evidence makes any position materially harder to establish, and it does so at the moment when the organisation has least time to reconstruct it.
The lesson extends beyond warranties. Rights that depend on asset history are only as usable as the evidence available to demonstrate that history.
Technical and contractual defect management drift apart
There is a related divergence that is easy to miss because both halves are working.
A technical defect-management process tracks condition, diagnosis, intervention and closure. A contractual one tracks obligations, notices, records and positions. They observe the same item, they are usually both competent, and they are almost never connected — so a condition can be technically closed while the evidence that would have supported a position was never captured, or a position can be maintained on evidence the technical record has since overwritten.
Keeping the two connected across an asset’s life is unglamorous work. It is also the difference between having a right and being able to use it.
Record quality changes maintenance decisions
Poor information creates operational cost long before anything is in dispute.
If configuration cannot be trusted, maintainers verify conditions physically before they work. If component identity is uncertain, spares decisions become conservative and stockholding rises. If maintenance history is incomplete, condition assessment requires additional investigation. If modification records are unreliable, every subsequent intervention takes longer to plan.
That cost rarely appears as an information cost. It appears as additional surveys, duplicated inspection, conservative maintenance, unnecessary replacement or delayed decisions — absorbed into the maintenance budget without anyone identifying what produced it.
Resolve the discrepancy, not just the document
When asset and record disagree, updating the database does not resolve the underlying issue.
The organisation first has to determine which evidence is authoritative — physical verification, commissioning information, supplier records, maintenance history, change records or other contemporaneous material. Only then can the controlled record be reconciled with the asset.
This matters more than it appears, because correcting a record can conceal the history of the discrepancy, and that history is frequently the evidence that matters. A record that now says the right thing, with no trace of what it said before or when it changed, has resolved a data problem and destroyed an evidential one. A good asset record shows the current position without erasing how it was established.
The decision point
There is an established way of framing what the information is for. The ISO 19650 information-management framework starts from the information an organisation needs to support operation of its assets, aligned to its organisational and asset-specific requirements — that is, requirements derived from what the organisation has to do, rather than from what happened to be produced.
That changes the test. The question is not whether a document exists or a field is populated. It is whether the organisation holds sufficient reliable evidence to operate, maintain and decide about the asset with appropriate confidence — and whether it knows which of its records would survive being relied on.
Where the physical asset and its record diverge, the discrepancy is an asset issue, not merely a documentation issue.
Westheath reviews asset-management arrangements where the information has to support a commercial position as well as a maintenance plan.