BEMSC — British Engineering & Manufacturing Standards Council All articles
Opinion

The Return That Keeps Returning: Service Records as an Early-Warning System

BEMSC — British Engineering & Manufacturing Standards Council
The Return That Keeps Returning: Service Records as an Early-Warning System

Photo: Kevin Larson, Fort Stewart Public Affairs Office., Public domain, via Wikimedia Commons

There is a particular kind of silence in a manufacturing business. It is the silence that follows a warranty claim once the replacement part has been despatched, the customer has been mollified and the job has been closed on the service system. The part goes back on the shelf or into the bin. The paperwork goes into a folder. Nobody asks the obvious question: what did that failure just tell us?

Usually the answer is nothing, because the information was never captured in a form that could answer it. That is not a failure of diligence. It is a failure of design — of the record, not the product.

The Signal Nobody Files

Consider the ordinary life of a service visit. An engineer attends, diagnoses, replaces, tests and leaves. The notes are written for invoicing purposes, not for analysis. They record what was done, not what was observed. And yet those notes, aggregated across a year, contain precisely the intelligence that standards committees spend months trying to assemble: which components fail in which conditions, which assumptions in a design calculation turn out to be optimistic, and which supplier declarations do not survive contact with a wet Tuesday in February.

Every manufacturer already holds this data. Very few hold it in a shape that can be read.

Why the Data Stays Silent

The reasons are practical and, in our experience, entirely reasonable. Service engineers are measured on response time and first-time fix, not on the quality of their descriptive prose. Warranty systems are built to adjudicate liability, not to generate insight. And there is a legitimate commercial nervousness about putting failure data into any system that might one day be disclosed.

To which we would add a fourth: nobody has been asked. Standards work can feel remote from the depot and the van. The engineer who has replaced the same seal four times in six months has no obvious route by which that observation reaches a committee in London.

A Modest Routine

BEMSC's proposal is deliberately unambitious, because ambitious data programmes tend to die in their second quarter. We suggest four steps.

Log. Add three fields to the existing service record: symptom as observed, operating condition at failure, and whether the failure was expected or surprising. No new system. No additional form. Three fields on a screen the engineer already completes.

Triage. Review the month's returns against a simple scale. A one-off is an incident. A repeat on the same asset type is a pattern. A repeat across different sites or operating conditions is a signal — and signals merit a named owner. The triage meeting need not last longer than half an hour, and the person chairing it should be the quality lead, not the service manager.

Anonymise. Strip customer identity, site location and any commercially sensitive detail before the data moves beyond the service function. What is retained is the technical substance: asset class, application, condition, mode of failure, age at failure. This is almost always sufficient for analytical purposes and removes most of the internal objection.

Route. Decide, for each signal, whether it is an internal matter, a supplier conversation, or something with a wider relevance that should go forward to a committee, a trade association or a live consultation. Most will be the first. A handful each year will be the third, and those handfuls are worth the whole exercise.

The Anonymisation Discipline

Anonymisation is where good intentions usually fail, because the obvious identifiers are removed and the indirect ones are not. A rare machine type, an unusual application and a specific month can together identify a customer to anyone who knows the sector. Members should apply a simple test: could a competitor, reading this, work out who it was? If the answer is even possibly yes, generalise further.

Equally, over-anonymisation destroys the signal. Removing the operating condition because it feels specific turns a useful observation into an anecdote. The art lies in keeping the technical context and discarding the commercial context.

Feeding the Loop

Where a signal appears to bear on a published standard — a tolerance that proves unworkable in service, a test method that fails to reproduce field conditions, a clause whose meaning has drifted in practice — the correct route is through the established channels: the relevant technical committee, the member's nominated representative, or a consultation response where one is open. Evidence from service data carries considerably more weight in those forums than assertion, and it is precisely the kind of evidence UK industry is often accused of not supplying.

We would encourage members to keep a short annual note of the signals they routed and the response received. Over time this becomes a record not only of what was found but of what changed as a result.

A Closing Observation

Standards are sometimes described as the codified experience of an industry. If that is true, then service records are where the experience accumulates before anyone has thought to codify it. Treating them as an administrative residue is a quiet waste. Treating them as an input to standards intelligence costs three fields and half an hour a month.

All Articles

Related Articles

The Clause That Pays for Itself: Putting Energy Performance at the Centre of the Standard

The Shop Floor as Seminar Room: A Proposal for Member-Hosted Standards Walkthroughs

The Shop Floor as Seminar Room: A Proposal for Member-Hosted Standards Walkthroughs

Reading Between the Clauses: Building Standards Literacy on the Shop Floor

Reading Between the Clauses: Building Standards Literacy on the Shop Floor