How to tag field service notes by product, symptom and failure
Summary
Tag every service note with product, equipment, symptom and action, then add failure mode once it is observed and mechanism and cause once they are confirmed. The scheme works for a chemical supplier's notes from lubricant, coolant, water treatment or coating visits when several engineers tag from one shared spelling list and each axis gets one tag per note.
ISO 14224:2016, edition 3, last confirmed in 2022, says its failure modes "can be used as a 'reliability thesaurus'" (ISO 14224:2016).
In the ISO 14224:2006 failure data table presented by its technical editor in 2014, failure mode is minimum data, while failure mechanism and failure cause carry no minimum mark (Selvik, 2014).
The same 2014 presentation lists 6 failure mechanism classes and 5 failure cause classes from ISO 14224:2006 Annex B (Selvik, 2014).
Microsoft Dynamics NAV 2013 builds fault reporting on a combination of 3 codes: symptom, fault area and fault (Microsoft, 2012).
This scheme is for making notes from customer visits and calls searchable. It is not an equipment reliability database. Plant owners in petroleum, petrochemical and natural gas collect reliability data under ISO 14224 itself (ISO 14224:2016). Regulated complaint and corrective action records follow their own procedures.
Which tag axes should technical service notes use?
Technical service notes should carry four tag axes, or categories of tag, on every note: product, equipment, symptom and action taken. Failure mode, mechanism and cause join them when they are known. The first four are facts the engineer has by the end of any visit; the last three are findings that may take weeks of testing.
The table shows what each axis records and when it is added. All examples are illustrative.
Axis | What it records | Example (illustrative) | When added |
|---|---|---|---|
Product | Product family or code | neat-cutting-oil | Every note |
Equipment | Machine or system type | cnc-lathe, cooling-tower | Every note |
Symptom | What the customer reported or saw | foam, odor, rust-on-parts | Every note |
Failure mode | How the failure shows on the equipment, as observed | sump-foam-overflow | Once the engineer has seen it |
Mechanism | Physical or chemical process | microbial-growth | Once tested |
Cause | Circumstance that led to it | wrong-mix-ratio | Once confirmed |
Action | What was done | concentration-adjusted | Every note |
The split between symptom, failure mode, mechanism and cause comes from ISO 14224; its 2006 edition defines failure mode as the "effect by which a failure is observed on the failed item" and failure mechanism as the "physical, chemical or other process that leads to a failure" (Selvik, 2014).
Most systems already store the customer site in a field on the note, so tag the site only where that field is missing.
Keep the axes in the same order on every note. Anyone scanning a list of notes then finds product, equipment and symptom in the same place each time.
Service note tags by situation
How far a case has gone decides which tags its note carries. Product, equipment, symptom and action go on every note, and the later axes come in as evidence arrives. The table sets the tags for a technical service note in five common situations.
Your situation | Tags to use | Why |
|---|---|---|
Every note | Product, equipment, symptom, action | These facts are known at the end of any visit or call |
Symptom seen, cause open | Add failure mode only; no cause tag | A guessed cause tag sends later searches to the wrong cases |
Cause confirmed by test or trial | Add mechanism and cause from the list | The evidence now supports them; note the test in the text |
Repeat complaint | Same tags as the first note, plus a repeat tag | Matching tags pull the earlier notes into one search |
Product change at the customer | New product tag; keep the old one in the text | Searches by the new product must not mix old results |
The open case matters most. The symptom has been seen but the cause is not confirmed, so tag the failure mode and leave the cause axis empty. An empty cause tag is honest. A wrong one hides the note from the very search that would have found it. Once the cause is confirmed, go back and add the cause tag to the earlier notes in the same case too.
For a repeat complaint, copy the tags from the first note rather than retyping them, so the spelling cannot drift between the two. A product change works the other way. The new product gets its own tag, and the note text records which product it replaced and on what date.
When should a service note get a failure mode or cause tag?
A service note gets a failure mode tag once the engineer has observed how the failure shows on the equipment. It gets a cause tag only after a test, trial or inspection has confirmed the cause.
In an illustrative coolant case, a phone report that "the coolant smells" is a symptom. When the engineer arrives and finds a black, slimy sump surface, that is an observed failure mode. A bacteria count above the customer's limit points to a mechanism. A dead leg in the piping, found on inspection, is a cause.
ISO 14224 draws the same line. In the 2006 edition's failure data table, failure mode is part of the minimum data that shall be collected, while failure mechanism and failure cause are recorded without that minimum mark (Selvik, 2014). That edition defines failure cause as "circumstances associated with design, manufacture, installation, use and maintenance that have led to a failure" (Selvik, 2014), which is something you establish, not something you see.
Mechanism and cause are easy to confuse. Maintenance documentation that builds on the ISO 14224 lists warns: "Due care is required so as not to confuse failure mechanism (describing the apparent, observed cause of failure) with failure cause (describing the underlying or 'root' cause of a failure)" (Tetra Pak). In a coolant case, microbial growth is the mechanism; poor sump circulation or a wrong mix ratio is the cause.
How do you keep service note tag spelling consistent across engineers?
Engineers spell tags the same way when they all work from one shared list of allowed tags. The list holds one preferred spelling per concept and maps its known variants to it.
Free typing drifts. In tagging research, items about television were tagged "television" or "tv," and if tags such as "cat" and "cats" are distinct, "a query for one will not retrieve both" (Golder and Huberman, 2005).
This method already has standards behind it, written for controlled vocabularies: fixed lists of approved terms. ISO 25964-1:2011, last confirmed in 2022 and expected to be replaced, gives recommendations for building and maintaining thesauri for information retrieval (ISO, 2011). ANSI/NISO Z39.19-2005 (R2010) gives guidelines for controlled vocabularies, including synonym rings and taxonomies (NISO, 2010). Commercial service software applies the same idea through code tables: Microsoft Dynamics NAV 2013 sets up symptom codes in a table and assigns them to service items (Microsoft, 2012).
Five spelling rules cover most drift:
One owner for the list; new tags are requested, not typed.
Lowercase, singular nouns, words joined by hyphens.
No abbreviations unless the abbreviation itself is on the list.
Each known variant points to the preferred tag.
Product tags use the product family, not each package size.
Lubinpla work logs accept hashtags, so engineers can put the same axes on each log and pull up past records for the same customer and symptom.
Why do too many tags make service notes harder to search?
Too many tags make service notes harder to search because each extra tag gives a note another way to match the wrong query or to be missed by the right one. Tagging research names the effects: a word with several related senses "dilutes query results by returning related but potentially inapplicable items," and synonyms mean a searcher cannot be sure "that all the relevant items have been found" (Golder and Huberman, 2005).
Take a note tagged "foam, foaming, air, air-entrainment, defoamer, sump, overflow" (illustrative). It turns up in searches for defoamer supply, air leaks and sump design. An engineer looking for past foaming cases has to read through all of them, and the next engineer may tag the same case with three different words.
The same research describes a "basic level" problem: people choose different levels of detail for the same thing, and experts choose more specific terms than others (Golder and Huberman, 2005). One engineer tags "corrosion," another "pitting," a third "rust." Set the level of detail for each axis on the shared list (symptom, for example, at the level the customer can see), and most of that spread goes away.
In practice the limit is one tag per axis, plus a repeat tag where it applies. A note that seems to need two symptom tags usually describes two problems and should become two notes.
Common mistakes in tagging service notes
The most common mistakes in tagging service notes are guessing the cause, tagging at different levels of detail and letting each engineer invent spellings. Each one breaks a search months later, when the person searching is no longer the person who tagged.
Tagging a suspected cause. A cause tag added on day one, before any test, stays on the note after the real cause is found. Put suspicions in the note text and keep the cause axis empty until confirmed.
Mixing mechanism and cause. "Corrosion" is a mechanism; "chloride carry-over from rinse water" is a cause (illustrative). Keep them on separate axes, as ISO 14224 does (Selvik, 2014).
Tagging the brand of a competitor's product. Tag the product family the customer uses instead, so the note is found by application rather than by a name that may change.
Tagging old notes with a new list without mapping. When old reports move into the same system, map their words to the current list first. The steps for turning years of reports into records you can search are in how to digitize old service reports into searchable records.
Never pruning the list. Review the tag list on a fixed schedule and merge tags that mean the same thing, keeping the old spelling as a variant.
Frequently asked questions
How many tags should a service note have?
One tag per axis is enough. Product, equipment, symptom and action go on every note; failure mode, mechanism and cause follow as they become known, plus a repeat tag where it applies. A note that needs more usually describes more than one problem.
Should tags be in the customer's words or in technical terms?
Use technical terms from the shared list for tags, and keep the customer's own words in the note text. The text preserves the original report, and the tag lets the note be found alongside similar cases.
Can we use the ISO 14224 failure mode lists directly?
Only partly. ISO 14224 covers equipment in the petroleum, petrochemical and natural gas industries (ISO 14224:2016), so its lists fit pumps or compressors, not symptoms such as foam or odor in a chemical product. Borrow its structure and write your own product-level list.