Axiom Precision

About

Who We Are and How to Verify Us

What we build, which records we hold against each claim, and the concrete steps a buyer can take to check us before the first order.

Application EngineeringLast updated 2026-08-04

Verification artefact

Verification artefact — A verification artefact is a document or test that a buyer can request and then check against something else, such as a dimensional report whose numbers can be re-measured on the delivered part or a wiring drawing whose terminal names match the delivered label.

Application-conditioned test

Application-conditioned test — An application-conditioned test is a measurement carried out at the conditions you will actually use — medium, temperature, flow, lighting, mounting — rather than at the favourable conditions a datasheet figure is normally quoted at.

What we can put in writing, and how each item can be challenged
RecordIssued byHow to challenge itWhen to ask for it
Dimensional and functional report per serialOur own metrology benchRe-measure the delivered unit and compareBefore a first article sign-off
Material and traceability certificate for wetted partsThe steel or elastomer supplierCross-check the heat or lot number on the partFor hygienic or regulated service
Ingress test record against a named ratingOur own test rig, witnessed on requestAsk which edition of the standard was usedAny washdown or submersion duty
Application trial summary under your conditionsBoth parties jointlyRepeat it yourself with the same protocolBefore volume commitment
Table contents describe the categories of record we hold; every figure inside those records is measured per order and nothing here is a substitute for them.
Documents available on request, before any order is placed
Dimensional drawing with datum schemeper part number
Wiring and terminal diagramper output variant
Statement of test conditions behind a quoted figureper parameter
Standard edition cited for a ratingper rating
Recommended spares and consumables listper order
Language of manuals and labelsask, do not assume
Illustrative share of pre-order questions we answer in writing
100percent of answered requests
Source: Illustrative sample dataset, not a measured value.

Will you publish certificates on this site?

Not as a wall of scanned PDFs. A certificate without its scope — the product family, the standard edition, the factory that produced the units you receive — proves nothing and is the kind of content that reads as decoration, so we would rather send the specific record against your enquiry and let you check its scope, its date and its issuer.

Can we audit the plant before ordering?

Yes, and we would prefer you did, including an unannounced second visit if a long supply relationship is intended. What we ask in return is that the audit covers the questions that decide your risk: which operations are performed here rather than bought in, how calibration is traceable, what is inspected per unit versus per lot, and how a returned unit is investigated rather than merely replaced.

How much of what you publish is measured rather than claimed?

Only figures measured under a stated condition are presented as numbers, and the illustrative figures on our pages are labelled as illustrative because they are example values for a demonstration site, not results. Where we quote a standard, we name it; where we quote a datasheet figure, we state the conditions it was taken at; where we have not measured something, we say we have not measured it rather than publishing an industry average.

The single figure shown in the statistics block above, and any example value elsewhere on this page, is an illustrative sample value for a demonstration site. Replace them with measured records before publishing, and do not cite them as evidence of performance.

What we actually make

We build industrial process and factory sensors — turbidity and colour measurement for water treatment, analogue and digital outputs for instrumentation loops, distance and position measurement for assembly stations — and we build them against a use case rather than a catalogue category. That choice has a cost, which is that our pages cannot list a range in a single table with one price column, and a benefit, which is that the person who answers an enquiry is the person who specifies the mounting, the seal material and the failure behaviour.

Most of our work is the unglamorous half of instrumentation: the gasket that survives the sterilisation cycle, the cable entry that does not wick water, the zero point that is still at zero after a fortnight. These are the properties that decide whether a sensor is still installed twelve months later, and they are also the properties no datasheet summarises, which is why we write about them in the guides.

How to check us, in the order that costs you least

Start with a written answer, not a certificate. Send the conditions of your application — medium, temperature range, pressure, what the instrument has to decide, what the maintenance window looks like — and read the reply for whether the answers name their assumptions. A specification that arrives without conditions attached is a marketing document, from us or from anyone.

Then ask for the record behind the specific figure that worries you: the drift figure, the immersion rating, the chemical compatibility line. Every one of those has a document behind it, and if it does not, you have learned something more useful than the number itself. Ask which edition of which standard was used, and ask for the conditions the number was taken at.

Then run one trial unit against your own protocol, written down beforehand, including the failure you are afraid of. A trial that only tests the happy path tells you both nothing.

What happens to a returned unit

A returned sensor is a measurement that most supply chains throw away. Ours is logged against its serial number, photographed as received, and bench-tested at the conditions it was ordered for before anything is replaced. If the fault reproduces, it is analysed down to the component or the interface that gave way — a seal, a potting joint, a connector, an input stage — and the finding changes the specification, the inspection or the advice given to the next buyer. If it does not reproduce, we say so and we say what was tested, because a no-fault-found report that does not declare its conditions is usually hiding a wiring or configuration problem that will simply be bought again.

What we will not do in a document

We will not quote an industry average as if it were our measurement, we will not put a logo on a page to borrow trust, and we will not publish a number we cannot attach a condition to. If you need a figure that we have not measured under your conditions, the honest answer is a trial plan with a date, and that answer costs you nothing to ask for.