I — Datum
The Digital Calibration Certificate
II — The Problem It Exists To Solve
Ask people who run calibration software what they dislike most about it and the answer is rarely the analysis. It is data entry. A laboratory calibrates an instrument and sends back a certificate; somebody then reads that certificate and types its contents into a register, point by point. On a fleet of any size this is days of work a year, and every one of those keystrokes is a chance to put a value in the wrong row.
The certificate arrives as a PDF because a PDF is what a person reads. It is also where machine-readable data goes to die: the numbers are there, but only as marks positioned on a page, and recovering them means guessing at a layout that changes with every laboratory and every template revision.
The industry's usual answer has been an integration per laboratory — a parser for this lab's format, another for that one, each breaking whenever the lab changes its report. The Digital Calibration Certificate is the other answer: one published format that any laboratory can emit and any system can read, so the work is done once rather than once per relationship.
III — What It Carries
A DCC is an XML document with two halves, matching the two things a certificate is actually for.
| Half | What it holds |
|---|---|
| Administrative | Which instrument, identified as an instance rather than a model. Who calibrated it and under what authority. Who it was calibrated for. The relevant dates — receipt, calibration and issue are three different things and the format keeps them apart. The equipment the calibration itself relied on. |
| Results | The measurement points, each carrying its value, its unit and its uncertainty as separate structured fields rather than as one printed line. This is the half a register wants and the half a PDF cannot hand over. |
The content requirements it satisfies are the ones ISO/IEC 17025 already places on a calibration certificate. The DCC does not add obligations; it gives the existing ones a structure a machine can traverse. Two major versions of the schema are in circulation, so a reader that only handles the newer one will refuse certificates that are perfectly valid.
IV — The Units Are The Hard Part
This is the part that surprises people, and it is the reason the format needs a standard underneath it rather than a convention.
A number in a certificate is worthless without its unit. But a unit written
as text is not a unit — it is a string. kg,
KG, kilogram, kilo and
1000 g all mean the same thing to a person and five different
things to a parser. Worse, the failure is silent: a receiving system that
guesses wrong does not stop, it records a value a thousand times off and
carries on.
The DCC is written in D-SI, a machine-readable encoding of the SI in which a unit is built from escaped identifiers rather than spelled out — a prefix identifier and a base identifier, composed, with exactly one legal way to write each one. A parser does not interpret those; it either recognizes the identifiers or refuses the document.
What this buys is not tidiness. It is the ability for a system to refuse. A quantity that knows its unit as structure can be compared to another only when the two are commensurable, so an attempt to read a length against a mass raises rather than producing a number. A silent unit conversion is one of the few failures in a measurement record that leaves no trace at all — the value looks reasonable, the certificate looks complete, and nothing anywhere says the comparison was meaningless.
V — What It Does Not Settle
A format is a smaller thing than it first appears, and three questions people expect it to answer are all outside it.
| Not this | Why not |
|---|---|
| Accreditation | A DCC from an unaccredited laboratory is a well-formed document making exactly the claim a PDF from the same laboratory would make. The format carries the authority the issuer states; it does not check it. |
| Signature | Structure is not attribution. Binding a document to the identity of the person who authorized it is a separate mechanism with its own standards. What a signature proves → |
| Conformity | A certificate can state a decision rule and the result it produced. Whether a reading passes, and what a guard band does to that answer, is arithmetic performed against the measurement and its uncertainty — not something a document format decides. |
VI — Adoption, Honestly
The DCC is led by PTB, Germany's national metrology institute, and work to harmonize its use across Europe runs through EURAMET. Adoption so far is concentrated where you would expect it: national metrology institutes and accredited calibration laboratories, predominantly in Europe.
A machine shop in the United States today will almost certainly receive a PDF. That is worth saying plainly, because the case for knowing about the DCC is not that it has arrived — it is that the direction is settled and the cost of being ready is now small. The shops that meet it first will be the ones whose laboratory happens to be European, or whose customer asks for machine-readable results and will not accept a spreadsheet.
Reference material. For what SVEND does with a DCC, see verifiable calibration certificates.