For Test Laboratories

Evidence captured once, at the moment of work. The Recommendation’s test procedures become a workbench your operators follow step by step; readings accrue as evidence with the operator and the timestamp attached; the test report is generated from that evidence, and your metrologists’ time goes into measurement instead of paperwork mechanics.

What you can do

Every act below was performed against the live demo by the scripted capture apparatus before it was listed; each figure names its capture date and links the live surface. An act that cannot be performed does not list.

Between dispatches, the same inbox carries your state honestly. The incoming bucket shows exactly what waits on you — and when nothing waits, the page says so plainly instead of paging stale noise.

Open the request and see exactly your share. The scope, the samples with their serials, the issuing authority’s context. Accept, or decline with a reason; silence is not a state the record allows.

Accept, and join the project’s dataspace. Accepting the assignment joins your laboratory to the Evaluation Project: the working panels (samples, assignments, the report) reveal, and what you produce accrues to the same record the IA evaluates.

See every report your laboratory has authored, with its status. Form instances, samples, acceptance state; your output is a list you manage, not a file share you search.

Register a report you performed offline, marked as what it is. Records mode: a report issued outside the platform enters as submitted to the IA, marked registered-offline; it never claims platform test-run evidence, and the IA’s review stays a real act.

Read the procedure the Recommendation actually requires. The test procedures are the Recommendation’s model, not a retyped PDF: the work instructions your bench follows derive from the same clauses the report traces to.

Bind a twin in the twin-lab workbench. Where a SMART-twin-capable instrument answers, the workbench reads its served surface by introspection and runs the guided conformance test against it; the twin’s readings auto-record. (Today the demo’s simulator runs locally; see the roadmap section.)

How it works

What this changes, and what it does not

Today a type evaluation means: read the procedures from a PDF, set up each condition by hand, record readings into a form, transcribe the form into a report, and answer findings about transcription. On the platform the evidence is captured at the moment of work and the report is generated from it; where evidence cannot be produced, the verdict reads INDETERMINATE, never a silent pass. What does not change: your measurement, your competence, your responsibility for the result. The platform removes the transcription, not the metrology.

Today (SMART) and the vision (SMART+)

What you can use, and what you can run

Your laboratory’s consoles on the hosted OIML-CS SMART Platform are yours today. Running an instance yourselves is your Member State’s entitlement to grant, quoted below from the program’s single entitlement source.

The honest questions

Does the workbench replace our metrologists’ judgment?

No. The workbench walks the procedure and captures what the operator does; the measurement, the setup, and the professional call remain the laboratory’s. Where the evidence does not support a verdict, the system says INDETERMINATE. Your laboratory never has to choose between silence and a pass.

We already have a LIMS. Is this another island?

The platform’s surfaces are data, not documents: reports, evidence rows, and dispatches are structured records. Interop is a first-class position (the standards map: ISO/IEC SMART, ReqIF, RDF/SHACL, W3C VC, AAS), and records mode lets a report authored in your own systems enter honestly. The deep LIMS integrations are roadmap work; the honest bridge is live.

What if our laboratory works for several Issuing Authorities?

The dispatch is one TestRequest per laboratory per engagement; your inbox shows exactly your share, per IA, and each accepted assignment joins only that project’s dataspace. Cross-authority visibility is the cone machinery’s job, and it is enforced, not configured per project.

We ran the evaluation before any of this existed. Do we redo it?

No. Register the offline report; it enters marked registered-offline and the IA reviews it as a real act. Nothing is retroactively digitized, and nothing claims evidence it does not have.

Where to go next