Validation toolkit
Use the four CSV files in order. For every row complete value, owner, status and notes; resolve missing mandatory evidence before moving to the next file, then retain the reviewed version with the project record.
Validation toolkit
Use TM-VT-01 to agree the reference, TM-VT-02 to define the application, TM-VT-03 before sampling and TM-VT-04 after the model is fixed. Keep all four under the same project identity.
Step 1 — record material/grade, result basis, method/revision, preparation, drying/endpoint, calculation, storage, duplicates, laboratory owner and approval. Complete value, owner, status and notes for each row.
TM-VT-01 · Rev 1 · 2026-08-11 · CSV · UTF-8 · English, French, Spanish and Chinese fieldsDownload CSV templateStep 2 — capture site/line, materials and suppliers, moisture and variable ranges, process stage, geometry, environment, safety/contact constraints, interfaces, decision owner and fallback. Attach photo references in notes.
TM-VT-02 · Rev 1 · 2026-08-11 · CSV · UTF-8 · English, French, Spanish and Chinese fieldsDownload CSV templateStep 3 — define population, increments, positions/depths, timestamps, online window and delay, sealing, composite rule, custody, laboratory route and deviations. Assign the collector and reviewer before sampling.
TM-VT-03 · Rev 1 · 2026-08-11 · CSV · UTF-8 · English, French, Spanish and Chinese fieldsDownload CSV templateStep 4 — record scope, dataset/model versions, calibration/validation split, count and range, criteria, bias/residual review, grouped results, failures, limitations, corrective actions and approval. Publish no performance result until review is complete.
TM-VT-04 · Rev 1 · 2026-08-11 · CSV · UTF-8 · English, French, Spanish and Chinese fieldsDownload CSV templateBefore the project starts
The downloadable files connect application definition, reference method, representative sampling and validation results. Keep them under one project identity so assumptions made during selection remain visible during installation and acceptance. Complete the application form before choosing hardware, agree the laboratory method before collecting calibration data, use the sampling plan during production, and reserve the validation report for data that were not used to adjust the model. Together they create a traceable path from the original process question to the accepted measurement and its operating limits. Review the completed pack at handover and keep it with the installed system throughout routine verification and future recalibration.
Application engineering, quality, site engineering and automation should each own the fields they can verify.
Date changes to material groups, geometry, sample method, coefficients, exclusions and acceptance rules.
Link drawings, photographs, laboratory reports, raw comparisons and signed decisions instead of copying conclusions alone.
Validation toolkit
Send the four reviewed files with the model or application reference so engineering can check remaining information before configuration, commissioning or acceptance.