Inspection and Test Records (ITRs) and Verification Evidence

What an ITR should prove

Inspection and Test Record is used here as a broad term for a controlled check sheet, inspection record or test-result form that supports completion and acceptance. Project terminology may differ.

A credible record allows a reviewer to establish:

• what equipment, tag, system or boundary was examined;
• which requirement and document revision governed the activity;
• what was observed, measured or demonstrated;
• whether the result met the stated criterion;
• which test equipment and temporary arrangements were used;
• which exceptions, failures or retests occurred;
• who performed, witnessed, reviewed and accepted the work;
• where the accepted evidence can be retrieved.

An ITR is normally an evidence record, not a substitute for a work method. Where the activity requires a controlled sequence, risk controls, isolations, test configuration or restoration instructions, those should be established in the appropriate approved procedure or work instruction and referenced by the ITR.

Define the exact scope and boundary

The record should identify a scope that can be inspected, tested and accepted without ambiguity. Begin with the project or asset, system and subsystem, equipment or tag, location and the applicable completion stage.

Define, as applicable:

• the physical and functional limits of the inspection or test;
• included components, circuits, loops, lines, interfaces or software functions;
• upstream, downstream and package boundaries;
• excluded equipment and separately controlled work;
• the starting configuration and required final condition;
• temporary supplies, instruments, jumpers, blinds, hoses or simulations;
• the controlled drawings, datasheets, specifications and procedures that define the scope.

Do not make one generic form cover several unrelated tags or activities merely to simplify reporting. If a common record is appropriate, each item must still have a distinguishable result and any exception must be traceable to the affected item.

Changes to a tag, boundary or acceptance basis after execution should be reviewed through the project’s document and change-control process. An annotation should not silently extend accepted evidence to work that was never inspected.

Write measurable acceptance criteria

Each check should connect four elements: the requirement, its approved source, the observed or measured result and the acceptance decision. Criteria may be quantitative, visual, documentary or functional, but they must be specific enough for another competent person to reach the same decision from the record.

Use the issued design, approved project specification, applicable code, manufacturer requirement or approved test procedure as the technical basis. The ITR should not invent a tolerance or operating limit, and a generic template should not override the governing source.

If a criterion is changed, record the authorised change and establish whether earlier results remain valid. Do not edit the criterion after a failure simply to make the recorded result pass.

Control N/A, blanks and deferred checks

Blank boxes and unexplained dashes do not show whether a check was completed, missed or considered irrelevant. Define the permitted result statuses and use them consistently.

An N/A decision should never be used to hide inaccessible equipment, incomplete work, a missing document or an unavailable witness. Where several identical checks are marked N/A, review whether the template or its allocation to that equipment type is wrong.

Record test-equipment and calibration evidence

Where a result depends on measuring or test equipment, identify enough information to demonstrate that the equipment was suitable and in an accepted calibration or verification state at the time of use.

Record as applicable:

• equipment type, unique asset number and serial number;
• manufacturer and model where needed for unambiguous identification;
• measurement range, resolution and accuracy required for the activity;
• calibration or verification certificate number;
• calibration date and due date, or the approved validity basis;
• status label condition and any pre-use or post-use check;
• measurement units, test connections and relevant environmental conditions;
• software, firmware, leads, reference devices or accessories that materially affect the result.

A current calibration label alone does not prove that an instrument has the correct range, accuracy, configuration or connection for the measurement. Suitability should be established against the approved activity.

If equipment is later found damaged, out of calibration or outside its approved verification interval, assess the validity of all affected results. Preserve the original records, document the impact review and repeat the work where the authorised technical or quality process requires it.

Make witness and hold points explicit

Required intervention points should be agreed in the inspection and test plan, procedure or other governing document before execution and reflected clearly on the ITR.

For each point, define the required notice, evidence, authority and release method. Record dates and names legibly. Distinguish a signature confirming attendance from one confirming technical acceptance; the responsibility should not be inferred from an unlabelled signature box.

If a required witness or hold point is missed, stop the affected progression and obtain an authorised disposition. Whether the evidence can be reviewed retrospectively, requires supplementary examination or must be repeated depends on the approved inspection plan and the nature of the requirement.

Keep failures visible through retest

A failed result is part of the technical history. Do not erase it, overwrite it with a later value or mark the original line as N/A after corrective work.

1. Record the actual result. Include the value or observation, date, equipment used and fail status.
2. Place the item in a controlled state. Prevent dependent acceptance or work from proceeding where the result affects safety, test validity, protection, boundary control or function.
3. Raise the appropriate exception. Link the project’s punch, non-conformance, technical-query, defect or other authorised control as applicable.
4. Complete authorised corrective action. Keep the repair, design change or concession under its governing process rather than hiding it in a comment field.
5. Retest against the approved criterion. Record a new result, date, personnel and test equipment, clearly identified as a retest.
6. Review and accept the complete history. Link the original failure, corrective evidence, retest and final disposition.

Limit a retest to only part of the original scope only when that boundary is technically justified and authorised. Corrective work may invalidate other results or require restoration checks. A “pass with comments” status should not conceal a result that fails a mandatory criterion.

Build traceability from the register to the dossier

Verification evidence should be retrievable in both directions: from a system or tag to its required records, and from an individual record back to the accepted boundary and governing requirement.

A practical traceability chain is:

Dossier index → ITR register and unique record ID → system, tag or boundary → criterion and controlled source → result and test equipment → exception or retest → authorised acceptance

The master register should distinguish records that are expected, issued, in progress, failed, submitted, returned, superseded and accepted according to the project’s defined workflow. It should also identify the current revision and the dossier or turnover-package location.

Useful controls include:

• one unique identifier for each controlled record;
• a defined relationship between templates, executed records and revisions;
• allocation by system, subsystem, tag and record type;
• links to exceptions, corrective evidence and retests;
• protection against duplicate or orphaned records;
• clear superseded status without destroying the audit trail;
• quality checks for missing pages, attachments, signatures and certificate references;
• an updated dossier index at each acceptance gate.

Minimum fields for a useful ITR

The exact form should reflect the equipment and activity, but the following information provides a practical starting point.

Use role-labelled sign-off fields rather than a collection of anonymous signatures. Electronic records should apply equivalent identity, authority, revision, audit-trail and retention controls.

Review the evidence before acceptance

A record can be complete administratively and still be weak technically. Review the executed ITR before it is accepted into the dossier.

• Does the record identify exactly what was checked?
• Does the field scope agree with the system and tag allocation?
• Are criteria explicit and linked to current controlled sources?
• Are actual observations and values recorded rather than only ticks?
• Are units, test conditions and measurement equipment clear?
• Are N/A entries justified and authorised as required?
• Are all failures, concessions and retests visible?
• Were hold, witness and review points correctly released?
• Do names, dates, roles and signatures correspond to assigned authority?
• Can every attachment and certificate be retrieved?
• Does the final status agree with the evidence and field condition?

Common weak-record patterns

ITR design and acceptance checklist

• Define the inspection or test purpose and completion stage.
• Identify the system, subsystem, tag and exact boundary.
• State inclusions, exclusions, interfaces and temporary arrangements.
• Reference controlled drawings, datasheets, specifications and procedures by revision.
• Write a clear criterion for every required check.
• Provide fields for actual observations, values and units.
• Define permitted result statuses and N/A controls.
• Identify required test equipment and suitability information.
• Capture calibration or verification evidence valid at the execution date.
• Mark hold, witness and review points visibly.
• Label every sign-off field by role and authority.
• Record the starting and final condition where relevant.
• Keep failed results visible and linked to their exception controls.
• Record retests separately against the approved criterion.
• Confirm corrective work has not invalidated other accepted results.
• Check that all attachments and certificate references are retrievable.
• Reconcile the record with the field condition and master register.
• Update the dossier index and accepted completion status.
• Retain superseded records and revisions according to the approved document-control process.

Make the evidence usable later

ITRs do not need unnecessary complexity. They need enough precision to show what was examined, which requirement applied, what result was achieved and who had authority to accept it.

When scope, criteria, results, equipment, exceptions and indexing are controlled as one evidence chain, the record can support completion reviews, commissioning decisions and handover without reconstruction from memory.