When a Perfect NDT Record Is a Lie: Verifying Inspection Documentation

Nondestructive testing (NDT) programs teach every technician the same rule: "In our industry, if you did not write it down, it did not happen." The harder question, one veteran examiner argues, is what happens when a record is written perfectly and is not true — because a missing record raises an alarm, while a polished false one gets filed, relied upon, and trusted by people who will never see the part.

The argument appears in a first-person analysis by a holder of an Associate of Applied Science degree in Nondestructive Examination Technology, who writes that a fabricated record is worse than no record at all. As inspection documentation moves from paper to digital systems, the author says the shift cuts both ways: digital records can be more verifiable than paper ever allowed, and they can also make a weak record look authoritative.

What is an inspection record actually for?

An inspection record is evidence, not a form. Its audience is a person who was not there: an engineer deciding whether a component is fit for service, an auditor reconstructing a decision years later, a regulator, or a future inspector comparing today's results against tomorrow's. That reader cannot ask the technician what happened.

The test of a record is not neatness or completeness. The test is whether someone can independently trace each claim back to something that actually happened, in the order it happened, performed by a person qualified to do it.

What do digital records do well?

The author identifies five design principles that strengthen the chain of evidence, none of which depends on a particular product:

• Raw data stays attached to the conclusion. A reviewer can examine what the inspector saw rather than rely on what the inspector wrote. The author calls this the most valuable change in the shift away from paper.
• Timestamps and metadata tell a story. Automatically generated dates, times, equipment identifiers and setup parameters are difficult to reconstruct after the fact, letting reviewers ask whether the sequence makes sense.
• Audit trails show change. A system that logs who altered an entry, when, and what it said initially turns a quiet correction into a visible event. Honest corrections are routine; hidden ones are the real concern.
• Signatures can be tied to identity. Electronic signatures linked to controlled credentials, with qualification status checked as of the date of work, close the gap between "someone signed" and "the qualified person signed."
• Configuration is captured, not assumed. Recording software versions, settings and hardware explains later why two results differ.

Any record system that lacks these features deserves questions regardless of who built it.

What can digital records not do?

Digital does not mean honest, the author writes. A system records only what it is given: data can be entered after the work, carried over from a prior job, or chosen from several acquisitions because it looks better. Software can format a fabricated entry with the same precision as a genuine one.

Training transmits risk too. When an experienced inspector trains the "new kid on the block," all habits are shared, good and bad, usually introduced as "this is how it has always been done." Without care, tribal knowledge replaces best practices and jeopardizes inspection integrity. A digital system will faithfully record whatever habits it is handed.

Reviewer fatigue compounds the problem. When records arrive tidy, complete and on time, reviewers tend to stop looking — and fatigue favors the record that looks perfect. In the author's experience, genuine inspection work is rarely flawless: retests, adjusted setups, noted limitations and honest corrections are the norm. A record with no friction in it should invite more scrutiny, not less.

Culture matters most of all. No audit trail replaces an environment where people can follow procedures as written and report problems without penalty. Where those conditions are missing, technology will faithfully preserve whatever the culture produces.

A practical verification checklist

For reviewers and surveillance personnel, the author offers seven questions that travel across methods and systems:

• Does the report match the raw data? Examine underlying data for a sample of reported indications and for areas reported as clean.
• Does the timeline hold together? Compare times, equipment in use and personnel on site against the record's claims.
• Was the person qualified for this method on this date?
• Does the data look too clean? Repeated identical values, missing noise or invariant results deserve a second look.
• Is there an independent verification point — a witness, a second review, or an artifact not dependent on the record's creator?
• Can you follow every change? Corrections, reprocessing and late entries should be visible and explained.
• Does the work as recorded match the procedure as written? Deviations should be identified, authorized and documented, not smoothed over.

The author illustrates with a composite case: a report states a component was examined on a Tuesday evening, and the signature is valid. The attached data shows the acquisition began at six that morning, and the equipment's calibration record shows verification was due the day before. Each document is tidy alone; together they cannot all be true. No reviewer reading the report alone would have caught it. The discrepancy surfaced only because timeline, equipment and raw data were compared against one another.

No single question proves integrity, the author concedes, but together they make fabrication harder and detection more likely — which is the realistic goal.

Whose compliance question does this raise?

Records are often described as protecting the organization. The author frames it differently: records protect the people downstream who rely on work they cannot see. Digital tools give reviewers a better chance than ever to justify that reliance, but the record that reaches them is only as honest as the system and culture that produced it. For quality departments specifying digital NDT documentation platforms, the open question is whether their procurement criteria include these five design principles — or only checklist completeness. A record, the author concludes, should be a witness, not a performance.