Skip to content

Scan and index quality

Digital records index quality across the shop-visit source file

A digitized shop-visit package is only usable if each scan can be found, read, and tied back to the record it images. This review measures the digital records index against those three tests across the closeout package: search fields populated and accurate, image quality sufficient for the signatures and stamps that matter, and every file attributed to the correct aircraft, work order, and document type. It is run at acceptance of the digital package. Files that fail come back listed by defect class with the rescan or re-index action required.

When this review is needed

  • The shop delivers records digitally and the operator's system will be the working copy.
  • Scanning was outsourced or batch-processed and nobody has sampled the output.
  • Records from multiple aircraft moved through the same scanning queue during the visit.
  • A future transaction will rely on searching this package rather than paging through it.

The problem

Scanning projects report volume, and volume hides defects. A thousand files can be delivered on time with hundreds unusable: metadata copied down a column so every file claims the same work order, stamps rendered illegible at the chosen resolution, double-feeds that silently dropped pages, files attributed to the wrong tail after a batch mix. None of this is visible in a file count, and the team accepting the package rarely opens more than a handful of documents.

What gets reviewed

  • Index metadata sampled for accuracy against the imaged documents
  • Search-critical fields checked for population: aircraft, document type, date, work order, part and serial where relevant
  • Image quality assessed at the points that carry evidential weight: signatures, stamps, certificate numbers, handwritten entries
  • Page counts of multi-page records verified against their paper or stated originals
  • File-to-aircraft attribution tested where multiple assets shared the scanning workflow
  • Duplicates, blanks, and misfiled images identified across the package

Scope this review

Tell us the asset, the event, and the evidence in scope, and we will outline a focused first engagement.

Send a representative, redacted record set and we will scope the review.

What gets validated

  • A targeted search by work-order number returns the correct documents and only those
  • Sampled metadata fields match what the imaged document actually says
  • Stamps, signatures, and certificate numbers are legible at the delivered resolution
  • Multi-page records are complete, in order, and free of double-feed page loss
  • No sampled file images a document from a different aircraft or visit

Evidence normally required

  • The delivered digital package and its index or metadata export
  • The scanning specification or records standard the delivery was made against
  • Access to paper originals or the shop's working copies for sampling
  • The work-order and document-type structure of the visit for search testing

Common discrepancies

  • Metadata propagated in bulk so entire batches carry one work-order number
  • Resolution adequate for text but too low for inspection stamps and wet signatures
  • Page loss inside multi-page test reports from feeder double-picks
  • Files from a concurrent aircraft's visit indexed into this package

What is at stake

Defective digital records fail at the moment of retrieval, which is always a worse moment than acceptance: an audit question, a diligence request, an AD accomplishment challenge. A record that exists but cannot be found is operationally identical to a missing record, and re-scanning requires the paper originals, which many programs destroy or return after digitization. Wrong-tail attribution is the most dangerous defect because it corrupts two aircraft files at once.

How the work runs

01

Fix the sampling frame

Stratify the package by document type and scanning batch, weighting evidential documents.

02

Test retrieval

Run known-item searches by work order, document type, and serial number and score the returns.

03

Inspect the images

Check legibility at signatures and stamps and verify multi-page completeness.

04

Issue the worklist

Report defects by class with the rescan and re-index actions, and document the sampling basis.

What the buyer receives

  • A defect report by class: index errors, image quality, completeness, attribution
  • A rescan and re-index worklist scoped to the failing files
  • A sampling statement documenting method and coverage for the acceptance file

Who uses the output

  • Records managers deciding whether to accept the digital package as the working copy
  • The scanning provider or shop records office executing the rescan worklist
  • Technical teams who will search this package under audit or transaction pressure

How the work fits into the transaction or program

Digitization is the point of no return for many programs, since paper is often released or destroyed after acceptance. This review is the gate before that step, and its sampling statement becomes part of the package's own provenance, evidence that the digital copy was tested before it replaced the paper.

Start with a single asset

Confirm release certificates and component traceability are complete.

Jurisdiction-specific considerations

FAA AC 120-78 sets expectations for electronic recordkeeping and signatures, and AC 43-9C shapes what a maintenance record must show regardless of medium. EASA-regime holders under EU 1321/2014 must keep continuing-airworthiness records retrievable for their retention periods, which makes searchability a compliance property rather than a convenience.

Regulatory limits

The review evaluates the digital package's fidelity and retrievability. It does not validate the content of the underlying maintenance records, approve an electronic recordkeeping system, or determine airworthiness; a perfect scan of a defective record is still a defective record.

What this review does not cover

  • Review of the substantive maintenance content of the imaged records
  • Selection or configuration of the operator's records software
  • Destruction decisions for paper originals

Specific to this review

  • Sampling must be stratified by document type, since certificates, cards, and reports fail in different ways; a flat random sample over-tests routine cards and under-tests releases.
  • Bulk metadata errors are cheap to make and cheap to detect: one search by a known work-order number exposes a propagated column instantly.
  • Wet-signature legibility is the binding constraint on resolution, and it cannot be judged from text legibility.
  • Attribution defects concentrate at batch boundaries, so sampling the first and last files of each scanning batch finds a disproportionate share.

Sources

Frequently asked questions

What sample size is enough?

It depends on batch structure more than package size. Stratified samples that cover every batch and every document type at modest depth outperform large flat samples, because the dominant defects are batch-level, and one hit in a batch justifies opening the whole batch.

Relevant glossary terms

Related pages

Where this fits

Talk to an engineer who has done this work

We will walk through your current state, the records or evidence involved, and a scoped first engagement.

Talk through the aircraft, records, evidence, deadline, and next useful step.