Correct context
Plant, timeframe, component, application, and process define whether a reference is relevant.

Research Methodology
Every meaningful conclusion should be traceable to the evidence, manufacturing context, comparisons performed, and reasoning that supports it.
Reserved for an approved photograph showing component, surface, and mark together.
The Method in Five Steps
Research begins with original evidence—not a desired authenticity conclusion. Appearance alone does not establish authenticity.
Preserve the source and separate what can be seen or measured from interpretation.
Identify the plant, timeframe, component, process, and application that produced the evidence.
Use relevant, independently supported references without turning frequency into proof.
Evaluate manufacturing significance, conflicts, limitations, and alternative explanations.
Record a reviewable finding whose reasoning and uncertainty remain traceable.
The Surface Comes First
For engine-pad research, surface preparation and applied marks are evaluated together in their plant, period, production, and comparison context. Surface lines, groove direction, interruptions, pits, and gouges are recorded as visible features. Explanations involving tool wear, tooling damage, chronology, or lineage require supporting evidence.
Apparent inconsistencies require investigation. Matching characters alone do not resolve conflicting surface evidence, and no universal coarse, smooth, new, or worn-broach appearance rule determines authenticity.
Documented factory exceptions and reasonable alternatives matter. A factory grind-out or unusual surface is not automatically inauthentic; later grinding, resurfacing, restoration, corrosion, damage, machining, and imaging effects must also be considered.
An approved complete source photograph would anchor all later surface observations.
An approved derived enlargement would label direction, interruptions, and tooling characteristics without replacing the source.
Context + Comparison
Useful comparison connects the subject to relevant manufacturing circumstances. Character form, stamping process, factory variation, and reference quality are considered together; one unusual character or visually similar mark is never the whole answer.
Plant, timeframe, component, application, and process define whether a reference is relevant.
Morphology, spacing, orientation, impression, tooling, and application sequence are evaluated together.
Authentic evidence may be imperfect, but anomalies are not automatically normalized as authentic.
Documented provenance and independent, context-qualified examples reduce circular comparison and false standards. Frequency or reputation alone does not establish truth.
Illustrative evidence forthcoming following public-use approval. No source, derivative, reference, or specimen evidence is displayed in this reserved presentation.
An approved subject view would remain separate from its comparison population.
An approved reference would retain documented, reviewable selection rationale.
Poorly supported examples remain outside the approved population.
AI + Human Review
Where approved and enabled, independent Gemini and Claude runs may organize evidence, locate relevant references, compare images and characters, identify possible anomalies, perform measurements and alignment, and help document analytical confidence. Each receives the same governed derivative without seeing the other provider's output. OpenAI is deferred and inactive.
Qualified reviewers interpret manufacturing significance, assess conflicts and historical context, preserve disagreement and uncertainty, and explicitly accept findings where authorized. Methodology, benchmark, publication, and governance status retain their separate approval requirements.
AI-derived observations remain attributable, advisory, traceable, and reviewable. AI does not independently determine authenticity, accept a human finding, confer benchmark status, approve publication, change methodology, or exercise governance authority.
Uncertainty + Traceability
Findings may be supported, inconclusive, affected by conflicting or insufficient evidence, or inconsistent with expected manufacturing characteristics. Unresolved questions are preserved—not converted into invented scores or forced determinations.
Original evidence remains authoritative.
Facts and limitations are recorded separately from interpretation.
Reference selection and exclusions remain reviewable.
Conflicts, uncertainty, notes, and revisions remain in the record.
The finding retains citations and a path back to its evidence.
Original evidence remains distinct from derivatives. Comparison examples, methodology, reviewer notes, revisions, citations, contrary evidence, corrections, and report history remain connected so later reviewers can understand what changed, what remains uncertain, and why a conclusion was reached.
Authenticity assessment, benchmark status, and public presentation remain separate. Restricted material does not become public or externally processable through a description, comparison, or analytical result.
Research • Review • Education