How to Assess Pit Fragments in a Frozen Peach Inspection Plan

Define the finding before choosing a limit

Frozen yellow peach halves with exposed pit cavities. Record the form and intended use in the inspection specification.

Begin with the actual order description. Is the product peach halves, quarters, slices, dice or mixed pieces? Is it described as pitted? What is its intended use, and which specification version applies? These details affect the meaning of a finding. A whole fruit sold with its stone is a different presentation from a pitted ingredient, even though both originate from the same fruit.

Use “suspected pit material” if identity has not yet been confirmed. A brown object in a photograph is not automatically a peach-pit fragment; colour alone cannot establish composition, hardness or sharpness. Keep the distinction between an observation and a confirmed identification in the report. If specialist examination is needed, the material and its traceability should be available for that examination.

Codex CXS 75-1981, Standard for Quick-Frozen Peaches, amended in 2022, defines a visual defect category for pit pieces that are both hard and sharp and measure at least 3 mm in any dimension. Its scope covers the defined product offered for direct consumption, apart from repacking, and excludes product indicated for further processing or other industrial purposes. The official Codex catalogue lists that standard and its 2022 amendment. This is a scoped definition, not a worldwide declaration that smaller pieces are safe.

For your inspection record, preserve the actual observations even where the applicable classification uses a defined size category. Otherwise, a later reviewer sees only “outside definition” without knowing whether the object was tiny and blunt, narrow and pointed, uncertain in identity or damaged during handling. The raw record should allow the responsible reviewer to apply another relevant requirement without having to recreate the finding.

Keep naturally coloured cavity tissue separate from suspected hard material. The presence of a darker area in a peach half can prompt closer examination, but it is not sufficient to count a fragment. Equally, the absence of a visible brown spot in a general product photograph does not demonstrate freedom from fragments. Product images help identify the supplied form; inspection evidence needs its own sample identity and method.

Keep product grading separate from hazard assessment

Three documents can mention pit material while answering different questions. A product grade describes a defined quality classification. A preventive-control document addresses a hazard in its named product and jurisdiction. An enforcement guide describes how an authority considers particular circumstances. Read the scope and conditions before extracting a number from any of them.

The USDA frozen-peach grade standard is voluntary and explicitly states that complying with it does not excuse failure to comply with applicable food law. The current official document is a corrected April 2026 reprint of standards effective in July 1961. Its defect provisions address freedom from pit material. Do not turn allowances for peel or blemished fruit in a grading table into allowances for pit fragments.

CFIA’s guidance on canned peaches says hard and sharp pit fragments of at least 2.0 mm may make that product non-compliant with the cited Canadian provision. It identifies hardness, sharpness, size and shape as relevant to injury risk. The page concerns canned peaches; copying its number into a frozen-peach specification does not establish a universal frozen-product acceptance rule.

FDA CPG 555.425 discusses US enforcement circumstances involving hard or sharp objects, including a 7–25 mm range for ready-to-eat foods or foods needing only minimal preparation that would not remove the hazard. Other circumstances include smaller objects for special-risk consumers and further processing that might affect the object. The guide also distinguishes natural defects covered by other guidance. It does not say that every peach fragment below 7 mm is acceptable, or that heating will neutralise a hard fragment.

Put the applicable document, version, product scope and buyer criterion in the inspection plan before inspection begins. Where the buyer asks for a stricter contractual requirement, state it clearly and agree how compliance will be assessed. Where a requirement is ambiguous, escalate that ambiguity before reporting a simple pass. A commercial concession cannot settle a legal or food-safety question that requires a competent assessment.

For example, a fictional order could have a buyer’s fragment specification and a separate requirement to meet a named product grade. If the grade report is satisfactory but the fragment investigation is unresolved, retain both statuses. Do not overwrite the investigation with the grade result. Each record should show the question it answered and the responsible person who accepted that conclusion.

Record how much product was actually examined

Yellow peach slices in a blue liner. Record both the inspected quantity and how portions were selected from the defined lot.

“One fragment found” is incomplete without an examination basis. Record the number of packages selected, where they came from within the defined lot, the quantity submitted and the quantity actually examined. Those quantities can differ. If part of a submitted bag was retained untouched, do not present its entire labelled weight as the examined quantity.

Codex CXS 75 uses different standard units for segregating and counting visual defects: 30 units for halved or quartered styles, and 300 g of drained fruit for sliced, diced or pieces styles. Those are units within that standard’s quality-assessment system. They are not, by themselves, a universal shipment sampling plan or proof that a whole lot is free of hazardous material.

A useful report can retain both the actual examined mass and the relevant count information. For a fictional sample of peach halves, an inspector might record the number of halves examined, their mass and the number containing suspected material. For slices, the record might instead identify the inspected mass and individual findings. The agreed method determines the formal calculation, while the underlying observations remain available.

Count fragments and affected fruit units separately where both matter. Two pieces attached to one peach half are not the same observation as one piece in each of two halves. A loose piece also has a different observed location from a piece still embedded in tissue. Record what was seen before converting observations into the categories required by a specification.

Keep the sample selection method visible. A bag deliberately taken after an operator noticed a problem is valuable investigative evidence, but its selection differs from a planned sample spread across the lot. Neither should disappear from the file. Label the purpose of each collection so an investigation sample is not presented as an unbiased estimate of the shipment’s overall condition.

When combining findings, preserve the package-level results as well as the total. If all findings came from one selected carton, an aggregate rate can conceal that concentration. The pattern may help define follow-up questions about production time, packing or material movement. It cannot independently prove the cause, but removing the pattern makes a useful investigation harder.

A no-finding report also needs a basis. State that no specified material was found in the examined quantity using the recorded method. Avoid turning that statement into “the lot contains no fragments.” The latter is a much broader claim than a selected sample ordinarily supports. Record any visibility, preparation or method limitation that affected the examination.

Measure and preserve the individual finding

Conceptual dimension-recording illustration. The illustrated pit material and unnumbered scale are not a real finding or usable measurement; no fragment size is declared acceptable.

Give each finding an identifier and connect it to its original sample. Photograph the material in context when practicable, then obtain suitable close views for examination. A close-up with no scale or identity may show an interesting shape while leaving its dimensions and origin unresolved. Keep the original image file and its connection to the observation record.

Use a suitable measuring approach and report the dimensions required by the applicable assessment. Explain what was measured: maximum length, width, thickness or another defined dimension. A flat photograph shows a projection; it may not capture the largest dimension of a curved or irregular object. If additional views or direct measurements are needed, record them rather than estimating a definitive size from a single perspective.

A scale should be readable and positioned so that the image can be interpreted correctly. Do not use a familiar object, such as a fingertip or coin, as the only size reference for a formal finding. Record the measuring device and the precision appropriate to it. If a result is close to a decision boundary, the responsible reviewer should consider measurement uncertainty and the method before rounding the observation into a convenient category.

The following is an illustrative record structure, not a completed inspection or a universal laboratory procedure. Adapt it to the agreed method and the evidence needed for the decision.

Preserve the object in a suitable identified container and document relevant handling. Avoid unnecessary manipulation that could break it or alter the surfaces being assessed. If separation from the fruit changes its condition, describe that change. A later observer should be able to distinguish the original finding from something created during recovery or transport.

Do not ask someone to test sharpness against their skin or mouth. Let the competent examiner select an appropriate assessment method. Similarly, a casual attempt to crush a suspected fragment can destroy useful evidence without answering the formal question. The inspection plan should define who evaluates uncertain material and how that assessment is recorded.

Review prevention and detector evidence separately

Illustrative product arrangements, not a controlled detector trial or X-ray images. Layering is one condition to document when evaluating the actual product and system; the picture establishes no detection capability.

A finding should trigger questions about how pit material might have remained with the product and how the relevant controls performed. Ask for the process information needed to investigate the identified lot: raw-material identity, pitting and trimming arrangements, relevant checks, production timing, material movement and recorded deviations. Request the records that bear on the finding rather than a generic statement that the factory has quality control.

The investigation may need to consider several hypotheses. A fragment could be associated with a raw-material condition, a pitting issue or a later handling event; the finding alone does not choose between them. Keep proposed explanations separate from confirmed causes. If records do not support a particular explanation, report the gap and identify the next evidence needed.

Equipment evidence answers another question. Southwest Research Institute describes X-ray monitoring in terms of differential absorption associated with material density, with pits in fruit pies among its examples. That example demonstrates an application of the principle, not a guaranteed detection size for fragments in your frozen peaches.

METTLER TOLEDO’s technical paper, How Do X-ray Systems Find Contaminants?, explains that product thickness, density and homogeneity, as well as object geometry, location and packaging, affect detection. This supports a practical request: ask what performance has been demonstrated with the actual product presentation. A machine model name or an unrelated test piece does not define every fragment it can find.

For a proposed detector review, describe the peach form, its condition at inspection, product depth or packing presentation, relevant operating conditions and the material the validation is intended to detect. Ask how detection and rejection were checked and how the records connect to production. The equipment provider and responsible technical team should determine the appropriate validation approach. Do not insert an unverified minimum detectable size into a purchase specification as though it were established performance.

Keep detection and removal distinct. A system may identify an event, but the review also needs to establish what happened to the affected product. If a reject record exists, connect it to the relevant handling and reconciliation information. If a later sorting step is proposed, assess whether it is suitable for the actual problem and how its effectiveness will be demonstrated. Calling a process “reinspection” does not establish that it controls the identified risk.

Buyer sampling remains a third layer. It can check selected material against an agreed plan and provide useful independent observations. It does not replace the supplier’s process controls or validate a detector merely because no fragment was found in the sample. Review these records together while preserving the different conclusion each can support.

Make the lot decision traceable

Conceptual record linkage. The pictured container, sample and carton have no real incident identifiers; retained evidence needs the actual identification and handling records.

Define who receives a finding, who assesses its significance and who can authorise the next action. If the assessment is unresolved, make the affected product’s status visible so that it is not treated as routinely cleared while the review continues. The scope of any restriction, investigation or further action should follow the evidence and applicable requirements, with responsibility clearly assigned.

A fictional example shows why identity matters. An importer receives a photograph of a suspected fragment, but the image has no sample code. The supplier has several production lots in the shipment. Before discussing a numerical tolerance, connect the physical object, original bag, examination record and shipment documents. A photograph that cannot be connected to a lot leaves both the investigation and any proposed disposition on uncertain ground.

If the finding is later confirmed, preserve that conclusion alongside the original observation. Record the affected scope, supporting evidence, decisions and outstanding actions. If the identity remains uncertain, state that uncertainty and the basis for the chosen next step. Neither status should be replaced by a convenient general assurance about the supplier’s history.

Further sampling can answer additional questions when the plan is justified, but a later no-finding result does not erase the earlier material. Explain what the follow-up was designed to establish, how it was selected and how the results were interpreted together. Avoid searching repeatedly until a favourable sample appears and then presenting only that sample as the conclusion.

Close the review with more than “accepted” or “rejected.” Identify the criteria used, the responsible reviewer, the supporting records and any restrictions, corrective actions or verification still required. Keep the decision connected to the order specification and lot. That gives the buyer and supplier a shared account of what was found, what was established and what remains to be done.

References

• Codex Alimentarius. CXS 75-1981, Standard for Quick-Frozen Peaches, amended 2022, sections 1 and 3.3. Official standards catalogue entry.
• USDA: Standards for Grades of Frozen Peaches — corrected reprint, April 2026.
• CFIA: Preventive controls for canned peaches — product-specific pit-fragment guidance.
• FDA: CPG 555.425, hard or sharp foreign objects — US enforcement guidance.
• METTLER TOLEDO. How Do X-ray Systems Find Contaminants? Technical paper, public full-text edition consulted September 2026.
• Southwest Research Institute: X-Ray Process Monitoring — scoped technology example.

More guides and related supply information