Inspection Lot
Inspection Lot is the defined population of units presented together for one inspection decision under stated acceptance criteria.

On this page · 11 sections
- 1What an inspection lot defines—and what it does not
- 2Compare the inspection population with its neighbouring records
- 3Set boundaries around the differences that affect the decision
- 4Record membership, quantities and the evidence-to-population connection
- 5Control the population from formation to human disposition
- 6Worked case: separate a mixed delivery with incomplete identity
- 7Worked case: divide one production lot by treatment history
- 8Worked case: form a successor inspection population after rework
- 9Separate current legal context, audit guidance and engineering analogy
- 10Recover identity and boundary errors across manufacturing records
- 11Review a population by testing its weakest assumptions
How does Inspection Lot apply to your shop floor?
Pick your industry and scale — Ask V5 rewrites the definition in your context, gives a worked example, and shows what V5 does on day one.
01What an inspection lot defines—and what it does not
Inspection Lot is a defined population of units presented together for one inspection decision against specified acceptance criteria. The term is used here in its manufacturing-practice sense, not as an FDA-prescribed lot definition. Its purpose is to make the reach of an inspection decision explicit.
The population consists of identifiable units included in that decision. Presenting units together means treating them as one decision population; it does not necessarily mean placing them on one pallet, receiving them together, or inspecting every unit.
The decision scope names the characteristic or set of characteristics being assessed, the applicable specification revision, and the product state to which the assessment applies. Criteria give that decision its basis: what constitutes an acceptable result for those characteristics in that state. An undated reference to the current specification can leave that basis ambiguous when specifications change.
For example, assume a hypothetical medical-device manufacturer assesses machined housings after deburring but before cleaning. The inspection scope covers dimensional conformity against an identified drawing revision. Cleanliness after a later cleaning operation is a separate characteristic and product state.
A useful inspection record therefore answers a practical question: exactly which units does this decision cover, and which units remain outside it? That question comes before choosing how to inspect. This article addresses population boundaries and decision coverage; sampling-plan selection, AQL tables, universal sample sizes, and incoming-inspection procedures are outside its scope. Those subjects concern how evidence is obtained, rather than which population the resulting decision covers.
02Compare the inspection population with its neighbouring records
In manufacturing practice, these records answer different questions: where material originated, how it moved, what was examined and what people authorised. Their identifiers describe different relationships, not interchangeable names for the same group.
| Term | What it groups or records | Relationship to an inspection lot | What it does not establish. |
|---|---|---|---|
| Inspection lot | Population presented for one inspection decision. | Defines the decision’s membership. | Every approval for subsequent use. |
| Supplier lot | Supplier-assigned source population. | May map to one or several inspection lots. | A suitable inspection boundary by itself. |
| Production lot or batch | Manufacturing population with a shared production identity. | Linked through production history or a batch record. | Identical treatment of every member. |
| Delivery or consignment | Material shipped together. | May contain several inspection populations. | Common source or product state. |
| Handling unit | Material physically handled together, such as a pallet. | Contains all or part of a population. | A defensible decision boundary. |
| Inspection sample | Members actually examined. | Drawn from the defined population. | Examination of every population member. |
| Inspection record | Observations, results and inspection context. | Connects evidence to identified membership. | Permission to distribute product. |
| Quality disposition | An authorised human decision about material status. | Applies an outcome to identified membership. | Completion of approvals outside its scope. |
Supplier-lot and production-lot relationships can each be one-to-one, one-to-many or many-to-one when the boundary rationale supports that arrangement. Combining source lots for inspection does not erase their separate identities; dividing a production lot does not create new manufacturing origins.
Keep the examined sample distinct from the population receiving the resulting decision. Inspection acceptance addresses the defined inspection scope, not every approval needed for subsequent use or distribution. A disposition code commonly labels the authorised outcome; its meaning depends on the decision and approval scope recorded with it.
03Set boundaries around the differences that affect the decision
Homogeneity is a justified assumption about the attributes relevant to an inspection decision—not a claim that every unit is identical or conforming. In manufacturing practice, a defensible boundary groups units whose relevant histories support evaluating them together. Start with the characteristic being assessed and the failure mechanisms that could affect it, rather than the convenience of a delivery, container or stock record.
Potential boundary attributes include part and revision, material identity, supplier process, production run, equipment or tooling history, treatment state, packaging condition and exposure history. Their significance depends on the decision. Tool wear may matter to dimensions; treatment history may matter to surface properties; damaged packaging or moisture exposure may matter to cleanliness or material condition. Relevant batch execution history can help distinguish a meaningful process change from an administrative label change.
A shared stock code or specification identifies intended product characteristics, not necessarily comparable manufacturing or storage histories. Conversely, different source-lot numbers do not automatically establish a meaningful difference for every characteristic. A compatibility rationale explains why the known differences do not undermine the particular decision; it does not simply assert that all material meets the same specification.
- One population: where relevant histories are compatible, document the basis for grouping, including which differences were considered and why they are not consequential.
- Separate populations: where known differences could change the failure mechanism or interpretation of results, establish boundaries that preserve those differences for human disposition.
- Unresolved material: where identity or relevant history is uncertain, keep that material outside a population receiving a decision until the uncertainty is resolved. Record the uncertainty rather than assigning an assumed history.
An identifiable mixed delivery contains distinguishable groups: labels, segregated packaging or reliable unit identifiers allow assignment to their respective histories. Untraceable commingling has lost that connection. Administrative splitting is useful only when actual units can be assigned to the proposed populations. Creating separate record numbers cannot reconstruct lost membership. Bin-level lot genealogy is useful only to the extent that recorded contents still match physical contents.
The aim is neither the largest possible population nor the greatest number of subdivisions. Over-broad grouping can conceal distinct failure mechanisms. Unnecessary fragmentation adds records without improving the decision. A useful review question is whether each proposed boundary changes how the inspection evidence should be interpreted for the units concerned.
04Record membership, quantities and the evidence-to-population connection
A practical inspection-lot record separates population identity, quantity accounting, observations and disposition. The following information model is a proposed manufacturing practice, not a prescribed regulatory schema. Its purpose is to make clear which material a decision covers, even after that material moves or changes state.
- Identity and scope: assign an inspection-lot identifier distinct from supplier-lot and production-lot identifiers; retain those identifiers as linked references. Include population version, decision scope, specification revision, declared product state, quantity and unit of measure.
- Membership: identify units, serials, containers or explicitly bounded quantities within identified source material. For container-based membership, preserve a dated contents snapshot or equivalent membership record. A container identifier alone becomes ambiguous when material can be added, removed or replaced.
- Evidence: associate each observation with the population version and, where useful, the member actually examined. Proposed fields include characteristic, method reference and revision, result with units where applicable, performer identity and observation time.
Keep the declared inspection population separate from current physical inventory and quantity available for subsequent use. Record testing consumption, scrap and transfers as events against the relevant members or bounded quantities rather than overwriting the original population quantity. A transfer changes location, not necessarily total inventory; consumed or scrapped material is no longer available. Linking quantity events to discrete consumption records helps prevent the same withdrawal being counted twice.
Reconcile quantities on a common unit basis. Pieces, kilograms and cartons cannot simply be added together. Use a controlled conversion basis tied to the relevant material and packaging state; do not assume every carton contains a fixed count. Where contents vary, retain the actual count or measurement supporting conversion.
Keep the inspection record separate from the human disposition record, including its disposition code, approver, date and covered population version. Model parent–child population relationships separately, recording transferred membership and the reason for the relationship. Historical results remain attached to the population examined; a successor relationship alone does not establish that those results apply to changed material.
Inspection population
Defines the versioned decision boundary.
- inspection_lot_id
- population_version
- decision_scope
- specification_revision
- declared_quantity
- uom
Membership
Connects included material with provenance.
- membership_id
- inspection_lot_id
- population_version
- unit_or_container_id
- supplier_lot_id
- production_lot_id
- member_quantity
Source lot
Preserves independently assigned source identity.
- source_lot_id
- issuer_id
- source_lot_type
- source_record_ref
Inspection evidence
Associates observations with population membership.
- evidence_id
- inspection_lot_id
- population_version
- membership_id
- characteristic_id
- result
Human disposition
Records the reviewed population decision.
- disposition_id
- inspection_lot_id
- population_version
- decision
- reviewer_id
Population relationship
Preserves split and rework genealogy.
- parent_inspection_lot_id
- parent_version
- child_inspection_lot_id
- child_version
- relationship_type
- transferred_quantity
- uom
05Control the population from formation to human disposition
A practical inspection-lot lifecycle keeps the decision population connected to its evidence. The stages below are conceptual controls, not mandatory software statuses or a receiving procedure.
Population formation → Compatibility review → Identity resolution → Membership commitment → Evidence association → Human disposition Boundary change → Reassess population and evidence → Renew membership commitment
Formation proposes the material covered by one decision. Compatibility review tests whether that grouping is defensible for the characteristic being assessed; identity resolution settles which units belong. Membership commitment establishes the cutoff: the point at which membership, quantity and product state are sufficiently defined for evidence to be interpreted consistently. It need not coincide with delivery completion.
Evidence association connects the committed population to its inspection record. An authorised person then evaluates that relationship and records the disposition; attaching results alone does not decide the population’s acceptability.
Treat added, removed or reassigned material as an explicit boundary-change event. Before extending existing evidence to the revised population, assess whether it still addresses the membership and state concerned. If not, identify the evidence gap rather than carrying the earlier conclusion forward.
Correcting a membership record repairs its description of the same actual population. A substantive change in product history or state, such as rework, instead calls for a successor population. Detailed recovery mechanics belong in the integration section. Receiving checks, instrument operation and sampling-plan execution remain outside this population-control diagram.
- Define scopeIdentify characteristics, criteria and product state.
- Identify membershipList units or containers and source relationships.
- Assess compatibilityEvaluate decision-relevant process and exposure differences.
- Resolve boundariesSeparate incompatible groups; isolate unresolved identity.
- Commit populationRecord membership version, quantity and cutoff.
- Associate evidenceConnect observations to the population examined.
- Record dispositionA human decides; preserve successor relationships.
06Worked case: separate a mixed delivery with incomplete identity
Assume a hypothetical medical-device manufacturer receives one delivery containing 1,200 moulded actuator housings of the same part revision. The unit of account is one housing. Labels and supporting records identify 700 housings as supplier lot A and 400 as supplier lot B; source-lot labels for the remaining 100 cannot be resolved.
Assume supplier records identify different moulding histories for A and B that are relevant to the fit characteristic being evaluated. For this decision, the manufacturer forms separate inspection populations for A and B. Their shared part revision in material master data describes the intended component, but does not establish a shared moulding history. Separation here follows the assumed fit-relevant difference, not a universal rule that every supplier lot needs its own inspection population.
The unidentified 100 housings remain outside both identified populations. Allocating them proportionally would turn a quantity calculation into an unsupported identity claim. Copying the nearest carton label would likewise substitute physical proximity for evidence of origin. Neither approach establishes which moulding history applies.
Quantity reconciliation is exact: 700 + 400 + 100 = 1,200 housings. Assume the inspection evidence supports acceptance of A, while B has an unresolved dimensional nonconformity. An authorised human accepts the 700-housing A population for the stated next operation, actuator assembly, and retains the other 500 housings on hold: 400 for dimensional investigation and 100 for identity resolution.
| Record linked to the delivery | Quantity | Separate disposition reference |
|---|---|---|
| Inspection population A | 700 housings | Human acceptance for actuator assembly |
| Inspection population B | 400 housings | Dimensional nonconformity hold |
| Unresolved identity group | 100 housings | Identity-resolution hold |
The resulting records preserve one delivery relationship, two identified inspection populations and one unresolved identity group. Each disposition reference points to its own membership and decision basis. Acceptance of A therefore does not change the status of B or the unidentified group merely because all three arrived together.
07Worked case: divide one production lot by treatment history
Assume a hypothetical medical-device manufacturer produces one production lot containing 900 stainless-steel clamp bodies, counted as individual bodies. Allocation records trace 600 bodies to electropolishing run E and 300 bodies to electropolishing run F. Assume no losses, scrap or unallocated bodies at this stage: 600 + 300 = 900 bodies. Both allocations retain the original production-lot identity.
The inspection decision concerns surface condition after electropolishing. For this example, the manufacturing procedure treats each treatment run as a separate process history relevant to that decision. The manufacturer therefore forms an inspection population of 600 bodies from E and another of 300 bodies from F. This is an assumed manufacturing practice, not a general legal rule requiring every treatment run to become an inspection lot.
The run allocation provides the distinguishing context. Linking each population to its electropolishing history through production context tagging makes clear which bodies the surface-condition evidence addresses. Shared upstream material identity does not erase the separate treatment histories.
Assume E has acceptable surface-condition evidence, while F has an unresolved surface-condition issue. A human reviewer accepts the 600 bodies from E for the next operation and places the 300 bodies from F on hold. F remains unresolved rather than becoming acceptable through its association with E.
The resulting genealogy preserves one parent production lot with two inspection populations distinguished by treatment history. Its batch record still accounts for all 900 bodies: 600 accepted for the next operation plus 300 held. Each decision points to its own population and supporting evidence. Neither the acceptance of E nor the hold on F silently extends to the other treatment run.
08Worked case: form a successor inspection population after rework
Assume a hypothetical medical-device manufacturer rejects an original inspection population of 500 machined instrument handles because of a dimensional feature. All quantities below count individual handles. The example illustrates manufacturing practice, not a legal mandate. The original rejection applies to that defined population; it does not establish that every handle has the same dimensional error.
The manufacturer selects 320 identified handles for one documented rework route. The remaining 180 handles stay unreworked on hold. During rework, 20 handles are scrapped and 300 complete the route. Completion means the documented operation has been performed; it is not itself an acceptance decision.
Form the successor inspection population from those 300 completed handles, using their identified membership and documented post-rework state. Establish membership before using the new inspection result to decide disposition. Do not construct the successor by selecting only handles with passing measurements: that would confuse the population being evaluated with the outcome of its evaluation.
The quantity reconciliation is 500 = 300 completed rework + 20 scrapped + 180 unreworked handles. These categories have no overlapping membership or unexplained quantity loss. The 320 transferred handles reconcile separately as 300 completed plus 20 scrapped.
Link the original population and rejection to the transferred membership, rework instruction revision, completed operation, scrap identities, successor population and new inspection record. Retain the original evidence and rejection rather than replacing them with the successor result. This record chain explains both where each handle went and which product state each inspection assessed.
For this example, the stated post-rework criteria are conformity of the dimensional feature to the applicable drawing tolerance and documented route completion with no unresolved rework deviations. Assume the successor population satisfies those criteria. At the human disposition step, a reviewer accepts those 300 handles for the next operation; the 180 unreworked handles remain separately on hold. The successor represents an intentionally changed product state—not an attempt to erase an unsuccessful inspection or obtain a different result by relabelling unchanged material.
09Separate current legal context, audit guidance and engineering analogy
For US medical-device applicability, consult the current text of 21 CFR Part 820 and the FDA QMSR final rule. These are legal-scope references, not an attribution to FDA of this article’s inspection-lot definition, sample size, proposed fields or retention interval. Keep the distinction between QMSR and ISO 13485 separate from the practical task of defining an inspection population.
Use MDSAP as medical-device audit guidance, not current US law. Check its country-specific references against the applicable jurisdiction; do not carry legacy Part 820 citations forward as current US obligations. For medical-device organisations assessed through MDSAP, Assessing conformity: Recognized acceptance activities expects records to identify who conducted acceptance activities. That supports identifying the person responsible for the activity; it does not establish a universal inspection-population schema.
For NASA engineering practice, section 5.2.1.2.7, Capture Product Integration Work Products includes updated product configuration and support documentation among integration work products. Use this only as an engineering analogy for preserving configuration context, such as through production context tagging, not as a requirement for device manufacturers. The proposed inspection-population model remains manufacturing practice: a way to connect membership, product state and evidence to a human decision. Do not extend medical-device audit guidance to pharmaceuticals, food, supplements or chemicals.
10Recover identity and boundary errors across manufacturing records
A practical integration design assigns record ownership explicitly rather than letting whichever system updates last define the population. These are proposed manufacturing controls, not legal mandates:
- Purchasing owns supplier identity, purchase references and their corrections.
- Warehouse owns physical container identity, location and movement records, including partial transfers.
- Manufacturing owns production genealogy and processing-history references.
- Quality owns inspection-evidence associations, population-version review and the authorized human disposition record.
Across MES–ERP integration, use the inspection-lot identifier and population version as decision references while retaining original identifiers from every source system. An inventory balance describes quantity, not membership: equal balances can conceal different containers or units.
- Duplicate receipt messages: compare the source transaction identity before adding membership. Preserve the duplicate-message history without counting the material again.
- Delayed acknowledgement: reconcile the existing transaction; interface silence does not justify issuing a fresh population identity.
- Late-arriving material: keep it outside the existing decision scope until quality reviews whether inclusion changes membership and decision coverage.
- Wrong source-lot assignment: preserve the original assignment, record its correction and identify affected populations, evidence links and dispositions.
- Partial inventory transfer: reconcile moved and remaining membership against physical containers. A location change alone need not create a new population.
- Results attached to an obsolete version: retain their original association and have quality assess applicability before linking them to the current version.
For exception recovery, preserve the original event, record a correction or successor relationship, and reconcile physical membership before closing the discrepancy. Through MES–QMS integration, identify which evidence associations and human decisions are affected. A real membership change calls for explicit review of decision coverage, not silent inheritance of an earlier disposition. Record the reviewer’s conclusion and any unresolved scope separately from interface delivery status.
11Review a population by testing its weakest assumptions
Use these suggested practice tests before signing an inspection decision. Ask a reviewer other than the preparer to reconstruct the population from its records; familiarity with the material is not a substitute for reproducible membership.
| Review question | Record that answers it | Action if uncertain |
|---|---|---|
| Can another reviewer identify the same included units and excluded remainder without the preparer's memory? | Membership register, unit or container identifiers, and exclusion list. | Resolve ambiguous identifiers and document the boundary before deciding. |
| Which known difference would justify dividing this population, and does its actual history contain that difference? | Grouping rationale compared with production, treatment and storage history. | Investigate the counterexample; divide the population or document why the difference does not affect this decision. |
| Can one split, one consumption or scrap event, and one remaining balance be traced? | Quantity ledger and linked event records; use the worked cases above as reconciliation references. | Investigate missing movements or double counting rather than adjusting a balance merely to agree. |
| Do the evidence's product state and population version match the decision being signed? | Inspection results, state/version identifiers and proposed release record. | Pause the decision and establish which evidence applies. |
| Can the successor be traced to the earlier decision without unresolved material inheriting acceptance? | Parent–successor links, earlier disposition and unresolved-material status in lot genealogy. | Correct the status linkage and obtain human review of the affected population. |
A useful review outcome is a person who can explain membership, boundary rationale, evidence coverage and disposition from the records—not by equating a supplier number, pallet or stock balance with the inspection population. Record unresolved questions and who will resolve them before approval.
Frequently asked questions
Q.Can separate deliveries from the same supplier lot share one inspection lot?+
Separate deliveries can share that population when their identity, relevant condition and inspection scope remain comparable. A common supplier-lot identifier alone does not establish comparability: transport exposure, packaging damage or storage history may justify separation. Define which deliveries are included before drawing conclusions, and retain their individual receipt identities. If another delivery arrives after human disposition, assess it separately rather than extending the earlier decision without review; preserve the relationship through lot genealogy.
Q.When can material from different suppliers belong to the same inspection population?+
Pooling is potentially defensible only for a clearly defined characteristic when technical evidence supports treating the sources as comparable for that decision. Matching part numbers or specifications alone is weak justification because supplier processes can produce different defect patterns. For hypothetical medical-device housings, a shared cosmetic assessment would not establish equivalence of resin identity or strength. Preserve supplier attribution within the population and document the technical rationale through the quality assurance process; otherwise, keep separate populations.
Q.What should happen when the physical count differs from the recorded inspection-lot quantity?+
Pause the affected disposition decision and reconcile the discrepancy before treating either quantity as authoritative. Check whether the difference reflects counting error, unit-of-measure confusion, omitted containers, test consumption or an unexplained movement. Correct the quantity with a traceable reason rather than overwriting the original entry. If reconciliation reveals changed membership, reassess whether the inspection results still represent the population. A human reviewer can then decide whether the existing evidence remains usable or further evaluation is appropriate, documenting unresolved differences through a deviation.
Q.Can different inspected characteristics justify different populations for the same material?+
Yes. Population boundaries can be characteristic-specific when the factors affecting those characteristics differ. In a hypothetical medical-device assembly, molded dimensions might be evaluated by molding run, while packaging condition is evaluated by shipping exposure. Record each population’s membership and decision scope explicitly in the inspection record. Avoid a single undifferentiated accepted status that hides an outstanding assessment. Human disposition of the material can consider the relevant decisions together without assuming that evidence for one characteristic establishes acceptability for another.
Q.Does destructive testing require a new inspection-lot identity for the surviving units?+
In common manufacturing practice, destructive testing alone does not normally justify a new population identity when it removes identified test units without changing the remaining material or its relevant condition. Retain the original population quantity, identify consumed units and reconcile the remaining quantity available for disposition. Link results to the population from which the test units came. If test preparation exposes or alters other units, assess that affected subset separately. Discrete consumption records can distinguish test consumption from unexplained losses.
Q.Does moving accepted material into different containers create a new inspection lot?+
Not ordinarily, if the move preserves identity, membership and the conditions relevant to the acceptance decision. Container identity and inspection-population identity serve different purposes. Record the transfer and maintain the relationship between the accepted material and its new containers through container status tracking. If repacking introduces a new concern, such as contamination exposure or damaged protective packaging, define an assessment for that concern. A human reviewer can determine whether the earlier disposition remains applicable rather than assuming either automatic acceptance or automatic invalidation.
Q.When does sorting a rejected lot create a new population rather than merely repeat the inspection?+
Sorting creates a new population when it deliberately changes membership by removing identified units under a documented selection rule. Record the retained and removed quantities, the sorting criterion and the relationship to the rejected parent. Rework likewise creates a changed inspection population when it alters the relevant condition of the units. Use parent–child lot relationships to preserve that history. Merely selecting another sample from unchanged material does not create a new population. Subsequent human disposition should address the resulting population, not erase the original rejection.
Q.How should a later-discovered nonconformity affect an inspection lot that already has a human disposition?+
Preserve the original disposition as a historical decision and open a linked assessment of the new information. Establish which units could be affected, including material already consumed, transferred or distributed, and consider whether related populations share the cause. Where appropriate, restrict further use of remaining stock while the assessment proceeds. The responsible human decision-maker can revise the current disposition with a recorded rationale. Use recall readiness to support downstream traceability without assuming that every later finding necessarily warrants a recall.
Primary sources
Want to see how Inspection Lot could fit into your own records and workflows? Explore the related V5 pages or talk to our team about what applies to your operation.
