Plan Incoming Inspection for an Undercarriage Parts Lot

An incoming inspection sample helps you make a controlled decision about a defined shipment lot. It does not prove that every uninspected part is free of nonconformities. A useful plan therefore starts before the boxes are opened: define the population, list the characteristics to check, identify the approved sampling basis, and record what acceptance, rejection or an inconclusive result will mean.

This distinction matters for undercarriage parts because one delivery can contain several part numbers, drawing revisions or production batches. If those populations are mixed in one record, a clean result from one group may be used incorrectly to release another.

Define a lot that can be inspected meaningfully

Give the inspection lot a unique ID and connect it to the purchase-order line, part number, approved drawing or specification revision, supplier lot, production batch and received quantity. The boundary should let you trace every selected unit back to the same defined population. The NIST description of lot acceptance sampling treats the sample as information used to decide the disposition of a lot. That only works when “the lot” is clear.

A mixed shipment can still be received under one purchase order, but it may require several inspection lots. Physical proximity on one pallet is not evidence that all parts share the same manufacturing or specification conditions.

Mixed part numbers

Separate different part numbers unless an approved plan expressly defines a valid combined population. A track roller shell, a bolt and a seal have different requirements and failure modes. Selecting several easy-to-reach bolts cannot support a release decision for the roller shells packed below them.

Also separate variants that look similar but have different drawings, material requirements or fitment. The receiving record can retain one shipment reference while giving each part population its own inspection-lot ID and decision.

Mixed production batches

Check heat, cast, forging, machining, assembly or supplier batch identifiers when they are part of the agreed traceability. If two batches are present, decide whether the approved sampling basis permits them to be treated together. If it does not—or if the basis is silent—hold them as separate populations until the responsible quality authority decides.

This prevents a result from production batch A from silently releasing batch B. It also preserves a useful investigation path if a later result points to only one batch.

Identify what must be checked

Translate the order requirements into inspection characteristics before selecting units. For each characteristic, record the requirement and revision, method, equipment or evidence needed, result format, nonconformity category and decision rule. “General inspection passed” is too broad to show what was actually verified.

Keep categories distinct. Identity, visible condition, dimensions, material or process documents, and functional or agreed tests answer different questions. Passing one category does not imply that the others passed.

Identity and visible condition

Identity checks may compare the item mark, part number, revision, quantity unit, supplier batch and packaging label with the released order record. Visible-condition checks can record damage, corrosion, contamination, incomplete protection or missing components, but only against criteria that the order or inspection instruction actually defines.

A visually sound part may still have an incorrect bore, material or heat-treatment record. Conversely, a paperwork mismatch may prevent release even when the surface appears acceptable. Record each result against its own requirement instead of collapsing them into one pass mark.

Critical dimensions and agreed tests

For dimensions, state the drawing feature, revision, tolerance, measurement method, unit, measurement location and applicable equipment requirement. “Diameter checked” is not enough if the drawing controls several diameters or sections. If a certificate or test report is required, identify the document and the characteristic it supports; document presence is not a substitute for the physical checks assigned to incoming inspection.

Characteristics with safety, regulatory or severe service consequences may need a dedicated control, specialist review, traceability or complete verification. Do not assume that a general lot-sampling plan covers them. Their treatment must come from the applicable specification, contract and authorized risk decision.

Choose an inspection approach with explicit risks

The approach must match the data and the decision. Record who approved it, the reference and edition, the lot size or boundary, the sampling unit, the selection method and every input required by that plan. Do not copy a sample size or acceptance number from another order simply because its quantity looks similar.

Attributes versus measurements

In attributes inspection, each inspected unit or occurrence is classified—for example, conforming or nonconforming—and the relevant count is used in the decision. In variables inspection, the actual continuous measurements are used. The ASQ overview of attributes and variables sampling explains this data distinction and notes that variables plans carry assumptions and administration requirements that differ from attributes plans.

Choose the method characteristic by characteristic. A packaging-label match may be recorded as an attribute, while a specified roller-shell diameter produces measurement data. The existence of measurement values does not by itself make a variables acceptance plan valid; its stated assumptions and required inputs still have to be satisfied.

Acceptance sampling limits

Acceptance sampling trades some inspection effort for a defined risk of making the wrong lot decision. NIST describes a lot acceptance sampling plan as a sampling scheme plus rules for accepting, rejecting or, in some multiple-stage plans, taking more samples. It also distinguishes producer’s risk—the chance of rejecting a lot at a specified acceptable level—from consumer’s risk—the chance of accepting a lot at a specified unacceptable level. These risks belong to the selected plan and its assumptions; they are not removed by a clean sample.

Suppose the approved plan selects units at random and every inspected unit conforms. The valid statement is: “No nonconformities were found in the inspected sample, and the lot met the preapproved acceptance rule.” It is not: “The lot contains zero defects.” Uninspected units remain unobserved, and the sampling plan’s decision risk remains.

Record the decision rule before seeing the sample

Write the rule before inspection so the result cannot influence the threshold. The following incoming-lot decision worksheet uses a hypothetical roller-shell delivery. It demonstrates the record structure; it does not prescribe an AQL, sample quantity or acceptance number.

Reference standard

Name the actual standard, contract procedure or approved internal plan and its revision. Record the characteristics to which it applies and the required inputs. A standard name by itself is incomplete if the edition, inspection level, risk points, distribution assumptions or switching state needed by the chosen method have not been established by the responsible authority.

If the order does not specify an adequate basis, keep the decision open and escalate the missing inputs. Do not select a convenient AQL or sample count during receiving. A bulk undercarriage parts sourcing guide can help with wider order preparation, while the current record should remain focused on this lot’s inspection decision.

Acceptance and rejection actions

Write the acceptable result, non-acceptable result and any permitted “continue sampling” route exactly as the approved plan defines them. Then connect each outcome to a physical action: release the defined lot, quarantine it, request the next authorized sample stage, or route it for technical disposition. Also state who can authorize that action.

Do not change a defect category after observing the sample merely to reach a preferred outcome. If a result was recorded against the wrong requirement or method, mark it invalid, preserve the original record and perform the authorized correction or reinspection.

Handle rejected or uncertain lots

A rejection means the lot did not meet the stated acceptance rule; it does not identify every nonconforming unit or automatically prescribe repair. An uncertain result means a required input, record or authorized decision is missing. Both conditions need controlled physical status and a traceable next action.

Quarantine

Identify the affected inspection-lot ID, quantity, location and physical containers. Prevent routine issue or installation while the decision is open. If only one supplier batch or part number is affected, avoid blocking unrelated populations without evidence, but do not split the affected lot after seeing results unless an authorized rule and traceability support that action.

Link photographs, measurement data, nonconformity records and supplier communications to the same decision record. Record discrepancies as observations rather than guesses about cause.

Authorized sorting, reinspection and release

The designated quality or engineering authority may choose sorting, additional inspection, supplier return, concession review or another documented disposition. The correct route depends on the characteristic, risk, agreement and traceability. A rejected sampled lot is not made acceptable simply by removing the sampled nonconforming pieces.

If 100% sorting is authorized, define the characteristic, method, acceptance criteria, identification of conforming units and treatment of rejected units before work starts. If reinspection is authorized, identify the new basis and explain why it is valid; repeating the same sample until it passes is not a decision rule.

Close the record only when the physical quantity, evidence, disposition and release authority agree. The final statement should say which lot was released under which rule. It should never turn a limited sample result into a claim that the entire shipment is defect-free.