Smart lock quality inspection in China is a five-pillar B2B decision the buyer designs before production starts, not a check the buyer performs at the

Julkaisuajankohta: marras 20, 2024

Smart Lock Quality Inspection China

Smart Lock Quality Inspection China — A Five-Pillar Decision Framework for B2B Inspection of Factory-Run Smart Lock Production

Smart lock quality inspection in China is a five-pillar B2B decision the buyer designs before production starts, not a check the buyer performs at the loading dock. The five pillars are (1) the timing decision — what is honestly in-process vs pre-shipment vs at-source, and what each tier actually catches; (2) the AQL framing — what Acceptance Quality Limit is, what it is not, and how the sampling plan must be written for a lock, not for a t-shirt; (3) the inspection-content pillar — what the inspector actually tests on each SKU and what the buyer must demand to see written before the inspector signs; (4) the documentation and traceability pillar — what records travel with the shipment, what the supplier retains, and what the buyer keeps for the next RFQ; (5) the disposition and corrective-action pillar — what the inspector recommends when a defect is found, what the supplier commits to, and how the buyer enforces the next lot.

Every per-SKU inspection figure lives on the product page; every per-lot figure lives on the inspector's report; every per-door figure lives on the door. The framework exists because the wrong pairing — a pre-shipment inspection scheduled against a finished-goods backlog, an AQL sampling plan written for a uniform-attribute good run against a multi-component SKU with electronic subassemblies, a sampling plan borrowed from a generic softline protocol applied to a lock with safety-relevant mechanical load paths, an inspector reading visual cosmetic criteria against a credential reader that fails after a 200-cycle stress — produces a pass report that does not catch the failure the buyer actually needed to catch. The pass report that misses the failure is the most expensive pass report a buyer can pay for.

This is a decision-pillar explainer. The subheads are pillar names, not questions. There is no FAQ section by design: every buyer question that belongs in a FAQ lives inside the pillar where the answer actually applies, so an AQL-sampling question is answered in the AQL pillar, a pre-shipment-vs-in-process question in the timing pillar, and a corrective-action question in the disposition pillar. Tables are the working surface of each pillar. The voice is the supplier who has walked more inspection reports than the brochure ever quoted and who has read what an inspector actually writes when the lot lands in front of them, not what the supplier's QC manager says the inspector saw.

For the sourcing context this framework lives inside — the four-stage sourcing flow, the supplier-verification work, and the factory-audit posture — read alongside How to Source Smart Locks From China, Smart Lock Supplier Verification China, and Smart Lock Factory Audit Checklist for International Buyers.

What an inspection actually catches — the honest frame

Before any pillar resolves, the buyer has to fix in writing what the inspection is for. A smart lock is not a single-attribute consumer good; it is a multi-component electromechanical assembly with at least four subsystems that each fail in their own way: (a) a mechanical subassembly (latch, bolt, cylinder, mortise, handle, spindle, fasteners), (b) an electronic subassembly (PCBA, microcontroller, EEPROM, radio module, antenna trace, power-management circuit), (c) a credential subassembly (fingerprint sensor, RFID reader, keypad, Bluetooth radio, app layer, optical/3D face module, mechanical-key override), and (d) a cosmetic-and-finish subassembly (housing substrate, surface finish, marking, packaging, labelling). Each subsystem has its own defect taxonomy, its own fail mode, its own inspection station, and its own inspector qualification. An inspection that treats the lock as one uniform-attribute good will systematically under-inspect on the mechanical subassembly and over-inspect on cosmetic criteria, because cosmetic defects are easier to count from across a table than mechanical-pull-strength defects.

The honest frame also includes the timing distinction. There are four honest timing tiers: (i) pre-production — when the supplier has the order and the raw materials but has not started assembling, used to verify in writing raw-material conformity and incoming-component acceptance; (ii) in-process / during production (DUPRO) — when the line is running, used to catch line-level defects before the lot accumulates; (iii) pre-shipment (PSI) — when finished goods are packed, used to catch lot-level defects before they leave the factory; and (iv) container loading / at-source supervision — when finished goods are loaded into the container, used to catch quantity, marking, and packaging defects. Each tier catches a different defect class. Each tier has a different cost of failure. The buyer has to choose which tier catches which defect class, and the inspector's scope is the written expression of that choice.

The framework pillar structure exists to force that distinction into the buyer's RFQ before the inspection is scheduled, and to make the distinction legible on the report the buyer reads.

The honest supplier acknowledges that. The honest buyer writes that acknowledgement into the RFQ.

Pillar 1 — Timing the inspection honestly: pre-production, in-process, pre-shipment, and at-source

The first pillar answers one question only: at which point in the production timeline is each inspection tier scheduled, what defect class does that tier actually catch, and what is the buyer's call between in-process and pre-shipment when only one tier can be funded? Timing is the buyer's input. Inspector capability is the lead's input. The choice between in-process and pre-shipment is the buyer's procurement call, and it must be written into the RFQ, the PO, and the supplier's quality agreement before the line starts.

1.1 The four timing tiers and what each one actually catches

Pre-production inspection sits at the moment the supplier has raw materials, sub-components, and electronic parts in house but has not started assembling. The inspection covers incoming-component conformity (PCBA lot, RFID reader batch, fingerprint sensor batch, cylinder blanks, lever castings, zinc-alloy or stainless-steel housings, packaging cartons, inner foam, manuals). This tier catches lot-level component defects that propagate through assembly if they are not caught early. A fingerprint sensor batch with a systematic defect — say, a sensor-window glass batch with a coating problem — will pass pre-production but fail every assembled lock it lands in, and the failure is invisible until the buyer tries to enrol a fingerprint. The pre-production tier is the cheapest tier per defect caught because the lot has not been assembled yet.

In-process / DUPRO inspection sits at the moment the line is producing finished or semi-finished units, typically when 20% to 50% of the lot is complete. The inspection covers line-level assembly defects: solder joint quality on through-hole and surface-mount components, fastener torque, adhesive cure, gasket seating, cable harness seating, screw torque on the cylinder cap, mechanical key insertion continuity on the override cylinder, and a sample of functional tests on finished or near-finished units. This tier catches line-level defects that the supplier's own first-article inspection may have missed and that lot-end inspection would catch too late to rework at reasonable cost.

Pre-shipment inspection (PSI) sits at the moment finished goods are packed and ready for container loading. The inspection covers lot-level cosmetic and functional conformity: carton and inner-pack count, marking and labelling, visible defects on the assembled unit (housing, finish, marking), functional spot-checks on a sample (card read, fingerprint read, mechanical-key rotation, latch actuation), packaging integrity, and per-carton quantity. This tier catches the lot-level defects that propagate through to the buyer's warehouse. PSI is the most common tier B2B buyers pay for; it is also the tier that catches the fewest defects per dollar spent, because most lot-level defects are downstream of a defect that pre-production or in-process would have caught.

Container loading / at-source supervision sits at the moment finished goods are loaded into the shipping container. The inspection covers quantity verification against the packing list, container condition, container-floor and wall integrity, loading sequence, sealing, and a final visual sanity check on the outermost cartons. This tier catches quantity and shipment-handling defects; it does not catch assembly defects because the units are already packed.

1.2 Why in-process and pre-shipment are not substitutes

A common error is to fund only PSI on the assumption that PSI catches everything the buyer needs. It does not. A defect that emerges at the in-process stage — say, a PCBA solder defect that causes a 3% read-failure rate after assembly — will be visible at in-process and invisible at PSI if the inspector runs only a visual sample. PSI's sampling plan catches defects that exceed the AQL threshold; an in-process defect rate of 3% on a 1,000-unit lot may not exceed PSI's accept threshold if the sampling plan is written loosely, but it will fail in the buyer's hands when 30 doors out of 1,000 reject a credential. The honest frame is to schedule in-process on at-risk SKUs and pre-shipment on every SKU, with the in-process output feeding the PSI sampling plan.

A second common error is to assume the inspector's report is the inspection. The report is the artefact that travels; the inspection is the event on the floor. An inspector who arrives at 10% of the lot complete and signs a report that says "passed" is reporting on what they saw at 10% — which is not the lot. An inspector who arrives at PSI and is given cartons that have been pre-selected by the supplier's QC team is inspecting the supplier's QC, not the lot. The buyer's call on inspector selection, inspector timing, and inspector scope is what the pillar is buying.

1.3 Timing tier scorecard

The scorecard below is the working surface for the timing pillar. The buyer completes one row per tier. Every column is the buyer's call against the supplier's production schedule.

TierWhen in the timelineDefect class it catchesWhat it does NOT catchCost per defect caughtBuyer's call
Pre-productionraw materials in-house, no assembly yetincoming-component conformity, lot-level component defectsline-level assembly defectslowestschedule or skip
In-process / DUPRO20-50% of lot completeline-level assembly defects, functional spot-checks, solder-joint and fastener-torque defectscosmetic-defect drift, packaging defectsmediumschedule on at-risk SKUs
Pre-shipment (PSI)finished goods packed, lot completelot-level cosmetic + functional defects, packaging integrity, markingsubcomponent lot-level defects that propagate unseenmedium-highschedule on every SKU
Container loadinggoods loaded into containerquantity, marking, container condition, sealingassembly defects, packaging integrity inside cartonshigh per defect caught (because the units are packed)schedule on every SKU

A SKU that lands on a single-tier timing plan (only PSI) is exposed to defects that tier does not catch. A SKU that lands on a multi-tier plan (pre-production + in-process + PSI + loading) is exposed only to defects that escape all four tiers — and at that point the defect is a process problem the supplier needs to fix, not an inspection problem the buyer needs to schedule against.

1.4 What the timing pillar does not say

The pillar states no single timing tier that fits every SKU. It does not promote PSI as a substitute for in-process on solder-joint-critical subassemblies. It does not promote pre-production as a substitute for in-process on assembly-critical subassemblies. It does not predict how many defects each tier catches on any particular lot. The timing tier is the buyer's call against the supplier's production schedule and the SKU's defect history; the AQL sampling plan of Pillar 2 is the statistical instrument that joins the timing tier to the inspector's sample; the inspection-content pillar of Pillar 3 is the per-SKU list the inspector runs against the sample; the documentation pillar of Pillar 4 is the records that travel with the lot; the corrective-action pillar of Pillar 5 is what happens when a defect is found.

For the broader in-process production context — what the inspector sees on the line, what an inspector signs off on a 25% sample, what the trip report reads — read Smart Lock Production Inspection and Smart Lock Factory Inspection Checklist.

Pillar 2 — AQL framing honestly: what the sampling plan actually does

The second pillar answers: what Acceptance Quality Limit is, what it is not, what a sampling plan written for a smart lock must include, and what the inspector must run against the buyer's lot. AQL is a statistical instrument. It is not a quality stamp. It is not a guarantee that any single unit is defect-free. It is a sample-size rule that bounds the probability of accepting a lot that exceeds the buyer's allowable defect rate, given the lot size and the inspection level. The honest use of AQL forces the buyer to write the sampling plan before the inspector signs anything; the dishonest use of AQL borrows a generic plan from a softline protocol and applies it to a multi-component electronic-mechanical assembly.

2.1 What AQL actually is and is not

AQL is a percentage figure that pairs with a sample-size code letter and a lot-size band to produce a sample size and an accept/reject number. The buyer picks (a) the AQL percentage, (b) the inspection level (general inspection levels I, II, III, and special inspection levels S-1, S-2, S-3, S-4), and (c) the lot size. From those three inputs the sampling plan produces a sample size and an accept number. The sample is drawn randomly from the lot; each unit in the sample is inspected against the defect classification; defects are counted; the lot is accepted if the count is at or below the accept number and rejected if it exceeds the reject number.

What AQL is not: it is not a measure of the supplier's quality, it is not a per-unit quality guarantee, it is not a substitute for in-process inspection, it is not a substitute for sampling on the supplier's own first-article and last-article checks, it is not a substitute for functional testing on critical-path subassemblies, and it is not a single number the buyer publishes across an entire catalog. The buyer must pick the AQL separately for each defect class (critical / major / minor) and must pick the inspection level separately for cosmetic vs functional defects.

2.2 The three defect classes a smart lock needs

A smart lock inspection runs against three defect classes, and the AQL figure must be written separately for each:

Critical defects are defects that create safety risk, security risk, or non-conformance with mandatory import regulation. For a smart lock, the canonical critical defect class includes: lock failure to engage the bolt under power, override-cylinder failure (mechanical key cannot open the door), electronic subassembly that fails the buyer-mandated functional spot-check, RFID credential reader that fails to read any valid card, fingerprint sensor that fails to enrol, battery-compartment polarity reversal that would damage the cell, unauthorised access through a backdoor command sequence, and any defect that prevents emergency egress. The AQL figure for critical defects is typically the strictest (lowest percentage) of the three classes, because a single critical defect that lands at the buyer's door is a recall-grade event.

Major defects are defects that affect function, durability, or merchantability but do not create safety risk. The canonical major defect class includes: cosmetic defects visible at arm's-length on the front face (scratches, dents, finish mismatch), marking defects (logo misprint, model-number misprint, batch-label misprint), latch-actuation torque out of spec, gasket-seating defects that allow dust ingress under IP-rated conditions (where IP-rating is a documented posture on that SKU), packaging-defect that allows the unit to be dropped or impacted during shipping, and battery-low-warning behaviour out of spec (where warning-channel class is documented on the SKU).

Minor defects are defects that affect appearance but do not affect function, durability, or merchantability. The canonical minor defect class includes: small cosmetic blemishes not visible at arm's-length, slight colour drift between units (within a tolerance band), inner-pack imperfections that do not affect the unit, and carton-printing blemishes that do not obscure critical information.

2.3 AQL worksheet — per-defect-class sampling plan

The worksheet below is the working surface for the AQL pillar. The buyer completes one row per defect class per SKU. Every column is the buyer's call against the supplier's lot size and the SKU's defect history.

| Defect class | AQL percentage | Inspection level | Sample size (per lot size band) | Accept number | Reject number | Source || --- | --- | --- | --- | --- | --- || Critical | check the model’s published specification | check the model’s published specification | check the model’s published specification | check the model’s published specification | check the model’s published specification | product page + buyer's RFQ || Major | check the model’s published specification | check the model’s published specification | check the model’s published specification | check the model’s published specification | check the model’s published specification | product page + buyer's RFQ || Minor | check the model’s published specification | check the model’s published specification | check the model’s published specification | check the model’s published specification | check the model’s published specification | product page + buyer's RFQ |

The buyer does not pre-fill any cell. The sampling plan the inspector runs against the lot is the plan the buyer wrote against the supplier's product page, the buyer's risk tolerance, and the SKU's defect history. A SKU with no published sampling plan on the product page is a SKU the inspector cannot sign a pass report against.

2.4 What the AQL pillar does not say

The pillar does not assert an AQL percentage for any defect class on any catalog SKU. It does not promote a single inspection level across the catalog. It does not predict a pass rate for any supplier's lot. Nor does it present AQL as a substitute for in-process inspection or for functional spot-checks on critical-path subassemblies. The sampling plan is the buyer's instrument; the inspector is the operator; the lot is the population; the supplier is the producer. The pillar names the framework; the figures the buyer puts in the worksheet cells are the buyer's call.

Pillar 3 — Inspection content: what the inspector actually tests per SKU

The third pillar answers: what does the inspector's written tests cover per SKU, what is the per-SKU differentiation between a generic AQL plan and a SKU-specific plan, and what is the buyer's call on what to demand the inspector tests against? Inspection content is the per-SKU inspection scope. A generic AQL plan borrowed from a softline protocol will test colour, marking, and visible defects; it will not test the credential reader's read-distance behaviour, the latch-actuation torque, the fingerprint enrolment latency, or the cylinder's mechanical-key cycle count. The buyer has to write the per-SKU scope before the inspector is briefed.

3.1 The seven inspection content areas a smart lock needs

A smart lock inspection covers at least seven content areas, and each area needs at least one written test the inspector can run against the sample. The areas are:

Visual / cosmetic conformity — housing-substrate finish, surface coating, brushed or polished direction, marking legibility, logo presence and clarity, batch-label accuracy, colour drift between units, edge burrs, parting lines.

Dimensional conformity — overall housing dimensions, mounting-hole pattern, door-thickness compatibility (the SLK-9936S spec block for example reads door thickness 36-100 mm; every other catalog SKU's door-thickness compatibility lives on the per-SKU product page), spindle interface, lever projection, backset, latch-bolt throw length, and any per-SKU dimensional envelope the lock must fit into.

Mechanical functional conformity — handle rotation, free-rotation posture (where the handle is designed to free-rotate under load to resist forced entry), spindle drive, latch actuation under power, deadbolt throw, mechanical-key override insertion and rotation continuity across the cylinder cycle count, three-point-latch engagement (where the SKU has a three-point latch), and mechanical-cycle behaviour under repeated load.

Electronic functional conformity — wake behaviour, credential authentication latency on each enabled method (RFID, fingerprint, PIN, Bluetooth/app, mechanical-key), motor-actuation behaviour, battery-low-warning posture on each enabled channel (LED / audible / app / server log), emergency-power handoff (USB, mechanical key, 9-volt clip), credential-storage integrity after power-cycle, audit-event-record count behaviour (the SLK-9936S spec block for example reads audit/event records up to 1,000; every other catalog SKU's audit-event posture lives on the per-SKU product page), and EEPROM retention behaviour across power interruption.

Credential method conformity — RFID read distance and read time (the SLK-9936S spec block for example reads RFID Mifare 1 @ 13.56 MHz / read distance 0-5 cm / read time <0.25 s; every other catalog SKU's RFID posture lives on the per-SKU product page), fingerprint enrolment-and-match latency, PIN entry acknowledgement, Bluetooth wake distance, app-layer credential-issue and credential-revoke latency, hotel PC-software credential synchronisation, and 3D-face-recognition enrolment and match posture where the SKU supports it.

Durability / stress conformity — cycle-test behaviour on the latch and the handle, the cylinder cycle count, the mechanical-key insertion-and-rotation cycle count, salt-spray posture where the SKU is rated for outdoor use, high-temperature dwell posture, low-temperature dwell posture, humidity-read posture, and drop-test posture on the packed unit.

Packaging, marking, and traceability conformity — per-carton quantity, inner-pack integrity, marking legibility on the carton, country-of-origin marking, model-and-batch labelling, manual-and-accessory inclusion, ISPM 15 marking on wood packaging where applicable, battery-cell shipping classification marking, and per-unit serial-number traceability.

3.2 Per-SKU inspection scope worksheet

The worksheet below is the working surface for the inspection-content pillar. The buyer completes one row per content area per SKU. Every column is the buyer's call against the per-SKU product page.

Inspection content areaWhat is documented
Visual / cosmeticSource: product page + buyer's RFQ
DimensionalSource: product page + buyer's RFQ
Mechanical functionalSource: product page + buyer's RFQ
Electronic functionalSource: product page + buyer's RFQ
Credential methodSource: product page + buyer's RFQ
Durability / stressSource: product page + buyer's RFQ
Packaging / markingSource: product page + buyer's RFQ

A SKU with an incomplete row — any cell reading check the model’s published specification without a per-model declaration attached — cannot be inspected against a written scope until the buyer has the declaration from the supplier. The buyer does not pre-fill any cell.

3.3 The SLK-9936S documentation-granularity yardstick

For the SLK-9936S hotel lock, the first-party spec block names door thickness 36-100 mm, RFID Mifare 1 @ 13.56 MHz, read distance 0-5 cm, read time <0.25 s, power 4 x AA, battery operating cycles >15,000, audit/event records up to 1,000, 5-pin cylinder, 3-point latch, stainless steel + alloy body, multiple finish options, mechanical emergency access, multilingual management software, standalone or centralized database modes, and 1-5 star hotel positioning. The yardstick is the shape of evidence the buyer should demand for every other catalog SKU; it is not a portable specification across siblings.

The 5-pin cylinder, 3-point latch, and stainless-steel-plus-alloy-body lines are scoped to the SLK-9936S only. They do not extend to the SLK-9936G, the SLK-9905A, the SLK-9951TH, the SLK-9962S, the SLK-G200, the SLK-8831, the SLK-1104 class, the SLK-3108, the cylindrical Tuya round-code lock, the 3D-face-recognition / remote-intercom lock, the B09 glass lock, the D11 / D12 / D15 / ODS / DQ outdoor series, the 9L5-R / Q7 / Future 1 face family, the 516 cat-eye, the Q7 intercom, the 4G model, the cabinet series, the RFTC / RF card-and-code 118 / 118E / 228 / 230 / 232 / 238 / 239 / 253 / 305 / 306 / 308 / 309 / 502 / 620 / 621 / 911 / 002 / 019 / 022 / 2020 / 760 family (with stainless 019 / 022 / 309 noted as a class), or the fingerprint 37-model family. Every per-SKU inspection scope those siblings might need lives on the per-SKU product page.

3.4 What the inspection-content pillar does not say

The pillar names no sample size, no accept number, and no pass criterion for any catalog SKU. It does not predict a defect rate on any supplier's lot. It does not promote any single content area above another for any catalog SKU without the buyer's own per-SKU scope. The per-SKU scope is the buyer's call; the inspector is the operator; the supplier's product page is the source of truth for per-SKU figures; the buyer's RFQ is the source of truth for per-lot accept/reject thresholds.

For the per-SKU inspection-content depth — what the inspector reads on the line, what they test on the bench, what they hand back to the supplier at the end of each day — read Smart Lock Product Sample Inspection and Smart Lock Factory Quality Control: What Buyers Should Check.

Pillar 4 — Documentation, traceability, and the records that travel with the lot

The fourth pillar answers: what documents travel with the lot, what records the supplier retains, what records the buyer keeps for the next RFQ, and what chain-of-custody posture the inspection report must establish? Documentation is the buyer's audit trail. A lot that ships without a written inspection report is a lot the buyer cannot defend if the shipment fails on arrival; a lot that ships with an inspection report that does not name the inspector, the sample size, the AQL plan, the defect-classification schema, and the supplier's corrective-action commitment is a report the buyer cannot use.

4.1 The six documents that should travel with the lot

The lot should travel with at least six documents, and each document should be named in the buyer's RFQ before the inspection is scheduled:

Inspection report. The inspector's signed report naming the inspector, the inspection company, the date, the tier (pre-production / DUPRO / PSI / loading), the lot size, the sample size, the AQL plan used, the defect classification schema, the defects counted per class, the accept/reject decision, and the inspector's signature with date.

Packing list. The supplier's signed packing list naming the SKU, the per-carton quantity, the total carton count, the carton dimensions, the per-carton gross weight, the total gross weight, the volume, and any per-carton serial-number range. The packing list is the document the loading inspector reconciles against the actual loading.

Commercial invoice. The supplier's commercial invoice naming the SKU, the unit price, the total value, the currency, the Incoterm, the port of loading, the port of destination, the country of origin. The commercial invoice is the document the customs broker reconciles against the HS-code declaration; it is also the document the buyer's finance team reconciles against the PO.

Certificate of origin. Where required by the destination country's customs regime, the certificate of origin naming the country of manufacture, the manufacturer, the SKU description, the HS code, and the issuing chamber of commerce. The certificate is the document the buyer's broker files for FTA preference where applicable.

Test summary. The supplier's dated test summary naming the SKU, the lot, the tests run (cycle, salt-spray, temperature, humidity, drop, packaging drop, RFID read, fingerprint enrolment, mechanical-key cycle, battery-cycle), the pass/fail per test, and the test-rig or test-method reference. The test summary is the document the buyer's per-SKU RFQ demands when the SKU has a documented cycle-test posture.

Per-unit serial-number list. The supplier's per-unit serial-number list naming the serial-number range per carton, the lot number, and the production date. The serial-number list is the document that lets the buyer trace a field failure back to a specific lot, a specific production date, and a specific test-summary.

4.2 The records the supplier should retain

Beyond the documents that travel with the lot, the supplier should retain at least five record classes the buyer can request post-shipment: (a) first-article inspection records, (b) in-process / DUPRO inspection records, (c) pre-shipment inspection records, (d) supplier-internal QC station records (often called IPQC records), and (e) raw-material incoming-inspection records. The buyer's RFQ should name which records the supplier retains, the retention period, and the buyer's right to request a copy post-shipment. The supplier's retention posture is the buyer's audit-trail recourse when a field failure surfaces three months after delivery.

4.3 Documentation worksheet — per-lot record set

The worksheet below is the working surface for the documentation pillar. The buyer completes one row per document class per lot. Every column is the buyer's call against the supplier's documentation discipline.

DocumentTravels with lotSupplier retainsBuyer retainsRequest cadence
Inspection reportyesyes (per supplier policy)yesper shipment
Packing listyesyesyesper shipment
Commercial invoiceyesyesyesper shipment
Certificate of originwhere requiredyesyesper shipment where required
Test summaryyes (when SKU has documented test posture)yesyesper shipment
Per-unit serial-number listyesyesyesper shipment

A document that does not travel with the lot, or that does not get retained by the supplier, is a document the buyer cannot defend at audit. The worksheet is the buyer's checklist against the supplier's per-shipment documentation discipline.

4.4 What the documentation pillar does not say

The pillar takes no position on a specific document format (paper vs digital) for any catalog SKU. Retention periods on any supplier are outside the scope of this article. It does not predict which documents the supplier will produce on the next shipment. The documentation discipline is the buyer's call; the supplier's documentation discipline is what the buyer's RFQ demands; the inspector's report is the inspector's call. The pillar names what documents should travel; the worksheet is the buyer's instrument.

For the per-shipment record discipline — what the loading inspector reconciles, what the broker reconciles, what the importer keeps — read Smart Lock Container Loading Inspection and Smart Lock Packaging Inspection.

Pillar 5 — Disposition, corrective action, and the next lot

The fifth pillar answers: when a defect is found, what does the inspector recommend, what does the supplier commit to, and how does the buyer enforce the next lot? Disposition is the buyer's contract with the supplier on what happens when the inspector finds something. Without a written disposition clause, the buyer has no contractual lever to demand rework, replacement, sorting, or refund; with a written disposition clause, the buyer has the lever but only if the clause names the action, the timeline, and the evidence the supplier must produce to close the loop.

5.1 The four disposition actions

When a defect is found at any tier, the inspector recommends one of four disposition actions, and the buyer's RFQ must name which action is acceptable for which defect class:

Accept (pass). The lot is accepted with no further action. Accept is appropriate when the AQL threshold is met for all defect classes and the inspector has found no critical defects. Accept is the default disposition on a clean lot.

Sort. The supplier sorts the lot — 100% inspection of every unit — and the sorted lot is re-presented for inspection. Sort is appropriate when a major-defect pattern is identified that the AQL sampling plan did not catch. Sort is expensive and is the buyer's call when the cost of sort is lower than the cost of rejecting the lot.

Rework. The supplier reworks the defective units in the lot to bring them into specification. Rework is appropriate when the defect is mechanical or cosmetic and the rework is technically feasible. Rework is not appropriate for electronic subassembly defects where rework may compromise the integrity of the rework itself.

Reject. The lot is rejected and either replaced by the supplier at supplier cost, refunded to the buyer, or scrapped. Reject is appropriate when the defect class is critical, the defect rate exceeds the AQL reject threshold, or the defect pattern is systemic. Reject is the buyer's strongest lever; reject without a written replacement/refund commitment is a paper exercise.

5.2 Corrective-action commitment

Beyond the immediate disposition, the buyer's RFQ should demand a corrective-action commitment from the supplier: (a) a written root-cause analysis for any critical or systemic defect, (b) a written corrective-action plan naming the action, the owner, the timeline, and the verification method, (c) a written commitment to apply the corrective action to the next lot and to communicate the change to the buyer, and (d) a written commitment to update the supplier's internal QC documentation to reflect the corrective action. Without the corrective-action commitment, the defect is repeated next lot. With the commitment, the defect is closed at root cause.

5.3 Disposition and corrective-action worksheet — per-defect finding

The worksheet below is the working surface for the disposition pillar. The inspector and the buyer complete one row per defect finding. Every column is the inspector's call against the buyer's RFQ and the supplier's quality agreement.

FindingDefect classDisposition recommendedCorrective-action ownerCorrective-action due dateVerification evidenceClosure status
Finding-1critical / major / minoraccept / sort / rework / rejectsupplier's QC managerper RFQper RFQopen / closed
Finding-2critical / major / minoraccept / sort / rework / rejectsupplier's QC managerper RFQper RFQopen / closed
Finding-3critical / major / minoraccept / sort / rework / rejectsupplier's QC managerper RFQper RFQopen / closed

A finding that lands in open carries over to the next lot. A finding that lands in closed is closed only when the verification evidence has been produced and the buyer has signed off on the closure.

5.4 What the disposition pillar does not say

Which disposition action applies to any catalog SKU is a per-lot call this pillar leaves open. It does not predict a defect rate on the next lot. It sets no corrective-action timeline either — that is fixed in the buyer's RFQ and quality agreement. It does not promote a single disposition action above another for any defect class without the buyer's own RFQ. The disposition is the inspector's call against the buyer's RFQ; the corrective-action commitment is the supplier's call against the buyer's quality agreement; the closure is the buyer's call against the verification evidence. The pillar names the framework; the worksheet is the buyer's instrument.

5.5 The five non-negotiables

The five non-negotiables govern the corrective-action discipline. Each one is non-negotiable because dropping it allows a defect to propagate to the next lot.

Non-negotiable 1. Per-SKU scope, not catalog scope. Every inspection pillar runs against the per-SKU product page. Catalog-level scope is rejected.

Non-negotiable 2. Per-lot disposition, not blanket disposition. Every lot gets its own disposition worksheet. A clean lot on SKU-1 does not pre-fill the disposition for SKU-2.

Non-negotiable 3. AQL per defect class, not single AQL. AQL is written separately for critical, major, and minor defects. A single AQL figure is rejected.

Non-negotiable 4. Documented corrective action, not verbal commitment. The corrective-action commitment is in writing on the supplier's letterhead, with a named owner and a dated verification step.

Non-negotiable 5. Closure on evidence, not on agreement. The closure status is closed only when the verification evidence has been produced, attached, and signed off by the buyer.

For the contractual envelope the corrective-action discipline sits inside — what the supplier commits to in writing, what the buyer enforces at the next lot, what the quality agreement names as a contract artefact — read Smart Lock Pre-Shipment Inspection China and the supplier-quality-agreement cluster sibling.

How the five pillars hand off to each other

The pillars are not five independent questions. They are a sequential pipeline where each pillar produces the input the next pillar consumes.

Pillar 1 produces the timing-tier plan that names when the inspection happens. Pillar 2 consumes the timing tier and the lot size to produce the AQL sampling plan and the sample size. Pillar 3 consumes the sample size to produce the per-SKU inspection scope. Pillar 4 consumes the per-SKU scope to produce the per-lot document set. Pillar 5 consumes the per-lot document set and the inspector's report to produce the per-defect disposition worksheet.

The five-pillar handoff reads as one pipeline when each pillar's artefact is lined up against the input the next pillar consumes:

flowchart TD
    P1["Pillar 1 · Timing<br/>timing-tier plan · buyer's call"] --> P2["Pillar 2 · AQL framing<br/>sampling plan · sample size"]
    P2 --> P3["Pillar 3 · Inspection content<br/>per-SKU scope"]
    P3 --> P4["Pillar 4 · Documentation<br/>per-lot document set"]
    P4 --> P5["Pillar 5 · Disposition<br/>per-finding worksheet"]
    PP["Per-SKU product page · buyer's RFQ"] -.-> P2
    PP -.-> P3
    P5 -.->|"supplier change re-opens the previous pillar"| P1

The discipline of the pipeline is that no pillar pre-fills its output until the previous pillar's output is closed. A sample size that pre-fills Pillar 2 before the AQL percentage is written in Pillar 2 is a sample size that fails at the inspector's bench. An inspection scope that pre-fills Pillar 3 before Pillar 2 is closed is a scope that fails at the bench. A document set that pre-fills Pillar 4 before Pillar 3 is closed is a document set that fails at audit. A disposition that pre-fills Pillar 5 before Pillars 1 to 4 are closed is a disposition that fails at the next lot.

The loop-back arrow in the framework diagram is the rule that any pillar can re-open the previous pillar if the supplier changes a SKU, a lot size, an AQL plan, an inspection-content area, a document class, or a corrective-action commitment. The re-open is a normal event, not a failure. The discipline is to re-open the previous pillar when something changes, and to close the next pillar only when the previous pillar's output is settled.

Illustrative scenario — a 1,500-unit hotel lock order across two door families

Consider a 1,500-unit order for a 220-room resort refurbishment, with a planned split of 1,200 standard hotel guestroom locks and 300 back-of-house service corridor locks. Two door families. Family A is 1,200 standard guestroom locks on interior corridor doors, 45 mm solid-core timber, lever-actuated, expected turnover moderate, environmental envelope temperate indoor. Family B is 300 back-of-house service corridor locks on a loading-dock side, 50 mm composite-clad aluminium, lever-actuated, expected turnover high, environmental envelope hot-humid service.

The buyer's timing read of Pillar 1 lands Family A on a two-tier plan (pre-production + PSI) because Family A's defect history is clean and the SKU is mature. Family B lands on a three-tier plan (pre-production + DUPRO + PSI) because Family B's target SKU carries a fingerprint reader (a credential subassembly with a higher defect-history variance than a pure-RFID reader) and an outdoor-exposed envelope (which raises the importance of in-process solder-joint and gasket-seating inspection).

The buyer's AQL read of Pillar 2 lands Family A on a critical-class AQL percentage that the buyer writes against the product page, a major-class AQL percentage that the buyer writes against the product page, and a minor-class AQL percentage that the buyer writes against the product page. Family B lands on the same AQL-class structure with the same documentation discipline — every cell the buyer fills comes from the per-SKU product page, not from a generic softline protocol.

The buyer's inspection-content read of Pillar 3 lands Family A on a seven-area scope where the credential-method area covers RFID read distance and read time (the SLK-9936S spec block for example reads RFID Mifare 1 @ 13.56 MHz / read distance 0-5 cm / read time <0.25 s, scoped to that SKU alone), and where the mechanical-functional area covers the 5-pin cylinder cycle and the 3-point latch engagement (scoped to the SLK-9936S alone). Family B's inspection-content scope adds fingerprint-enrolment latency, fingerprint-match latency, salt-spray posture (because Family B's envelope is hot-humid service with periodic cleaning-chemical exposure), and gasket-seating inspection — each of these added scopes is filled from the per-SKU product page, not from a generic lock protocol.

The buyer's documentation read of Pillar 4 lands both families on the six-document set, with Family A's test summary reflecting the SLK-9936S spec block on RFID / cycle count / audit-event posture (scoped to that SKU alone), and Family B's test summary reflecting the per-SKU test posture for the fingerprint-reader SKU (filled from that SKU's product page).

The buyer's disposition read of Pillar 5 lands both families on a four-action disposition worksheet with a corrective-action commitment. Family A's worksheet ends at closed on every finding because the lot lands clean. Family B's worksheet ends at closed on every critical finding because the lot lands within the AQL threshold, and at open on one major finding — a fingerprint-sensor enrolment-latency drift on 4% of the sample — that the supplier sorts and re-presents within the contractual rework window.

The 1,500-unit outcome is 1,200 Family A units accepted on the two-tier plan, 300 Family B units accepted on the three-tier plan after one sorted-and-reworked lot, six documents travelling with each family, one open finding closed within the contractual window, and one corrective-action commitment on file with the supplier's QC manager for the next lot. The pillars caught what a single-tier PSI plan would have ignored: the Family B fingerprint-sensor drift would have slipped past a loose AQL plan and surfaced as a 12-door enrolment-latency complaint three months after installation.

Cross-link map — how the pillars travel through the cluster

The five pillars are the buyer's inspection-decision structure. The C12 cluster is the supplier's sourcing-and-supply-chain space. The cross-link map names which cluster sibling each pillar reads against when the pillar needs to go deeper.

Pillar 1 (timing) reads against Smart Lock Production Inspection for the in-process / DUPRO deep dive, and against Smart Lock Factory Inspection Checklist for the day-of-inspection discipline.

Pillar 2 (AQL) reads against the supplier-quality-agreement cluster sibling for the contractual envelope the AQL plan sits inside.

Pillar 3 (inspection content) reads against Smart Lock Product Sample Inspection for the per-SKU sample-test deep dive, and against Smart Lock Factory Quality Control: What Buyers Should Check for the supplier-internal QC station discipline.

Pillar 4 (documentation) reads against Smart Lock Container Loading Inspection for the per-shipment document discipline at loading, and against Smart Lock Packaging Inspection for the per-carton marking and packaging posture.

Pillar 5 (disposition) reads against Smart Lock Pre-Shipment Inspection China for the contractual envelope the corrective-action commitment sits inside, and against How to Source Smart Locks From China for the broader sourcing-flow context this framework lives inside.

Per-model firewall — the rule this article will not break

The five pillars route every per-model answer to the product page. The article will not state an AQL percentage for any catalog SKU. The article will not state an inspection level for any catalog SKU. The article will not state a sample size, an accept number, or a reject number for any catalog SKU. The article will not state a defect-classification schema for any catalog SKU. The article will not state a pass criterion for any catalog SKU. The article will not state a pull-strength threshold in Newtons, a torque threshold in Newton-metres, a cycle-test cycle count, a salt-spray hours figure, a temperature-drop hours figure, or a humidity-read cycles figure for any catalog SKU. The article will not assert a certification, a price, a MOQ, a lead time, or a warranty year for any catalog SKU.

The SLK-9936S hotel lock is the only catalog SKU whose first-party spec block is invoked in this article. The block names the door thickness, the RFID Mifare 1 frequency and read distance and read time, the 4 x AA cell format, the >15,000 operating cycles, the audit/event record count, the 5-pin cylinder, the 3-point latch, the stainless steel + alloy body, the multiple finish options, the mechanical emergency access, the multilingual management software, the standalone or centralized database modes, and the 1-5 star hotel positioning on that model. All thirteen lines are scoped to the SLK-9936S only.

None of the thirteen lines extends to the SLK-9936G, the SLK-9905A, the SLK-9951TH, the SLK-9962S, the SLK-G200, the SLK-8831, the SLK-1104 class, the SLK-3108, the cylindrical Tuya round-code lock, the 3D-face-recognition / remote-intercom lock, the B09 glass lock, the D11 / D12 / D15 / ODS / DQ outdoor series, the 9L5-R / Q7 / Future 1 face family, the 516 cat-eye, the Q7 intercom, the 4G model, the cabinet series, the RFTC / RF card-and-code 118 / 118E / 228 / 230 / 232 / 238 / 239 / 253 / 305 / 306 / 308 / 309 / 502 / 620 / 621 / 911 / 002 / 019 / 022 / 2020 / 760 family (with stainless 019 / 022 / 309 noted as a class), or the fingerprint 37-model family. Every per-SKU inspection figure those siblings might need lives on the per-SKU product page.

What this framework does not do, by design

The framework does not state a universal AQL percentage. There is no such universal answer. There is a per-SKU, per-defect-class, per-risk-tolerance answer that joins a per-lot size and a per-inspection level into a per-SKU sampling plan. The framework exists to force that join to happen on paper, before the inspector is briefed.

The framework does not predict a defect rate. A supplier's defect rate is the supplier's input, derived from the supplier's process capability and the supplier's quality history. The framework reads the supplier's history; the history is not a substitute for the framework. The two artefacts are paired: the framework names what to ask for, the supplier's history names what the supplier has actually produced.

The framework does not replace the product page. Every per-SKU input lives on the product page. The framework reads the product page; the product page is not a substitute for the framework. The two artefacts are paired: the framework names what to ask for, the product page names what the supplier must answer.

The framework does not assert any certification. No certification claim — ISO 9001, ISO 2859-1, AQL, inspection-level letter, BHMA, ANSI, IEC, UL, CE, FCC, IP rating, ISTA procedure, salt-spray hours, high-temperature dwell hours, low-temperature dwell hours — is made in this article. Every qualification the buyer imposes must be specified by the engineering behaviour the supplier must demonstrate, not by naming a credential.

The framework does not name a price, a MOQ, a lead time, or a warranty year. Commercial terms belong in /request-a-quote. The framework names the inspection discipline the buyer applies against whatever the supplier's commercial terms and product page commit to.

Closing — what to do next

The buyer who runs this framework starts by writing the per-tier timing plan (Pillar 1) against the supplier's production schedule, names the per-defect-class AQL plan (Pillar 2) against the product page, demands the per-SKU inspection-content scope (Pillar 3) against the product page, builds the per-lot document set (Pillar 4) against the supplier's documentation discipline, and writes the per-finding disposition worksheet (Pillar 5) against the supplier's quality agreement. The five artefacts together are the buyer's inspection protocol.

For the per-tier execution — what the inspector sees on the floor at pre-production, at DUPRO, at PSI, and at container loading — read Smart Lock Production Inspection, Smart Lock Container Loading Inspection, and Smart Lock Packaging Inspection.

For the per-SKU sample-test deep dive, read Smart Lock Product Sample Inspection and Smart Lock Factory Quality Control: What Buyers Should Check.

For the contractual envelope the corrective-action commitment sits inside — the supplier quality agreement, the per-shipment record discipline, the broker-and-forwarder handoff — read Smart Lock Pre-Shipment Inspection China and How to Source Smart Locks From China.

For the commercial-terms envelope — what the inspection discipline costs, how the inspection cost is bundled with the unit price, what the supplier commits to in writing, and how the next lot is scheduled — request a quote at Request a quote and request a sample kit at Order samples.

Ostoskorisi
Kokemuksesi tällä sivustolla paranee, kun sallit evästeet. Evästekäytännöt
Chat on WhatsApp