What Should I Check Before Smart Lock Shipment?
The direct answer: before a smart lock shipment leaves the factory's hands, a buyer should clear five gates in order — production-complete notice, line-end sampling checks, packing list and carton-mark verification, documentation assembly, and release to carrier. Each gate needs three things written down before you start it: an entry test (what has to be true to begin), an exit test (what proves you may stop), and one owner (a single person who signs or blocks). Nothing on this list is exotic; the reason shipments go wrong is that these checks exist as opinions instead of as signed gates. This article puts the whole pre-shipment sequence in the buyer's seat and gives you the timeline, the tables and the sign-off worksheet to run it yourself.
One boundary first, because this is gate five of your purchasing life, not gate one. The pre-order checklist — doors, access methods, platform binding, quantities — lives in What Should I Check Before Ordering Smart Locks?. If your purchase order is already cut and the factory floor has the work, this timeline is the one that owns the weeks that follow.
Where This Timeline Ends and the Neighbouring Articles Begin
Four other guides on this blog cover shipment territory, and it matters which slice each one owns, because a buyer who reads all four without a map gets the same advice in four fonts. Here is the division, stated plainly:
| Thread | What it owns | Where it sits relative to this article |
|---|---|---|
| Pre-shipment inspection China (PSI) | The hands-on unit testing event itself: sampling size, per-unit protocols, defect dispositions | Deep-dive inside Gate 1 territory; this timeline tells you when the gate opens and closes, that article tells you how to test |
| Container loading inspection | The loading day: cartons into pallets into container, humidity, seal, load plan | Downstream of Gate 4 release — the cargo-moving discipline that begins once paperwork allows booking |
| Shipment planning | Mode choice, volume math, urgency mapping, incoterm cost architecture | Runs in parallel, mostly before Gate 4; it decides how the cargo travels, not whether it is ready |
| This article | The buyer-side sign-off sequence from production-complete to release-to-carrier | The connective timeline that turns loose checks into gates with owners and exit criteria |
If you only remember one sentence from this section: the other articles are disciplines, this one is the schedule. A loading inspection you perform on boxes that were never reconciled against the packing list is theatre. A document packet assembled before the sampling checks pass is paperwork about units that may not survive. Sequence is the product here.
The Gate Map at a Glance
Five gates, one arrow between each, and a hold-loop back from every failure. Owners are stated as roles, not names, so you can pour your own org chart into them — on small teams, two gates may share a person, and that is fine as long as the signatures stay separate.
flowchart LR
G0["Gate 0 — Production-complete notice<br/>Owner: factory planning + your order tracker"] --> G1["Gate 1 — Line-end sampling checks<br/>Owner: factory QC against your written criteria"]
G1 -- pass --> G2["Gate 2 — Packing list & carton marks<br/>Owner: packing supervisor + your docs desk"]
G1 -- fail: HOLD --> RW["Rework / replace units<br/>re-enter Gate 1"]
RW --> G1
G2 -- reconciled --> G3["Gate 3 — Documentation packet<br/>Owner: export desk + your customs broker"]
G2 -- mismatch: HOLD --> G1
G3 -- complete --> G4["Gate 4 — Release to carrier<br/>Owner: YOU sign; forwarder books"]
G3 -- missing paper: HOLD --> G3
G4 --> OUT["Handoff to loading inspection<br/>and freight lifecycle"]
And the compact reference table for the whole map:
| Gate | Trigger to open | Entry test | Exit test (sign to pass) | Primary owner | Buyer's role |
|---|---|---|---|---|---|
| 0 — Production complete | Factory declares the build finished | All PO line items report built and function-tested on the line | Written production-complete report listing model, quantity, batch IDs per line | Factory planning | Receive, file, confirm it matches your PO line-for-line |
| 1 — Sampling checks | Gate 0 signed | Test criteria sheet agreed in writing before testing starts | Sampled units meet the per-model spec sheet; defect rate within the tolerance you set | Factory QC, against buyer-written criteria | Define criteria, review results, accept or HOLD the lot |
| 2 — Packing list & carton marks | Gate 1 signed | Units in approved cartons, marks printed, packing list drafted | Packing-list arithmetic ties to PO quantities and carton marks exactly, per model | Packing supervisor, verified by your docs desk | Reconcile totals, spot-check photos of marks |
| 3 — Documentation | Gate 2 signed | Document checklist locked before assembly begins | Every required paper exists, matches the packing list, and reads cleanly | Export desk / supplier admin | Verify content, route to broker, chase gaps |
| 4 — Release to carrier | Gate 3 signed | Booking confirmed, incoterm duties settled | Your written release note exists; nothing moves without it | You, jointly with the forwarder | Sign. This gate's owner cannot be delegated |
Notice what the map does not contain: vibes. "The factory says it looks good" is not an exit test at any gate. Each gate closes with a document — a report, a signed criteria sheet, a reconciled list, a complete packet, a release note — because a shipment problem you cannot date and attribute is a shipment problem you will re-buy.
Gate 0 — Production Complete: What the Notice Must Actually Contain
Production "finished" is a claim until a report says which models, which quantities, which batch identifiers, and on what date the line-level function tests ran. The entry test for Gate 0 is trivially met — the factory opens its mouth — but the exit test is on your side of the desk: the written production-complete report must match your purchase order line-for-line.
Three failure modes show up at this gate more often than anywhere else:
- The silent substitution. A line item shipped under a revised model code because "the platform is basically the same." Different credential formats, different battery configuration, different mounting hardware. Your doors do not care that the platforms are basically the same.
- The partial declaration. Three of five lines finished, the notice reads as if everything is. Sampling starts on finished lines while the unfinished line quietly loses its place in the queue — and its place in your booking window.
- The batch blind spot. No batch or serial ranges recorded. When Gate 1 finds a defect, you cannot tell whether it is one bad carton or one bad production run, and the disposition decision becomes a coin flip instead of an arithmetic exercise.
The buyer-side check at Gate 0 takes twenty minutes and prevents all three: open your PO, open the report, put a tick next to every line where model code, quantity and batch range match, and write one sentence for every line where they do not. That sentence goes back to the factory the same day. Do not let Gate 1 begin on a lot whose Gate 0 report you have not ticked; sampling a quantity that was never actually built is how inspection reports become fiction.
Gate 1 — Line-End Sampling Checks: Per-Model, Spec-Sheet-First
Gate 1 is where a shipment is either earned or lost, and this timeline only owns its scheduling; the deep testing methodology belongs to the dedicated pre-shipment inspection guide at Smart Lock Pre-Shipment Inspection China. What Gate 1 demands from the buyer's seat is a different question: what are you testing against, and did you write it down before anyone picked up a sample?
The answer must be a per-model criteria sheet, assembled from each model's own product page — not from a category average, not from the sister model that "shares the hardware." Take SLK-9936S, because its spec block is a good example of how specific a line-end check can be when you use the first-party data properly. For that model, the published page states: door thickness fit 36–100 mm, RFID Mifare 1 at 13.56 MHz, read distance 0–5 cm, read time under 0.25 seconds, power from 4 × AA cells, battery operating cycles above 15,000, audit and event records up to 1,000, a 5-pin cylinder with a 3-point latch, stainless steel and alloy body with multiple finish options, mechanical emergency access, and multilingual management software in standalone or centralized-database modes. Every one of those lines becomes a verification point on a sampled unit of that model: present a card at the stated distance and time, cycle the latch, pull the event records out of the log, fit the jig to a within-range door section, confirm the emergency mechanical access actually actuates.
Now the discipline that makes this a firewall instead of a wish list: none of that block applies to any other model. A sampled SLK-9905A gets checked against SLK-9905A's own page. If a model's page does not state a number, your criteria sheet does not invent one — it carries the question to the factory instead, in writing, before testing starts. The same rule prunes nonsense checks: nothing in this catalog's published data supports verifying Matter, Thread, Zigbee, Z-Wave, Apple Home Key, ultra-wideband or vein-pattern features, and Wi-Fi presence is only ever a per-model product-page confirmation, never an assumption to test for.
The gate's exit test has two halves. Half one: sampled units meet the written criteria sheet. Half two — the one everyone skips: the defect findings are quantified per model and per batch, so the disposition at the end of this section is data, not debate. A criteria sheet without batch tagging cannot separate "eight units were mishandled" from "eight thousand share a solder problem," and those two findings have opposite consequences for your shipment date.
Gate 2 — Packing List and Carton Marks: The Arithmetic Before the Boxes Move
A shipment can pass every functional test and still fail customs, still arrive short, and still make your warehouse unusable — all for problems that live in this gate. Gate 2 checks two paper-ish objects against one physical reality: the packing list (the arithmetic) and the carton marks (the labels the receiving team will trust for the next decade of inventory work).
Work it as three reconciliations, in order:
Reconciliation one — packing list to purchase order. Every model line's quantity must tie exactly, including how spares and accessories appear. If mounting templates, strike plates, cards or keys ship alongside locks, they need their own lines or an explicit statement of inclusion; "assorted hardware inside" is how a distributor discovers six weeks later that the property's installer expects 200 cards and the cartons hold 100.
Reconciliation two — carton marks to packing list. Marks should carry, at minimum, the model designation as ordered, quantity per carton, gross/net weight consistent with what is measured, a carton number against a total ("CW 14 of 62"), batch or serial range, and destination reference. Spot-check by photographing a handful of cartons across the stack and reading the marks back against the list. One mis-marked carton is a fix; a systematically mis-marked batch means the packing process itself is the defect.
Reconciliation three — carton content discipline. No mixed-model cartons unless the packing list says so carton by carton. Receiving docks reconcile by the mark on the box, and a mixed carton is an undiscovered short shipment forever.
The battery note that people get wrong: shipping-paperwork treatment of cells depends on cell chemistry per model. For an AA-powered unit like SLK-9936S — four alkaline AA cells, per its own page — the configuration is not the same problem as a model shipping with rechargeable packs installed, and cell chemistry is a per-model confirmation you make before the carrier books, not in a port-side email thread. Whatever each model uses, the packing list and the marks should agree about whether cells are installed, packaged separately, or excluded.
| Check | Question that closes Gate 2 | Evidence to file |
|---|---|---|
| Packing list ↔ PO | Do per-model quantities tie exactly, spares and accessories included or explicitly excluded? | Annotated list with ticks per line |
| Marks ↔ list | Do five photographed cartons per model read identically to the list? | Photo set, filed |
| Contents ↔ marks | One model per carton, or declared otherwise? | Packer's statement or corrected repack |
| Cell status | Is declared battery configuration per model, and agreed with the carrier's rules? | Written confirmation per model |
| Weights | Do listed gross/net weights match a scale reading on at least two cartons per model? | Scale ticket photos |
Gate 2's exit test is a reconciled set, not a factory assertion. Do not outsource the arithmetic — a packing list checked only by the party that wrote it is a packing list with no control at all. The loading-day discipline that follows this — pallets, container hygiene, humidity, seal numbers — is owned by Smart Lock Container Loading Inspection; this gate ends when the numbers inside the warehouse are true.
Gate 3 — Documentation: Locked Before Assembly, Not Improvised in the Port
The document packet is where the term "pre-shipment" earns its prefix. The planning-side article at Smart Lock Shipment Planning covers what your destination's customs regime demands as part of mode and incoterm decisions; this gate is the buyer-side verification that the papers exist, match the cargo, and match each other.
Lock the checklist before assembly starts — a per-destination, per-product list you and your customs broker agree in writing. The generic skeleton most international buyers recognise:
- Commercial invoice consistent with the PO and the packing list.
- Packing list, the same reconciled artifact from Gate 2, not a regenerated version with drifting numbers.
- Transport document instructions ready for the carrier — bill of lading or air waybill draft details: shipper, consignee, notify party, marks exactly as printed.
- Per-model supporting sheets: what the goods are, HS classification as advised by your broker, country of origin stated on the goods or packaging per your import rules.
- Any test reports, conformity documents or product papers your market requires — requested and received as documents to verify, and checked against the actual model and batch numbers being shipped. A report that references a different model revision than what is in the cartons is worse than no report, because it will be believed at the worst possible moment.
The verification pass that catches real problems is a cross-document consistency read, not a completeness tick. Take one number — say, the total quantity of hotel locks for the 9936S line — and follow it through invoice, packing list, bill-of-lading draft and certificate-of-origin draft. Four documents, one number, and the mismatch is the finding. Do that for quantity, model codes and marks, and you have run Gate 3 honestly.
A word on timing traps: documents lag cargo by a habit. If your carrier needs paper forty-eight hours before the vessel and you treat Gate 3 as something to finish while the truck loads, every late amendment becomes a amendment fee, a rolled booking, or a container sitting at the port accruing days. Gate 3 exists precisely so the release at Gate 4 has nothing left to wait for.
Gate 4 — Release to Carrier: The Only Signature You Cannot Delegate
Gates 0 through 3 can be run by capable people on your side and honest paper on the factory side. Gate 4 is different: it is the moment the buyer decides the shipment may leave. Booking confirmations, container appointments and truck schedules are logistics inputs — the release itself is your written statement that Gates 0–3 closed with evidence, and nothing moves until that statement exists.
The release note should be short and dumb-simple: PO reference, the five gate documents it relies on, date, name, and the words "release authorised." Its value is not ceremony. In the inevitable later dispute — a shortage, a customs query, a warranty-adjacent claim about whether a batch was ever modified after test — a dated release note reconstructs what you knew and when. Buyers who say "the factory just shipped it when the truck came" have, at that moment, no gate timeline, no evidence chain, and no position.
Two settlement items belong at Gate 4 because they are incoterm duties rather than paperwork: who books the carrier and who pays whom to move cargo from the factory floor to the vessel or the freight station, confirmed against the term you traded on. Mode urgency and cost architecture were decided in planning; here you only confirm the decision is executed.
Then hand off cleanly. From the release, the cargo's life belongs to loading inspection and the freight-lifecycle guides; your job shifts to receiving. The worksheet at the end of this article is the artifact that makes all of the above one page instead of five intentions.
Who Owns Which Gate
Role clarity beats enthusiasm. Here is the ownership grid in RACI form, with the buyer's org in the left column and the factory's in the second — the single most common cause of a shipment released on a failed gate is a role both sides assumed the other held.
| Gate | Accountable (signs) | Responsible (does) | Consulted | Inputs from the other side |
|---|---|---|---|---|
| 0 — Production complete | Buyer's order tracker | Factory planning | Factory QC lead | PO line table, batch scheme |
| 1 — Sampling checks | Buyer (accept/HOLD) | Factory QC to buyer's written criteria | Your inspection contact | Criteria sheet, defect tolerance, batch tags |
| 2 — Packing list & marks | Buyer's docs desk | Packing supervisor | Warehouse/receiving lead | Mark artwork spec, carton policy |
| 3 — Documentation | Buyer (content verified) | Export desk / supplier admin | Your customs broker | Locked document checklist, HS advice |
| 4 — Release to carrier | You, personally | Forwarder books; factory stages | Factory planning | Booking confirmation, incoterm duty map |
The pattern worth internalising: on every gate, "responsible" sits with whoever physically touches the work, and "accountable" sits with whoever pays for the mistake. Those are almost always different people, and shipments leak exactly through that gap.
When a Gate Fails: The Disposition Matrix
A failed gate is not an emergency if you already decided what failing means. Decide it now, in table form, and paste it into your project file.
| Finding | Severity | Decision rule | Who decides | Clock |
|---|---|---|---|---|
| One sampled unit fails a functional check | Low | Isolate, re-test per criteria, note batch; if isolated, proceed | Factory QC with buyer review | Same day |
| Defect rate exceeds your written tolerance | High | HOLD the line; rework or replace; Gate 1 restarts for that line | Buyer | Before Gate 2 opens |
| Packing list total disagrees with PO | High | HOLD; recount before any box is sealed against the list | Buyer's docs desk | Before Gate 3 |
| Carton marks contradict the list | Medium | Repack/relabel affected cartons; re-verify full batch of marks | Packing supervisor | Before Gate 3 |
| A required document cannot be produced | High | HOLD release; escalate to broker: alternative paper, restatement, or destination problem | Buyer | Before Gate 4 |
| Quantity/model mismatch discovered after booking | High | Amendment with carrier before sailing; do not "fix it at destination" | Buyer + forwarder | Hours, not days |
The reason to hold at all is that every defect is cheaper on the factory floor than in your warehouse. A unit caught at Gate 1 is a bench rework. The same defect caught at your receiving dock is freight claims, replacement logistics, and an installer standing at a hotel door with no lock.
Illustrative Scenario: A 180-Room Hotel Order, Two SKUs
Consider a fictional importer — call the project the Marrow Bay Hotel order to keep names straight; it describes no real customer — with a purchase order of 180 hotel locks for a 180-room property, split as 160 units of SLK-9936S for guest rooms and 20 units of a second hotel model for back-of-house, plus a small batch of spares. The booking is already held for a weekly service, and the production-complete report arrives on a Tuesday.
Gate 0, Tuesday: the report matches the main lines but lists the spares with no batch range. One email, one corrected report Wednesday morning. Small, but if uncorrected, a later Gate 1 finding could never have been attributed.
Gate 1, Wed–Fri: the criteria sheet for the 9936S line quotes its own model's published numbers — read distance and read time at 13.56 MHz with Mifare 1 cards, 4 × AA power, event-record depth, latch and cylinder behavior — and the second SKU gets its own sheet from its own page, no cross-contamination. On the second model, three sampled units show inconsistent emergency-access feel. Under the disposition table this is a HOLD on that line only; the 160-unit line proceeds to packing because its batch-level data is clean. The reworked line re-enters sampling the following Monday.
Gate 2: the packing list arrives showing total quantities right but "assorted hardware" unspecified; the docs desk forces per-line statements for cards and templates before any carton count is trusted, photographs twelve marks across three stack positions, and catches one carton marked with the guest-room model code that packs back-of-house units. Repack. The arithmetic now ties.
Gate 3: the broker's checklist requires per-model supporting papers; the draft origin certificate quietly carries the old revision code of the second model. The quantity trail — 160 and 20 followed across four documents — is checked line by line. One correction cycle, documents finalised Thursday.
Gate 4: Friday release note, signed after all four gate artifacts are filed. The container that sails is one the buyer can describe from memory, gate by gate.
What actually went wrong in this week? Three things — a missing batch field, an unspecified hardware line, a stale revision code — and none of them is a lock failure. All three are timeline failures. The property's opening date never learned they happened, which is the entire point.
Failure Modes of a Sign-Off That Looks Complete
Experienced importers do not skip gates; they weaken them, which is worse because the paperwork still exists. Watch for these five:
Test criteria written after the fact. "We'll compare against the sample" with no sample-approved spec sheet means the benchmark is a memory. Write the per-model sheet before testing.
The verbal pass. A gate cleared by an email that says "all good" carries no numbers. Exit tests produce a document or they did not happen.
Owner drift. Gate 2 signed by whoever was online. The RACI table above exists so the signature has a role attached; when the usual person is unavailable, the gate waits or a named delegate signs. Silence never signs.
The back-office merge. Letting the factory produce the packing list, the marks and the invoice, then signing all three together, converts five checks into one check performed by the counterparty of your shipment. Cross-document consistency only works when you do it.
Optimistic sequencing. Starting Gate 2 while Gate 1 tests are "probably fine" saves hours and destroys evidence: units packed around an unresolved finding are units you have formally accepted while saying you have not.
The Tear-Off Sign-Off Worksheet
Print this, fill it per shipment, file it per PO. It is the whole article compressed into five rows.
| Gate | Entry satisfied? (PO/line match, criteria sheet locked, checklist locked, duties settled) | Exit document produced? | Anomalies noted | Owner signature + date |
|---|---|---|---|---|
| 0 | ☐ Production-complete report matches PO line-for-line, batch ranges listed | ☐ Report filed | ||
| 1 | ☐ Per-model criteria sheet agreed in writing before testing | ☐ Sampled results per model and batch; defects quantified; tolerance held | ||
| 2 | ☐ List drafted, marks printed | ☐ Three reconciliations tied: list↔PO, marks↔list, contents↔marks; cell status confirmed per model | ||
| 3 | ☐ Locked document checklist from broker in hand | ☐ Packet complete; quantity/model/marks followed across every document; drafts match cargo | ||
| 4 | ☐ Booking confirmed, incoterm duties settled | ☐ Release note signed and filed; handoff date logged |
Run it once honestly and the second shipment already feels different, because you stop asking "did they check it?" and start asking "which gate closed, against what exit test, signed by whom?"
Handoff — and What You Check Next
The five gates do not end the shipment; they end the part of the shipment you can still control by refusing to sign. After release, your attention moves to loading-day verification and the freight lifecycle — covered by the container-loading guide and the shipment-planning guide linked above — and finally to receiving, where every gate you skipped arrives as a pallet you cannot explain.
If you are mapping this timeline onto your first order with a supplier, keep the artifact habit from day one: a written production-complete report, a per-model criteria sheet you actually authored, reconciled packing arithmetic, a locked document checklist, and one signed release line with your name on it. The factory-side buyer-resources page at Buyer resources collects the templates and confirmations this sequence leans on, and specific quantities, configurations and commercial terms for your project resolve through Request a quote so that the numbers in your gate documents are the numbers in your PO.
Shipment quality is not discovered at the destination. It is signed, gate by gate, at the origin — and the signature that matters most is the one you withheld until the document existed.

