Acceptance sampling decides whether to accept or reject a whole lot from a random sample of it. Attribute plans count nonconforming units or nonconformities; variable plans measure a characteristic and estimate how much of the lot falls outside its specification. ANSI Z1.4 is the standard for attribute sampling and ANSI Z1.9 for variable sampling. In both, the lot size and the inspection level give a code letter, and the code letter and the Acceptable Quality Level (AQL) give the sample size and the acceptance criteria.
Both standards change the sample as a supplier's results change. Lots start on Normal inspection, move to Tightened after two of five lots are rejected, return to Normal after five accepted in a row on Tightened, and can move to Reduced after ten accepted lots when production is steady and Reduced inspection has been approved. Five lots rejected on Tightened discontinue inspection until the supplier has corrected the cause. c=0 plans, from Squeglia's zero acceptance number tables, accept a lot only when the sample has no nonconformities.
Incoming inspection checks what suppliers deliver, and outgoing inspection checks what you ship. Quality management systems such as ISO 9001 and IATF 16949 expect evidence that product was verified before it is used or released, which is the record an inspection lot leaves behind.
Six problems inspection and supplier-quality teams describe in their own words, and what changes when the sampling standard is built into the workflow.
A plan is set up once per part and inspection type. Every lot then follows the same path from receipt to a final decision.
Step 1 · Masters
Define parts, parties (suppliers, customers and internal factories), the characteristics you check, inspection stations and the inspection teams that work them.
Step 2 · Plan
Create a plan for a part as incoming or outgoing. Add attribute groups on Z1.4, c=0, fixed, percentage or a custom table, and variable groups on Z1.9, then put each characteristic at its station in inspection order.
Step 3 · Activate
Bind the suppliers or customers the plan applies to, with their own switching settings if needed, and name the approvers. A plan needs a group, a party and an approver before it can go live.
Step 4 · Lot
An inspector enters the lot number, quantity, batches and receipt references. The switching level is fixed when the lot is created, and a Discontinued supplier cannot get a new lot.
Step 5 · Inspect
Count nonconformities as totals or unit by unit, enter or paste readings, and record serial numbers where the station asks for them. Each characteristic shows its live status against Ac and Re unless the plan uses blind inspection.
Step 6 · Submit
Zometric derives the sample, judges each characteristic or group and rejects the lot if any group is rejected. Plans can finalise accepted lots, or rejected ones, automatically; the rest wait for an approver.
Step 7 · Approve
The approver accepts or rejects with a reason, returns the lot to the inspector for corrections, or asks for a resubmission, which starts from Tightened. Reports and performance by part, party and plan follow.
Each capability is tied to a decision your team is already trying to make.
ANSI Z1.4 (attribute AQL — Normal, Tightened, Reduced), ANSI Z1.9 (variable AQL, standard deviation method), c=0 (Squeglia), fixed sample plans, percentage plans and your own custom lot-size tables. Pick the regime, the system handles the math.
Incoming checks what suppliers deliver, outgoing checks what you ship — one plan per part and inspection type, with every characteristic assigned to the station where it is checked.
Forward-computed switching between Tightened, Normal and Reduced regimes based on the lot history. Dual decision model when both attribute and variable plans apply.
Every inspection record links back to lot, batch, PO, GRN, Challan and Certificate of Conformance. When the customer asks 'show me the evidence,' it is on the lot record and in its PDF report.
Who accepted, who rejected, with what sample, against what AQL, with what reason. Every change is kept in the record history with who made it and when.
Replace paper and Excel inspection with a browser-based workflow, with station worklists opened from a QR code. The inspector captures once; QA and supplier quality see the same record.
For a lot of 1,200 parts at General inspection level II, ANSI Z1.4 gives code letter J. With an AQL of 1.0, the level the lot was created at decides how many parts are sampled and how many nonconforming parts are allowed. A c=0 plan with the same index AQL moves one column instead and always accepts on zero.
| Level | Z1.4 n | Ac | Re | c=0 n | c=0 AQL |
|---|---|---|---|---|---|
| Normal | Z1.4 n80 | Ac2 | Re3 | c=0 n34 | c=0 AQL1.0 |
| Tightened | Z1.4 n80 | Ac1 | Re2 | c=0 n47 | c=0 AQL0.65 |
| Reduced | Z1.4 n32 | Ac1 | Re3 | c=0 n27 | c=0 AQL1.5 |
Inspection plans and lot results are shared with the modules that plan, investigate and answer for quality.
Released Control Plan rows arrive in an inspection hand-off inbox to be set up, linked to an existing characteristic or declined, and each row shows its floor status back in the Control Plan.
A CAPA proof can point at an inspection characteristic of the part and compare its results before and after the action took effect.
A complaint on a part links straight to that part's inspection plan, or to the lot register when there is none.
Build, version, activate and deactivate inspection plans, bind parties, name approvers and mark a supplier steady and eligible for Reduced inspection, with who and when recorded. They do not record inspections.
Members of the plan's inspection team create lots, record results at their stations, submit lots and raise corrective actions.
Named on the plan, they accept or reject pending lots, return a lot for corrections, request resubmissions and approve the corrective action that lifts a Discontinued supplier.
Resources
Step-by-step guides to configuring the module, running it day to day, and reading what it records.
| Zometric Inspection vs paper check sheets and a spreadsheet | Check sheet and spreadsheet | Zometric Inspection |
|---|---|---|
| Sample size | Looked up in the standard's tables by hand | Derived from lot size, inspection level and AQL, and stored with the lot |
| Switching | Tracked from memory, if at all | Worked out from each supplier's lot history for the part |
| Variables | Mean and standard deviation on a calculator | Z1.9 quality indices and estimated percent nonconforming, with the reason for a rejection |
| Decision | Written on the sheet by the inspector | System decision, then the approver's, with a reason when they differ |
| Traceability | Receipt numbers kept in a separate register | Batches, PO, GRN, challan and CoC reference on the lot record |
| Plan changes | A new sheet printed while old copies stay in use | Versioned plans; sampling changes need a new version |
| History | Crossed-out corrections | Every change kept with who made it and when |
| Reports | Typed up afterwards | Final Inspection Report and Report for Approver as PDF |
Plans, groups, characteristics, party bindings, approvers and notification subscribers are set up in flat sheets pasted from Excel, with a check column that shows each row's errors before saving.
Measurements and unit-by-unit results can be pasted into the lot entry grid from a spreadsheet.
The Final Inspection Report, the Report for Approver and the approver's decision preview, each named by part, lot and date.
Performance tables by part, party and plan download as CSV, keeping the filters applied on screen.
Teams or users subscribe, per plan or for all plans, to events such as lots pending approval, lot decisions and corrective actions.
Straight answers to the questions teams ask before a demo.
Attribute groups use ANSI Z1.4 at inspection levels I to III and S1 to S4, c=0 (Squeglia) zero acceptance number plans, fixed sample plans, percentage plans or a custom lot-size table that you define. Variable groups use ANSI Z1.9 with the standard deviation method. Attribute groups in the same plan can use different standards, while a plan built on a custom table uses it for every group and has no switching.
When a lot is created, Zometric reads the supplier's earlier lots for the same part and inspection type, across every plan version. Two rejections in the last five lots on Normal suggest Tightened, five accepted lots in a row on Tightened return to Normal, and ten accepted lots on Normal suggest Reduced if the supplier has been marked steady and Reduced-eligible. The level is fixed on the lot when it is created.
For ANSI Z1.9 it computes the sample mean and standard deviation, the upper and lower quality indices, and the estimated percent nonconforming from the standard's table. The characteristic is accepted when that estimate does not exceed the maximum allowed by the standard, and rejected automatically when the mean is outside a specification limit.
Yes. The approver records their own decision next to the system's, and a reason is required whenever the two differ. Accepting a lot the system rejected is stored as an approval under deviation. A plan can also finalise accepted lots, or rejected lots, without an approver.
After five lots are rejected on Tightened inspection, Zometric suggests Discontinued and blocks new original lots for that supplier and part. Once an approver approves a corrective action, the next lot starts on Tightened. A resubmission of a rejected lot is suggested Tightened and can never run on Reduced.
Yes. A plan can switch on blind inspection, so the acceptance and rejection numbers are not sent to the inspector's page and no live accept or reject status is shown. Approvers and reports always show them.
Sample-defining settings, such as standards, AQLs and specification limits, are locked once a plan is live, so changing them creates a new plan version. When it is activated, open lots with nothing recorded move to the new version, lots already started finish on the old one, and suppliers keep their switching history.
Yes. Plans, groups, characteristics, party bindings, approvers and notification subscribers can be pasted from Excel into flat sheets keyed by part number and inspection type. A check column validates every row before it is saved. A guided step-by-step builder is also available.
Zometric's modules share a single data backbone — adopting more of them compounds the signal-to-noise advantage.
Catch process deviations before they become defects.
Every gauge calibrated on time, signed off by the right people and proven by MSA.
More than 230 statistical tools in the browser, with guided assistants for choosing one and an AI you can ask about the results.
Extract data from Excel, PDF, CSV or JSON, match against templates, push to Zometric API.
A 30-minute discovery call. A 2–3 week pilot on real factory data. Then scale at your pace.
