When Does the Mid-Run Inspection Happen on a 500-Piece Bag Order?

Final inspection is a verdict on a lot that already exists. The mid-run check is a different instrument: it is a measurement taken while the outcome can still be changed, and it is the only quality control in the schedule with that property. Buyers who understand the difference ask for it by name. Buyers who do not discover their systematic error in the final report, at which point the only remaining options are re-work or acceptance.

When Exactly Does the Mid-Run Check Happen?

At a defined share of completion rather than on a fixed day, and the share matters. Too early and the line has not settled, so the tally measures the set-up rather than the run. Too late and there is no run left to correct, so the tally is a verdict dressed as a check. On a 500-piece order with a sewing block of about twelve days, the working answer is when roughly half the units are assembled.

The date follows from that rather than the other way round. On a 40-day schedule the sewing block runs from about day 23 to day 34, which puts the gate at day 31 or 32 and leaves two days of run in which a correction can still be absorbed. On a compressed schedule the same gate sits closer to the end, which is one of the quiet costs of compression: the check still happens, but with less run behind it to fix.

What Does the Check Look At?

Three things, in this order: the dimension-critical features, the repeat operations, and the assembly steps that are expensive to undo. Dimension-critical means anything a customer will measure or a retailer will reject on: an overall dimension, a strap length, a pocket position. Repeat operations means anything done 500 times, because a setting error there is systematic rather than random.

The expensive-to-undo list is the one that justifies the gate. A webbing length that is 4 mm short, a binding that is being set 2 mm narrow, a print sitting 3 mm off registration: each is invisible in one unit, obvious in twenty laid side by side, and each is sewn into every unit produced before somebody counts it.

How Do You Read a Defect Class Back to a Cause?

The class tells you the cause, and the cause tells you the response. Critical defects point at compliance or function, major defects at a setting or a component, and minor defects at an operator or a finish. Reading the tally that way converts a list of counts into a decision, which is the only reason the tally is worth producing.

What Should the Output Contain?

A defect tally by class, with counts, not a pass verdict. You want to know that four units had a skipped stitch, one had a hardware misalignment and two had a print registration outside tolerance, because the class tells you where to look. Critical, major and minor are the classes used at final inspection under ISO 2859-1 and they are the right vocabulary here as well, because the same definitions will be applied six days later.

A tally that says only "acceptable" has discarded the only information the gate exists to produce. The major and minor defect dictionary is the reference we classify against, and holding the mid-run tally and the final report in the same vocabulary is what lets you compare them and see whether a correction worked.

What Happens After the Tally Is Issued?

Three possible outcomes, and the schedule should name which one applies within a stated number of days. If the tally is clean, the run continues and nothing changes. If it shows a correctable systematic error, the line is stopped, the setting is changed, and the units already made are either corrected or segregated. If it shows a component problem, the run stops until the component is resolved.

The second outcome is where most of the value sits and where most programmes fail to act. Segregating the affected units is the discipline that matters: a corrected setting on the machine does nothing for the 150 units already made, and those have to be identified, held and either re-worked or accepted knowingly rather than disappearing into the carton count.

There is a fourth outcome that deserves naming because it is the one buyers least expect: the gate finds nothing and the run still fails final inspection. That happens when the defect is introduced after the gate — in finishing, in packing, or in the carton itself. A gate at half the run protects what happens before it, and the schedule still needs a packing check and a final draw to protect what happens after.

Who Should Read the Tally?

The buyer, not only the programme desk. This is the point in the schedule where the buyer has real leverage and usually does not use it, because a defect tally with unfamiliar vocabulary in it reads as a technical document rather than a commercial one. Read the counts and the classes, and ask one question: is this an operator, a setting or a component.

The answer determines what happens next. An operator problem is a training conversation and is usually self-correcting. A setting problem is a correction plus a segregation decision. A component problem is the serious one, because it means the material or hardware arriving is not the material specified, and the units already made cannot be corrected at all.

One more field belongs in the response rule, and it is the one that protects the buyer commercially. When a correction is made, ask how many units were produced before it and what happened to them. That number is the size of the exposure, and a tally that reports four defects without reporting the population they came from and the disposal of the affected units is a tally that has answered a question nobody asked.

How Does the Mid-Run Gate Interact With Final Inspection?

They use the same vocabulary and different maths. Final inspection is a statistical draw: on a 500-piece lot at General Inspection Level II under ISO 2859-1 the code letter is H, 50 units are drawn, and at AQL 2.5 the accept/reject pair for major defects is 3 and 4. The mid-run gate is not a draw at all; it is a look at every unit available at that moment, or at a defined sample of them, with no accept/reject decision attached.

That difference is deliberate. A random sample of 50 from a run that is producing a systematic error at a rate of 6 per cent will frequently pass, because 50 units is not enough to distinguish 6 per cent from 3 per cent reliably. The gate exists to catch what the draw is not powerful enough to catch, which is why the two are complementary rather than redundant.

What Does the Gate Cost and What Does It Return?

One to two days on the schedule and a few hours of inspection time, and it is the highest-return control in the whole programme. The comparison is not against doing nothing; it is against discovering the same error at final inspection, where the options are a full re-work across 500 units, a waiver the buyer signs, or a shipment in which the customer finds the defect first.

There is a second return that is harder to quantify and often larger. A programme with a visible mid-run gate produces fewer defects at final inspection, because a line that knows it will be counted at the halfway point behaves differently from one that will only be counted at the end. That effect is real, it is the reason in-line gates exist, and it costs nothing beyond the tally.

What If the Buyer Cannot Attend the Gate?

Ask for the tally and for photographs that prove something. A photograph of a finished bag proves nothing; a photograph of a measured dimension, of twenty units in a row, or of a segregated hold tray proves that the gate happened. The three images worth requesting are the dimension record, the row of units, and the hold tray with its count.

Ask also for the units inspected rather than only the units found defective. Four skipped stitches out of 200 is a different article from four out of 40, and a tally without the population is a list rather than a rate. That one field is what lets the buyer compare this run against the last one and tell whether the line is actually improving.

The QC documentation and reports post sets out what a usable report contains, and the same field list applies here: a date, a quantity inspected, a classification vocabulary, counts by class, and a name. A mid-run tally without those five fields is a message rather than a record.

There is one further use for the gate that buyers rarely exploit: it is the natural moment to confirm the packing specification. Carton marks, polybag requirements, barcode placement and the carton count are all settled by the packing stage, and a mistake in any of them is discovered at the receiving warehouse rather than at inspection. Checking a packed sample at the gate costs an hour and prevents the most annoying category of rejection.

How Should the Gate Be Written Into the Schedule?

Trigger

A defined share of completion, usually 40% to 60% of the run

Method

Every unit available at that point, or a defined sample of them

Vocabulary

Critical, major and minor, matching the final inspection classification

Output

Counts by class, issued within a stated number of days

Response

Continue, correct and segregate, or stop for a component problem

Escalation

A component problem goes to the buyer the same day

Evidence

Dimension record, row photograph, hold tray count

One caveat on the gate is worth stating plainly, because it is the difference between a control and a ritual. A gate that is always passed, on every order, regardless of what is on the bench, is not a gate. If a programme has run twenty orders and the mid-run tally has never once found anything, the tally is being filled in rather than filled out, and the buyer should ask to see the counts behind the last three.

Seven fields, and the gate stops being a favour somebody does and becomes a row in the schedule with a date and a consequence. We do not promise the fastest 20 days. We promise a delivery date you can calculate. The mid-run gate is one of the rows that makes the calculation worth doing.

Frequently Asked Questions

People Also Ask