Why Do Reorder Batches, Machines and Staff Drift?

Drift is not the same as a mistake. A mistake is a unit that was made wrong and can be pointed at; drift is a run that was made differently and cannot. Mistakes are found by inspection. Drift is found by comparison, and it is found late, because every unit in the run is equally affected.

That is why drift costs more than defects. A defect is a percentage of a lot and is handled by the sampling plan. Drift is the whole lot, and by the time it is visible the goods are usually on a ship.

The word matters because it sets the response. Teams that call drift a defect go looking for bad units and find none. Teams that call it drift go looking for a movement between two lots, which is where the evidence actually is and where the correction has to be applied.

What Does Drift Mean in a Reorder Programme?

Drift is movement in a measured property between one run and the next, small enough to pass inspection and large enough to be noticed by a customer. It accumulates: each run moves a little, each movement is inside the band, and the total across several runs is outside it.

The definition matters because it tells you where to look. Drift is not found in the inspection report, which only says whether this lot passed. It is found in the comparison sheet, which says what this lot measured against the approved run.

One consequence deserves emphasis: because drift is uniform, a bigger inspection sample does not find it. Sampling finds units that differ from their lot. Drift is the lot differing from the previous one, and only a comparison against a retained reference finds that at all.

Why Does a Batch Drift Even When the Recipe Is Right?

Because a recipe is a set of proportions and a batch is a physical event. Water quality, raw-material strength by shipment, the load size in the bath, the temperature profile and the substrate absorbency all vary within tolerance, and each one moves the result a little.

None of them is out of specification. That is the point: a batch can be compliant and still be different from the last one. The only defence is to compare each batch against a retained reference rather than against the recipe that produced it.

There is a second reason batch drift goes unnoticed for so long: the certificate that comes with a purchase describes the grade, not the lot. A reader checking a certificate against the specification will find it compliant, because the question being asked is the wrong one.

Why Does a Machine Drift Between Runs?

Because a machine is set, and a setting survives only until the machine is needed for something else. Feed timing, thread tension, presser-foot pressure, needle size, blade wear and fixture wear all move, and most of them move while the machine is doing other work.

Tooling wear deserves a note of its own because it is the one machine factor that moves in one direction only. A setting can drift up or down; a cutting die or a print screen only wears, so a dimension or a print position that moves steadily in one direction across several runs is usually tooling rather than material.

Why Does a Change of Operator Move the Product?

Because sewing is a manual skill executed at speed. How firmly a panel is guided, how a corner is turned, how much a seam allowance is trimmed and how a webbing tail is folded are habits, and habits differ between people.

The effect is local rather than uniform. An operator who stretches a panel slightly into the foot changes the finished dimension of that panel on every unit they make, which is why a drift that appears on part of a run often traces to one station rather than to a material.

It is worth adding a fourth source that sits between machine and people: tooling. Cutting dies, print screens, welding fixtures and jigs wear with use, and the wear moves a dimension or a print position gradually across a year rather than suddenly between two runs.

Which of the Three Is Easiest to Control?

The machine, by a wide margin. A setting can be checked in a minute, corrected in another, and recorded on a sheet. The people are next: training, a named station and a first-piece check cover most of it. The batch is hardest, because it is bought rather than set.

That ordering is worth knowing because it tells a programme where to spend its effort. First-piece checks are inexpensive and catch two of the three sources; batch control requires a reference, a certificate and a comparison, and it never fully eliminates the difference.

One practical consequence follows from that ordering: the least costly drift control is a first-piece check, because it addresses the two sources that are easiest to correct. A programme that does nothing else should still do that, and it is remarkable how many programmes do nothing else and not that.

There is a third test that is quicker than either: ask whether the difference is present in the first pieces. If it is, the cause was there before the run started - a material, a setting or a station. If it is not, the cause entered during the run, which narrows the list to wear, a lot change or a correction that was applied and forgotten.

How Is Drift Distinguished From a Mistake?

By whether the difference is uniform across the run. A mistake produces a few units that differ from the rest of the lot. Drift produces a lot that differs from the previous lot, uniformly, with nothing inside the run to point at.

The second test is whether a correction fixes it. A setting is corrected and the next units are right, which is a machine drift. A material is changed and the next units are right, which is a batch drift. Neither correction works if the cause was the other one.

There is a fifth document that helps and is rarely kept: the shift record. Where the run is long enough to cross a shift, knowing which station was staffed when the change entered narrows a people-drift investigation from a week to an hour, and it costs one line on a sheet somebody already writes.

What Evidence Tells You Which One Moved?

Four documents, read in order: the first-piece sheet, which says whether the machine was set correctly at the start; the in-process check, which says whether the setting held; the material certificate, which says which batch was used; and the comparison sheet, which says what the run measured.

First-piece sheet

Wrong at the start: a machine or an operator issue

First piece right, run wrong

A setting drifted, a blade wore, or a lot changed

In-process check

Shows whether the change entered partway through the run

Material certificate

Names the batch, which links a change to a purchase

Comparison sheet

Gives the direction and the size of the movement

Previous sheets

Show whether this is a step or a continuing slope

The first-piece sheet is the most valuable of the four and the most often missing. It is the only document that can say whether the run started correctly, and without it every drift investigation begins by guessing.

A useful habit is to write the suspected cause on the comparison sheet before investigating. A guess recorded at the time can be checked against the evidence later, and a programme that does this for a year ends up with a record of how often its own first guess was right.

How Is Each One Bounded in Writing?

Batch: a retained reference, a certificate with a batch number, and a comparison before the lot is cut. Machine: a first-piece check with measurements and two signatures, and a named setting recorded per style. People: a named station per operation and a first-piece check that covers every station.

That repetition is also what makes the programme resilient to a staff change. A new account manager, a new line supervisor and a new buyer can all read the same four records and reach the same conclusion, which is the practical definition of a programme being under control rather than being remembered by one person.

What Does a Drift-Resistant Programme Consist Of?

Boring, and slightly repetitive. A first-piece check on every run, a comparison sheet with readings on every run, a retained unit from every run, and an annual review that compares the current run with the approved one rather than with the previous one.

None of those four is difficult. What makes them rare is that each one is easy to skip on any given run without anything visibly going wrong, which is exactly how drift becomes established.

There is one protection worth naming explicitly, because it is the only one that catches accumulated drift: keep a retained unit from every run. Five units in a cupboard, dated and in order, show a movement that no single comparison sheet can, because the eye reads a series far better than it reads a table.

What Goes Into the Drift Record?

One sheet per run: which batch, which machine, which stations, the first-piece readings, the in-process readings, the final comparison readings, and a note on any correction applied during the run. Seven fields on one sheet.

There is a final point worth making because it decides whether any of this gets done. None of the four habits takes more than an hour per run, and all four are easy to skip without anything visibly going wrong. The programmes that hold them are the ones where someone has written the four lines into the booking routine rather than into a manual.

Frequently Asked Questions

People Also Ask