The Film Ran Fine. The Seals Did Not.

The film that ran fine

A Midwest protein processor was qualifying a new sealing film. Cheaper stock, the same supplier conversation every plant has when resin moves. On the call the question that mattered came out plainly: does it physically run, and does it hold the quality. Two questions, not one. The film ran. Similar line speed, similar downtime, the operators were comfortable. By the first test the change looked like a win.

Then the second question. Leakers. The only way to know whether the new film actually sealed was a leak-test panel against the current protocol, and that work had not been scoped the way the run test had. The run was easy to see on the floor inside an hour. The seal integrity needed a baseline, a sample set, and a disposition path, and it was sitting a step behind. The plant was one signature away from rolling a film wide on the strength of half a qualification.

Two gates, and most plants only test one

Every material or spec change clears two independent gates. The run gate asks: does it feed, seal, and index at line speed with downtime no worse than the incumbent. The quality gate asks: does the output hold the spec, leakers, seal strength, shelf life, transit survival. They are independent. A film can run beautifully and seal poorly. A board can fold clean and fail the transit test. Passing the run gate predicts nothing about the quality gate, and the run gate is the one you can see, so it is the one that gets tested.

Here is the part that makes it expensive. A leaker is not a reject you toss and forget. It is a hold-and-rework cycle, and rework runs on the same line as first-pass production. The defect does not just cost you the pack; it costs you the minutes to find it, hold it, disposition it, and run it again. Disposition latency, not the defect rate, is usually the real constraint. A plant can carry a low leaker rate for years because disposition clears inside a shift. Raise the rate one point with a half-qualified film and the holds stack faster than quality can clear them, WIP spikes, and the downstream line that "gained capacity" on cheaper film is now starved behind a quality hold. The savings were modeled on the run. The losses live on the seal.

Qualify against the production condition, not a sample

The same failure shows up far from a meat line. An outdoor-furniture brand was packing out a new shipper and could not get the product to stop moving inside the box. The packaging engineer found the easy fix in minutes, reposition two blocks, but the real callout was upstream of the box: they were validating the pack against a sample while the production parts were still on order. The production unit should have been a stronger board grade than the sample in their hands. They were proving the design against the wrong artifact, then planning to run it on a different one. No transit test against the production-condition pack had been done. The qualification they trusted did not describe the thing they were going to ship.

So the application is narrow, and it is the same in both plants. Write the qual with two gates, and do not let the visible one stand in for the invisible one. For the run gate: line speed and downtime within a stated band of the incumbent, measured on the floor. For the quality gate: a leak-test or transit panel run at production speed, on production-grade stock, against the current protocol's pass rate, with a disposition path defined before the panel runs. Both gates signed before the change goes wide. The film that runs fine gets no credit until the seal panel clears. The pack-out that looks tight gets no credit until the production-grade box survives the transit test. If the qual file has a run note and no quality panel, the change is not qualified; it is half-qualified, and the other half is going to bill you downstream.

What a clean qualification reads like

The qual file names both gates and a number for each. Line speed and downtime sit within 5 percent of the incumbent, measured on the production line, not a pilot. The quality panel runs at production speed on production-grade material, and the leaker or transit-failure rate comes in at or below the incumbent before anyone signs. Disposition for any fail is defined ahead of the panel, and it clears inside a shift. Quality and operations both sign, because one owns each gate and neither gate stands in for the other. When a change goes wide on that file, the floor sees the same leaker rate next month that it saw last month.

The cheaper film usually is cheaper. The savings on the run are real. So are the leakers on the seal, and nobody put the two in the same model.