Inline vs Endline QC: Where to Spend the Checkers
A defect found at the operator's station costs a re-stitch. The same defect at the end of the line costs unpicking through every operation added since. In the finished carton it costs a repack; at the buyer's DC it costs a chargeback and a reputation. Quality control is one discipline practiced at four prices, and the whole game is moving detection left.
Inline: quality where it is created
Inline QC checks work at the stations, during production: the first pieces after every changeover, then a patrol rhythm through the day. Its output is not a score but a conversation: a defect trend at station six is caught within the hour, traced to a guide, a tension, a method, and stopped at the source. The measure of good inline work is defects that never multiplied.
Two disciplines make inline real rather than ceremonial. First-piece verification after any change, machine, operator, or method, because changes are where defect runs begin. And traffic-light rhythm: a station generating repeats gets flagged and revisited, so patrol attention follows risk instead of a fixed circuit.
Endline: the honest count
Endline checks completed garments, every piece or a heavy sample, and produces the day's truth: defects by type, by operation, by hour. Its numbers feed everything downstream: the hourly production and quality capture that runs the variance card, rework minutes into the loss tree, and the trend that predicts final inspection results before the inspector is booked.
The endline trap is letting it become a repair station: checkers quietly fixing what they find, defects vanishing from the record, and the line learning nothing. Found means recorded, and recorded means attributed to the operation that made it, or the count is theater.
The pincer, not either-or
Factories sometimes staff one and starve the other. The pincer needs both: inline shrinks the defect rate at the source; endline measures what escaped and aims tomorrow's inline attention. A rising endline count for an operation is the instruction to move a checker upstream to it; weeks of clean endline results on a stable style justify lightening the patrol. The two feeds regulate each other, which is exactly why they belong in one system with one defect taxonomy rather than two clipboards that never reconcile.