Inline Quality Inspection and Defects-per-Hundred-Units Analytics in Apparel Manufacturing: A Review and Implementation Model

Quality control in garment factories has traditionally relied on end-of-line inspection and final audits based on acceptance sampling. These methods determine whether a lot can be shipped but do little to prevent defects, because they detect problems after the cost of making them has already been incurred. Inline inspection, in which garments are checked at or near the operations that create defects, offers earlier detection, but its value depends on how the resulting data is recorded and used. This paper reviews the principles of acceptance sampling and statistical quality control as they apply to apparel, defines common quality metrics such as defects per hundred units (DHU) and first-pass yield, and proposes an implementation model in which inline, end-of-line and final inspection data are linked by operation, operator, defect type and lot. We describe how Pareto analysis and control limits can be applied hour by hour and how corrective actions can be tracked to closure. The paper is a review and conceptual model; numerical examples are illustrative. Intelactory Working Paper Series, WP-2026-04. https://intelactory.com