Name: AI Quality Control Lab
Author: MySimulator

Inspection policy

Defect-flag sensitivity: 65%

This case study models a 3,000-unit production run inspected by a computer-vision defect scanner. Higher sensitivity scraps more units for rework, catching more true defects, but also scraps good units that were never actually defective, cutting into yield.

Run outcome

Units scrapped/reworked—

True defects caught—

Defects that escaped to customer—

Units by defect score (orange = scrapped, red ring = true defect)