Production Test Effectiveness of Combined Automated Inspection and ICT Test Strategies

When manufacturers implement automated inspection systems that can detect subtle anomalies in the shape of a solder joint, there is a tendency for process indicators to fail inspection and be reported as defects. Data from a test vehicle experiment demonstrates the limitations of inspection machines when programmers subjectively define acceptance criteria. If engineers do not distinguish between defects and process indicators when performing test effectiveness studies, misleading conclusions can result. The classification of process indicators as defects in a test effectiveness study can be used to support an "inspect everything, repair everything" approach. The Production Test Effectiveness (PTE) metric is indicative of the economic advantage of a test strategy. Case study data on high complexity double-sided boards with real production failures demonstrates that the Test Effectiveness (TE) of two different test strategies can be identical at 97% while their PTEs are 97% and 32%; this represents a three-fold difference in the total cost of ownership.