Quality Management
Efficiency metrics in high-volume manufacturing depend on the precise identification of where materials or units are rejected during the assembly sequence. A yield loss allocation enables the engineering team to identify which production steps are responsible for the majority of the scrap. By breaking down the total losses into discrete categories like wafer fabrication, assembly, dicing or packaging, the facility can prioritize its improvement efforts.
This systematic approach ensures that the most costly problems are addressed first and provides a clear roadmap for engineering investments.
Data Attribution
Automated test equipment generates the raw data needed to populate the loss model. In yield loss allocation, the software correlates the location of a failed die on the wafer with the known defect density of the lithography tool. This correlation helps in distinguishing between random particles and systematic equipment malfunctions.
Accurate mapping requires a precise coordinate system that is consistent across all the inspection tools in the factory.
Economic Consequence
Financial impact of each loss category is calculated to determine the return on investment for process changes. Yield loss allocation at the final test stage is generally more expensive than losses occurring early in the wafer fab. This is because the units at the final stage have already absorbed the costs of packaging and assembly.
Verification Loop
Corrective actions are monitored to confirm that the changes in the process lead to a measurable improvement in the output. The effectiveness of yield loss allocation depends on the feedback loop between the test floor and the production management.