Multi-stage Process Diagnosis Based on Advanced Morphing

The process control for high accuracy manufacturing needs to diagnose the feature variation from high definition metrology (HDM) data.
The state-of-the-art shape description methods are inadequate for HDM data processing due to their low accuracy.
Morphing-based shape description technologies can effectively capture the feature dissimilarities between and with manufacturing stages from HDM data.
By examining the relationship between the feature variation from morphing and processes, it is possible to diagnose the root cause.

To develop a morphing based approach to diagnose feature dissimilarities within and between manufacturing operations to reduce ramp-up time.

Define morphing functions for part feature variation.
Examine the relationship between part feature variation and manufacturing process.
Develop algorithms for using HDM data to fit the morph function.
Get correlations between morphing function parameters and root causes of manufacturing variation.