Ranking routes in semiconductor wafer fabs

Area of Science:

• Semiconductor Manufacturing
• Process Engineering
• Data Science

Background:

• Wafer fabrication involves complex processing routes.
• Identifying optimal routes is crucial for efficiency and quality.
• Current methods for route evaluation are limited.

Purpose of the Study:

• To develop a method for estimating and ranking the quality of processing routes in wafer fabrication.
• To provide a practical approach for identifying "best" and "worst" routes for recipe adjustments.
• To support the development of efficient scheduling algorithms in semiconductor plants.

Main Methods:

• A statistical model is used to generate a "local" ranking of individual tools.
• A heuristic procedure combines local rankings to create a "global" route ranking.
• The method utilizes count-based metrics, such as wafer defect counts.

Main Results:

• The proposed method provides an approximate ranking of wafer fabrication routes.
• This ranking is valuable for operational decision-making, even with limited data.
• The approach facilitates the identification of high-performing and low-performing routes.

Conclusions:

• Approximate route rankings are highly beneficial for semiconductor manufacturing operations.
• The developed method offers a practical solution for quality estimation in wafer fabrication.
• This work contributes to improving efficiency and decision-making in complex manufacturing environments.