Semiconductor Process Management System for Real-Time Yield Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Semiconductor manufacturing processes require high accuracy and efficiency, but existing systems lack effective methods to identify and address process deviations and defects in real-time, leading to reduced yield and reliability.
Innovation Solution
A semiconductor process management system that includes process apparatuses, measuring apparatuses, and a management system with a communicator, determination units, and a recipe generator to establish mutual influence models, detect out-of-spec processes, identify defective apparatuses and factors, and generate new process recipes to improve yield and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a great number of unit processes are performed through a series of semiconductor manufacturing processes, then the complexity of the manufacturing process increases, but the process yield decreases due to difficulty in identifying and addressing process deviations
Solution Approach 1:
The patent implements a feedback mechanism where measured values from measuring apparatus are compared against specification ranges, and when deviations are detected, the system automatically identifies defective apparatus and process factors, then generates corrected process recipes that are fed back to the process apparatuses. This closed-loop feedback system enables real-time process correction without manual intervention, maintaining high yield despite process complexity.
Solution Approach 2:
The process management system acts as an intermediary between multiple process apparatuses and measuring apparatuses. It receives process recipes from process apparatuses, receives measured values from measuring apparatus, determines spec-out conditions, identifies defective apparatus and process factors, and generates corrected process recipes. This intermediary system coordinates the complex interactions among numerous unit processes, making the overall system manageable and controllable.
2Manufacturing precision
If multiple process apparatuses perform unit processes according to a planned process sequence, then the manufacturing precision is maintained, but the difficulty of detecting and measuring process deviations increases
Solution Approach 1:
The patent segments the complex manufacturing process into individual unit processes, each with its own process recipe and measurable parameters. The process management system evaluates each unit process separately by comparing measured values against specification ranges, making it easier to detect deviations in specific processes rather than analyzing the entire manufacturing sequence as a whole.
Solution Approach 2:
The patent replaces manual detection and analysis of process deviations with an automated information processing system. The process management system automatically receives measured values, compares them against specifications, determines spec-out conditions, and identifies defective apparatus using algorithmic analysis rather than mechanical or manual inspection methods.
3Reliability
If process deviations are not identified and addressed in real-time, then the ease of operation is maintained, but the reliability of the semiconductor product decreases
Solution Approach 1:
The process management system operates autonomously without requiring continuous manual intervention. It automatically receives process recipes, receives measured values from measuring apparatus, determines whether processes are out of specification, identifies defective apparatus and process factors, and generates corrected process recipes that are automatically transmitted to process apparatuses. This self-service capability maintains reliability while minimizing operational complexity.
Data Source
AI summary
A semiconductor process management system is provided. The semiconductor process management system includes a communicator that receives a process recipe from one or more process apparatuses and receives a measured value for each sampling point from one or more measuring apparatus, and a first determination unit that establishes a mutual influence model between the process recipe and the measured value for each sampling point based on the process recipe and the measured value for each sampling point.


