Semiconductor Dispatch System Using PCA and Linear Regression

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current real-time dispatch systems in semiconductor manufacturing fail to optimize tool utilization efficiently, leading to idle tools and suboptimal production efficiency due to reliance on manual task allocation based on operator priorities and tool capacities.

Innovation Solution

A dispatch integration system that utilizes Principal Component Analysis (PCA) to sort tool parameters and defines factor weights, combined with a linear regression weight formula to calculate tool weights, determining the most suitable tool for dispatching based on current process conditions, thereby optimizing tool allocation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual task allocation based on operator priorities and tool capacities is used, then ease of operation is maintained, but productivity decreases due to idle tools and suboptimal production efficiency

Engineering Contradiction:

Improveproduction efficiencyVSAvoidmanual task allocation

Core Design Contradiction:

ProductivityVSEase of operation

Solution Approach 1:

The system enables tools to autonomously send idle messages when they become available, and the real-time dispatch system automatically calculates tool weights and determines optimal task allocation without human intervention. This self-service mechanism eliminates manual operator involvement while maximizing productivity through automated, data-driven decisions.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the mechanical/manual dispatch system with an automated computational system that uses principal component analysis and linear regression algorithms. The real-time dispatch system substitutes human operators' decision-making with mathematical models that objectively evaluate tool parameters and determine optimal task assignments, thereby improving productivity without sacrificing operational simplicity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If real-time dispatch system uses automated calculation with PCA and linear regression, then productivity improves through optimized tool allocation, but device complexity increases

Engineering Contradiction:

Improveproduction efficiencyVSAvoidreal-time dispatch system

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent introduces a real-time dispatch system as an intermediary layer between the tool control system and the manufacturing execution system. This intermediary automatically collects idle messages, performs PCA analysis on tool parameters, calculates tool weights using linear regression, and determines optimal task assignments. While this adds computational complexity, it enables sophisticated automated optimization that dramatically improves productivity through data-driven decision-making.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20070129833A1Dispatch integration system and method based on semiconductor manufacturing

Publication Date: 2007.06.07 POWERCHIP SEMICON MFG CORP

AI summary

A dispatch integration method based on semiconductor manufacturing. When multiple tools are idle, idle messages are sent to a real-time dispatch system through a tool control system and a manufacturing execution system. The real-time dispatch system sorts tool parameters relating to tool efficiency based on a currently executed process using a principal component analysis method, defining weights of each tool according to the defined tool parameters and corresponding factor weights using a linear regression weight formula, calculating tool weights of each tool parameter using a linear regression weight formula, and determining a desired tool for dispatch according to the calculated tool weights.