Defect Cause Analysis Using Device Influence Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the manufacturing process, identifying the cause of product defects is challenging due to complex production processes and large volumes of products, leading to timeliness and accuracy issues in data processing that cannot meet increasing demand.
Innovation Solution
A data processing system that calculates influencing parameters of sample production devices based on defect degree and device information, using methods like Gini coefficient and chi-square distribution to determine the impact of each device on defect occurrence, and displays these parameters to facilitate targeted inspections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual inspection methods are used to identify defect causes, then labor costs are high, but inspection efficiency is low
Solution Approach 1:
The patent replaces manual mechanical inspection with an automated data processing system that uses computer algorithms to calculate influencing parameters. The system automatically processes production data, calculates Gini coefficients and chi-square values, and identifies defect causes without human intervention, thereby improving efficiency and reducing time loss.
Solution Approach 2:
The system enables self-service by allowing the data processing system to automatically analyze production data and identify defect causes without requiring manual inspection. The automated calculation of influencing parameters and generation of inspection priorities allows the system to serve itself in identifying problems that previously required human labor.
2Measurement precision
If comprehensive data processing is performed on all production devices, then accuracy is high, but processing time is excessive
Solution Approach 1:
The patent transforms the raw production data into meaningful influencing parameters through mathematical transformations. By calculating Gini coefficients and chi-square values, the system changes the parameter representation to highlight the most significant factors, achieving high accuracy without processing every detail of the raw data comprehensively.
Solution Approach 2:
The system extracts only the most relevant information from the comprehensive production data by calculating influencing parameters that specifically measure the relationship between production devices and defects. This extraction approach maintains accuracy by focusing on key indicators while reducing overall processing time by ignoring less relevant data details.
3Reliability
If traditional defect analysis methods are used, then all devices must be inspected, but the complexity of inspection increases
Solution Approach 1:
The patent applies local quality by assigning different weights and priorities to different production devices based on their calculated influencing parameters. Instead of treating all devices uniformly, the system identifies and focuses inspection resources on specific devices with high influencing parameters, thereby maintaining reliable defect detection while reducing overall inspection complexity.
Solution Approach 2:
The system segments the inspection process by dividing all production devices into priority groups based on their influencing parameters. This segmentation allows the system to handle complex multi-device inspections by breaking them down into manageable priority levels, reducing the perceived complexity while maintaining comprehensive coverage.
4Measurement precision
If detailed analysis of all production parameters is performed, then identification accuracy is high, but data processing complexity increases
Solution Approach 1:
The patent transforms complex raw production parameters into simplified influencing parameters through mathematical calculations. By converting multiple detailed parameters into aggregate metrics like Gini coefficients and chi-square values, the system maintains high identification accuracy while reducing data processing complexity through parameter aggregation and transformation.
Data Source
AI summary
A data processing method includes: obtaining a defect type of a sample set in response to a first input of a user on a first interface, the sample set including samples, each sample having a first parameter used to represent a defect degree of the sample with regard to the defect type and a second parameter used to represent device informations of sample production devices through which the sample passes; calculating yield purity indexes of sample production devices on the samples based on first parameters and second parameters of the samples, so as to obtain influencing parameters of the sample production devices, an influencing parameter of each sample production device being used to represent an influence degree to which the sample production device affects an occurrence of the defect type on the samples; and displaying the influencing parameters of the sample production devices on a second interface.