# US 2021/0364999 A1 · System and method for analyzing cause of product defect, computer readable medium

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0364999 A1
- Document ID: US20210364999A1
- Status: Published application
- Kind: A1
- Application number: 16/972,402
- Filed: 2019-11-29
- Published: 2021-11-25
- Earliest claimed: 2019-11-29
- Assignees: Boe Technology Group Co., Ltd.
- Applicants: Boe Technology Group Co., Ltd.
- Inventors: Tian LAN; Hong WANG; Dong CHAI; Haohan WU; Guoliang SHEN; Weihe LIU
- Web page: https://patentgrep.com/patent/US20210364999A1

## CPC

- G05B 19/406: characterised by monitoring or safety (G05B19/19 takes precedence)
- G05B 19/41875: {characterised by quality surveillance of production}
- G05B 23/0237: {based on parallel systems, e.g. comparing signals produced at the same time by same type systems and detect faulty ones by noticing differences among their responses}
- G06F 3/147: using display panels
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 20/00: Machine learning
- G06N 20/20: Ensemble learning
- G06Q 10/06395: {Quality analysis or management}
- G06Q 50/04: Manufacturing
- G05B 2219/31314: Store in workpiece detected defects
- G05B 2219/32221: Correlation between defect and measured parameters to find origin of defect
- G05B 2219/32222: Fault, defect detection of origin of fault, defect of product
- G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR

## Abstract

A system for analyzing cause of product defect, including: a distributed storage device configured to store production data generated by a factory device; an analysis device including one or more processors configured to perform: acquiring a production record from the production data; the production record includes information of processing devices used during production procedures for producing the products and information of defects occurring, where each product is processed by multiple processing devices, and each processing device participates in only the production procedures of a portion of the products; determining a correlation weight of the processing device to be analyzed corresponding to a defect to be analyzed according to the production record, and determining a correlation between the processing device to be analyzed and the defect to be analyzed according to the correlation weight; a display device configured to display an analysis result of the analysis device.

## Claims

### Claim 1

A system for analyzing cause of product defect, comprising: a distributed storage device, an analysis device, and a display device, wherein,
the distributed storage device is configured to store production data generated by a factory device;
the analysis device includes one or more processors configured to perform following operations to determine a correlation:
acquiring a production record from the production data stored by the distributed storage device; the production record comprises information of processing devices by which a plurality of products are processed during the production procedures for producing the products and information of defects occurring, wherein each of the products is processed by multiple ones of the processing devices during the production procedure, and each of the processing devices participates in only the production procedures for production a portion of the products;
determining a correlation weight of the processing device to be analyzed in the processing devices corresponding to the defect to be analyzed according to the production record, and determining a correlation between the processing device to be analyzed and the defect to be analyzed according to the correlation weight;
the display device is configured to display an analysis result of the analysis device.

### Claim 2 (depends on 1)

The system of claim 1, wherein the determining the correlation weight of the processing device to be analyzed in the processing devices corresponding to the defect to be analyzed comprises:
for each processing device to be analyzed, taking the products which are processed by the processing device to be analyzed in the production procedure as a first group, and taking the products which are not processed by the processing device to be analyzed in the production procedure as a second group;
for each of the processing device to be analyzed, determining a sub-correlation weight of the first group and a sub-correlation weight of the second group.

### Claim 3 (depends on 2)

The system of claim 2, wherein the sub-correlation weight Qi of the ith group is calculated by a formula:
Q
i
=
ln
⁡
(
G
i
/
G
total
B
i
/
B
total
)
;
wherein i is 1 or 2, Gi is the number of the products with the defect to be analyzed in the ith group, Gtotal is the number of all the products with the defect to be analyzed, B1 is the number of the products without the defect to be analyzed in the ith group, and Btotal is the number of all the products without the defect to be analyzed.

### Claim 4 (depends on 2)

The system of claim 2, wherein the determining the correlation between the processing device to be analyzed and the defect to be analyzed according to the correlation weight comprises:
for each processing device to be analyzed, determining the correlation between the processing device to be analyzed and the defect to be analyzed according to a sum of the sub-correlation weight of the first group and the sub-correlation weight of the second group.

### Claim 5 (depends on 2)

The system of claim 2, wherein the determining the correlation between the processing device to be analyzed and the defect to be analyzed according to the correlation weight comprises:
for each processing device to be analyzed, determining a value parameter according to the sub-correlation weight of the first group and the sub-correlation weight of the second group, and determining the correlation between the processing device to be analyzed and the defect to be analyzed according to the value parameter.

### Claim 6 (depends on 5)

The system of claim 5, wherein the value parameter J is calculated by the formula:
J=ΣJi, and Ji=(DGi−DBi)*Qi,
where Qi is the sub-correlation weight of the ith group, i is 1 or 2, DGi is a proportion of the products with the defect to be analyzed in the ith group to all the products with the defect to be analyzed, and DBi is a proportion of the products without the defect to be analyzed in the ith group to all the products without the defect to be analyzed.

### Claim 7 (depends on 1)

The system of claim 1, wherein the determining the correlation weight of the processing device to be analyzed in the processing devices corresponding to the defect to be analyzed comprises:
determining at least a portion of the processing devices as the processing devices to be analyzed corresponding to the defect to be analyzed according to a preset corresponding relation.

### Claim 8 (depends on 1)

The system of claim 1, wherein the production record further comprises: parameter values of at least a portion of process parameters used by at least a portion of the processing devices, by which the product is processed during the production procedure, for processing the product;
after acquiring the production record from the production data stored in the distributed storage device, a preset machine learning model is configured to perform: determining the correlation between at least the portion of process parameters used by at least the portion of the processing devices and the defect to be analyzed, according to the production record.

### Claim 9 (depends on 8)

The system of claim 8, wherein the machine learning model comprises any one of a random forest model, a gradient lifting tree model and an XGboost model.

### Claim 10 (depends on 8)

The system of claim 8, wherein the determining the correlation between at least the portion of process parameters used by at least the portion of the processing devices and the defect to be analyzed is performed during determining the correlation between the processing device to be analyzed and the defect to be analyzed according to the correlation weight;
the determining the correlation between at least the portion of process parameters used by at least the portion of the processing devices and the defect to be analyzed comprises: determining the processing device to be analyzed, which has a correlation with the defect to be analyzed greater than a first preset value or ranked before a first preset position, as a key processing device to be analyzed, and only determining the correlation between at least the portion of process parameters of the key processing device to be analyzed and the defect to be analyzed.

### Claim 11 (depends on 8)

The system of claim 8, wherein after the determining the correlation between at least the portion of process parameters used by at least the portion of the processing devices and the defect to be analyzed, determining a recommended parameter value range of at least the portion of process parameters according to the correlation between the process parameters and the defect to be analyzed.

### Claim 12 (depends on 11)

The system of claim 11, wherein the determining the recommended parameter value range of at least the portion of process parameters comprises:
determining the process parameter which has a correlation with the defect to be analyzed greater than a second preset value or ranked before a second preset position, as a key process parameter, and only determining the recommended parameter value range of the key process parameter.

### Claim 13 (depends on 1)

The system of claim 1, wherein the product is a display panel.

### Claim 14

A method for analyzing cause of product defect, comprising:
determining a correlation weight of a processing device to be analyzed in a plurality of processing devices corresponding to a defect to be analyzed according to a production record, and determining a correlation between the processing device to be analyzed and the defect to be analyzed according to the correlation weight; the production record comprises information of processing devices by which a plurality of products are processed during production procedures and information of defects occurring, wherein each of the products is processed by multiple ones of the processing devices during the production procedure, and each of the processing devices participates in only the production procedures for producing a portion of the products.

### Claim 15 (depends on 14)

The method of claim 14, wherein the determining the correlation weight of the processing device to be analyzed in the processing devices corresponding to the defect to be analyzed, and determining the correlation between the processing device to be analyzed and the defect to be analyzed according to the correlation weight comprises:
for each processing device to be analyzed, taking the products which are processed by the processing device to be analyzed in the production procedures as a first group, and taking the products which are not processed by the processing device to be analyzed in the production procedures as a second group;
for each of the processing device to be analyzed, determining a sub-correlation weight of the first group and a sub-correlation weight of the second group;
for each processing device to be analyzed, determining the correlation between the processing device to be analyzed and the defect to be analyzed according to a sum of the sub-correlation weight of the first group and the sub-correlation weight of the second group.

### Claim 16 (depends on 15)

The method of claim 15, wherein the sub-correlation weight Qi of the ith group is calculated by the following formula:
Q
i
=
ln
⁡
(
G
i
/
G
total
B
i
/
B
total
)
;
where i is 1 or 2, Gi is the number of the products with the defect to be analyzed in the ith group, Gtotal is the number of all the products with the defect to be analyzed, B1 is the number of the products without the defect to be analyzed in the ith group, and Btotal is the number of all the products without the defect to be analyzed.

### Claim 17 (depends on 14)

The method of claim 14, wherein the determining the correlation weight of the processing device to be analyzed in the processing devices corresponding to the defect to be analyzed, and determining the correlation between the processing device to be analyzed and the defect to be analyzed according to the correlation weight comprises:
for each processing device to be analyzed, taking the products which are processed by the processing device to be analyzed in the production procedures as a first group, and taking the products which are not processed by the processing device to be analyzed in the production procedures as a second group;
for each of the processing device to be analyzed, determining a sub-correlation weight of the first group and a sub-correlation weight of the second group;
for each processing device to be analyzed, determining a value parameter according to the sub-correlation weight of the first group and the sub-correlation weight of the second group, and determining the correlation between the processing device to be analyzed and the defect to be analyzed according to the value parameter;
the value parameter J is calculated by the following formula:
J=ΣJi, and Ji=(DGi−DBi)*Qi;
where Qi is the sub-correlation weight of the ith group, i is 1 or 2, DGi is a proportion of the products with the defect to be analyzed in the ith group to all the products with the defect to be analyzed, and DBi is a proportion of the products without the defect to be analyzed in the ith group to all the products without the defect to be analyzed.

### Claim 18 (depends on 14)

The method of claim 14, wherein the production record further comprises: parameter values of at least a portion of process parameters used by at least a portion of the processing devices, by which the product is processed during the production procedure, for processing the product;
the method further comprises:
determining, by a preset machine learning model, the correlation between at least the portion of process parameters used by at least the portion of the processing devices and the defect to be analyzed, according to the production record.

### Claim 19 (depends on 18)

The method of claim 18, further comprises:
after the determining the correlation between at least the portion of process parameters used by at least the portion of the processing devices and the defect to be analyzed, determining a recommended parameter value range of at least the portion of process parameters according to the correlation between the process parameters and the defect to be analyzed.

### Claim 20 (depends on 14)

A computer readable medium storing a computer program, which, when being executed by a processor, implements the method for analyzing cause of product defect according to claim 14.
