Quality problem management method and system based on knowledge graph

Abstract

The invention belongs to the technical field of quality problem management, and particularly discloses a quality problem management method and system based on a knowledge graph, and the method comprises the following steps: carrying out the code definition of all possible quality problems, and automatically generating a task management record table according to an early warning rule for quality problem records; the method comprises the following steps: designing and constructing an RCA root cause analysis map and a quality defect analysis map, aiming at generated quality defects, performing problem information definition, reason analysis, measure formulation and verification according to the quality defect analysis map, and directly calling the RCA root cause analysis map according to defect types to automatically perform root cause identification. And new experience formed for the quality problem can form knowledge and fill the knowledge into a quality defect analysis map to assist in supporting business. By adopting the technical scheme, the RCA root cause analysis atlas is utilized to help business personnel to quickly identify the root cause of the quality problem and make corresponding measures, so that the problem processing efficiency and quality are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of quality problem management and relates to a quality problem management method and system based on knowledge graph. Background Art

[0002] The automobile industry will have production quality, parts quality, manufacturing process, QCM (Quality Chain Management), control map and after-sales quality problems in various links such as early production, parts delivery, manufacturing process and market sales.

[0003] Generally, quality problems are managed online through problem definition, problem input, problem generation, problem tracking and processing (offline processing), problem escalation, problem closure and problem data sharing. In the process of problem tracking and processing, the problem information is clarified according to the standard processes such as 8D / FORM4 / A3 / 6sgima, the root cause of the problem is confirmed according to the RCA root cause analysis method, and measures are formulated to handle the problem and close the loop.

[0004] Generally, quality problems in the automotive industry come from IQS systems, DDM systems, AI quality application systems, MOM systems, etc. The quality problem management application receives relevant problem data and enters quality problems online according to the quality problem definition rules to generate problem records. In the problem tracking process, an offline processing method is adopted. According to the different levels and severity of different problems, the problem information is clarified according to the standard processes such as 8D / FORM4 / A3 / 6sgima, the corresponding influencing factors are found to determine the root cause and formulate corresponding solutions. After the problem measures are verified, the online problem loop is closed and recorded.

[0005] In the process of quality problem management, online and offline methods are used for problem management. In the process of problem management, it is necessary to look through historical data to query and analyze the root causes of different quality problems and formulate solutions. The efficiency is relatively low and the process is not convenient enough. The quality problem records are not complete and the experience and knowledge formed by different problems are not uniformly managed and accumulated, which cannot be used for reference or rapid reuse in the future. Summary of the invention

[0006] The purpose of the present invention is to provide a quality problem management method and system based on knowledge graph to improve the efficiency and quality of problem handling.

[0007] In order to achieve the above object, the basic scheme of the present invention is: a quality problem management method based on knowledge graph, comprising the following steps:

[0008] All possible quality problems are coded and the quality problem records are automatically generated into a task management record table according to the early warning rules;

[0009] Design and build RCA root cause analysis maps and quality defect analysis maps;

[0010] In quality task management, for quality defects that occur, problem information is clarified, causes are analyzed, measures are formulated and verified according to the quality defect analysis map;

[0011] When conducting cause analysis, the knowledge graph is used to directly call the RCA root cause analysis graph to automatically identify the root cause based on the defect type. New experience gained from quality problems can be transformed into knowledge and filled into the quality defect analysis graph to assist in supporting business.

[0012] The working principle and beneficial effects of this basic solution are as follows: This technical solution is based on the quality problem management method of the knowledge graph and optimizes the quality problem management process. In the process of problem tracking, there is no need to look through a large amount of historical data to query and analyze the root causes of quality problem defects. The RCA knowledge graph can be directly called through the defect name in the task management record to select the corresponding root cause, quickly locate the cause of the problem and formulate corresponding solutions, thereby improving the efficiency and quality of problem handling. Problem measures are verified and the system is closed online. The new historical experience generated can be recorded to form knowledge, which can be reused by the business or assist in supporting the business later.

[0013] Furthermore, when all possible quality problems are defined by codes, all quality problems are standardized by unified coding according to defect name + defect location + defect description.

[0014] All quality issues are standardized so that when similar problems occur, relevant knowledge can be quickly retrieved and used as a reference.

[0015] Furthermore, the early warning rules are: according to the various quality defect codes, process quality monitoring indicators, and eight abnormality judgment rules defined in the SPC monitoring early warning process:

[0016] Any point greater than three standard deviations from the center line;

[0017] 9 consecutive points are on the same side of the center line;

[0018] All 6 consecutive points increase or decrease;

[0019] 14 consecutive points staggered up and down;

[0020] There are 2 points on the same side with a distance greater than 2 standard deviations from the center line);

[0021] 4 of the 5 points are on the same side and are more than 1 standard deviation away from the center line;

[0022] 15 consecutive points are within 1 standard deviation of the center line;

[0023] 8 consecutive points are greater than 1 standard deviation from the center line;

[0024] Follow the above rules to trigger quality task management.

[0025] Simple operation and easy to use.

[0026] Furthermore, according to the early warning rules, some of the original problem record information generated is filled into the task management record table, including the problem number, defect type, problem location, problem area, problem time, problem level, importance, and number of times the problem occurred on the day;

[0027] Problem information is clearly filled in by engineers according to the 8D process. Root cause identification comes from the data matching that already exists in the RCA root cause analysis map. Data that does not exist in the RCA root cause analysis map is filled in by engineers based on the 8D process identification. Definition / verification of corrective measures, implementation of corrective measures, and effect verification are filled in by engineers; control standards and control areas are derived from the quality defect analysis map or data updated by engineers.

[0028] Obtain relevant parameters and place them in the task management record table for easy use.

[0029] Furthermore, the RCA root cause analysis map is designed and constructed. The specific method is as follows:

[0030] According to the definition of quality RCA root cause analysis, the metadata in the root cause analysis process is summarized into five categories: defect type, cause analysis, root cause analysis, data source, and solution measures;

[0031] Based on the defined ontology types, the relationship names are customized according to the actual relationships between ontologies to form an ontology-relationship-ontology structure:

[0032] Defect type-correspondence-cause analysis, one defect type corresponds to multiple causes;

[0033] Cause analysis - includes - root cause analysis, one cause includes multiple root causes;

[0034] Root cause analysis-corresponding-data source, different root causes correspond to the same data source;

[0035] Data source - corresponding - solution, different data sources correspond to multiple solutions;

[0036] According to the structure of the structured data table, use the knowledge graph tool to create tasks, select data sets, map and configure the ontology types and relationships, and map the header fields of the structured table to the ontology one by one. For relationship mapping, select the entity structured header field corresponding to the start ontology and the entity structured header field corresponding to the end ontology according to the relationship structure.

[0037] According to the mapping configuration rules, start the construction task, fill the data of the structured data table into the quality analysis map, and set the incremental strategy. According to the time dimension, automatically execute the task every two weeks, extract the relevant data of the quality problem record table and fill the customized quality defect analysis map. As new problems are continuously generated, the map is filled with data, and the data of the map is used in the quality defect problem analysis process.

[0038] In the process of quality problem analysis, the RCA root cause analysis map is used to help business personnel quickly identify the root causes of quality problems and formulate corresponding measures, thereby improving the efficiency and quality of problem handling.

[0039] Further, the method of designing and constructing the quality defect analysis map is as follows:

[0040] According to the definition of root cause analysis, identify the ontology types: professional name, control standard, defect type, defect generation point, defect influencing factor, and solution measures;

[0041] According to the defined ontology types, the relationship between ontologies is confirmed to form an ontology-relationship-ontology structure:

[0042] Professional name - control standard - control standard, one major corresponds to multiple control standards;

[0043] Professional name-defect type-defect type. One professional corresponds to multiple defect types.

[0044] Defect type-defect generation point-defect generation point. One defect type contains multiple defect generation points.

[0045] Defect type-influencing factor-defect influencing factor. One defect type contains multiple defect influencing factors.

[0046] Defect generation point - influencing factor - defect influencing factor. One defect generation point corresponds to multiple influencing factors.

[0047] According to the structure of the structured data table, use the knowledge graph tool to create tasks, select data sets, map and configure the ontology types and relationships, and map the header fields of the structured table to the ontology one by one. For relationship mapping, select the entity structured header field corresponding to the start ontology and the entity structured header field corresponding to the end ontology according to the relationship structure.

[0048] According to the mapping configuration rules, start the construction task, fill the data of the structured data table into the quality analysis map, and set the incremental strategy. According to the time dimension, automatically execute the task every two weeks, extract the relevant data of the quality problem record table and fill it into the quality defect analysis map.

[0049] The quality defect analysis map is constantly accumulated with new problems and new solutions and added to the knowledge map as knowledge. When the same type of problems occur, relevant knowledge can be quickly retrieved and used as a reference.

[0050] The present invention also provides a quality problem management system based on a knowledge graph, comprising a processing unit, wherein the processing unit executes the method of the present invention to perform quality problem management in the automotive industry.

[0051] Through processing units, this system uses RCA root cause analysis maps and quality defect analysis maps to form a large amount of company-level knowledge assets as knowledge is continuously filled. It can provide auxiliary support for other business scenarios and provide learning experience for new employees who have just joined the company, helping them to quickly understand the relevant knowledge of the quality problem sector. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flow chart of the quality problem management method based on knowledge graph of the present invention. DETAILED DESCRIPTION

[0053] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0054] In the description of the present invention, it is necessary to understand that the terms "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0055] In the description of the present invention, unless otherwise specified and limited, it should be noted that the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a mechanical connection or an electrical connection, or it can be the internal connection between two components. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to the specific circumstances.

[0056] The present invention discloses a quality problem management method based on knowledge graph, such as Figure 1 As shown, the following steps are included:

[0057] All possible quality problems are defined by codes, and the quality problem records are automatically generated into a task management record table according to the early warning rules; preferably, the task management record table includes problem number, defect type, problem location, problem area, problem time, problem level, importance, number of times the problem occurs on the day, problem information, responsible person, containment measures, root cause identification, definition / verification of corrective measures, implementation of corrective measures, effect verification, control standards, completion time, control area, and knowledge accumulation fields.

[0058] Design and build RCA root cause analysis maps and quality defect analysis maps;

[0059] In quality task management, for quality defects that occur, clarify problem information, analyze causes, formulate and verify measures according to quality defect analysis charts (such as 8D / FORM4 / A3 / 6sg ima and other standard processes);

[0060] When conducting cause analysis, the knowledge graph is used to directly call the RCA root cause analysis graph to automatically identify the root cause based on the defect type. New experience gained from quality problems can be transformed into knowledge and filled into the quality defect analysis graph to assist in supporting business.

[0061] In a preferred solution of the present invention, when all possible quality problems are defined by code, all quality problems are standardized by unified code according to defect name + defect location + defect description. All quality problems are only standardized, and when the same type of problems occur, relevant knowledge can be quickly retrieved and used for reference.

[0062] In a preferred embodiment of the present invention, the early warning rules are: according to the various quality defect codes, process quality monitoring indicators (including FTT, C / 1000, etc.), and eight abnormality judgment rules defined in the SPC monitoring early warning process:

[0063] Any point greater than three standard deviations from the center line;

[0064] 9 consecutive points are on the same side of the center line;

[0065] All 6 consecutive points increase or decrease;

[0066] 14 consecutive points staggered up and down;

[0067] There are 2 points on the same side with a distance greater than 2 standard deviations from the center line);

[0068] 4 of the 5 points are on the same side and are more than 1 standard deviation away from the center line;

[0069] 15 consecutive points (on either side) within 1 standard deviation of the center line;

[0070] 8 consecutive points (on either side) are greater than 1 standard deviation from the center line;

[0071] Follow the above rules to trigger quality task management.

[0072] In a preferred solution of the present invention, part of the original problem record information generated is filled into the task management record table according to the early warning rule, including the problem number, defect type, problem location, problem area, problem time, problem level, importance, and number of times the problem occurred on the day;

[0073] Problem information is clearly filled in by engineers according to the 8D process. Root cause identification comes from the data matching that already exists in the RCA root cause analysis map. Data that does not exist in the RCA root cause analysis map is filled in by engineers based on the 8D process identification. Definition / verification of corrective measures, implementation of corrective measures, and effect verification are filled in by engineers; control standards and control areas are derived from the quality defect analysis map or data updated by engineers.

[0074] In a preferred embodiment of the present invention, the RCA root cause analysis map is designed and constructed, and the specific method is as follows:

[0075] According to the definition of quality RCA root cause analysis, the metadata in the root cause analysis process is summarized into five categories: defect type, cause analysis, root cause analysis, data source, and solution measures;

[0076] Based on the defined ontology types, the relationship names are customized according to the actual relationships between ontologies to form an ontology-relationship-ontology structure:

[0077] Defect type-correspondence-cause analysis, one defect type corresponds to multiple causes;

[0078] Cause analysis - includes - root cause analysis, one cause includes multiple root causes;

[0079] Root cause analysis-corresponding-data source, different root causes correspond to the same data source;

[0080] Data source - corresponding - solution, different data sources correspond to multiple solutions;

[0081] According to the structure of the structured data table, use the knowledge graph tool to create tasks, select data sets, map and configure the ontology types and relationships, and map the header fields of the structured table to the ontology one by one. For relationship mapping, select the entity structured header field corresponding to the start ontology and the entity structured header field corresponding to the end ontology according to the relationship structure.

[0082] According to the mapping configuration rules, start the construction task, fill the data of the structured data table into the quality analysis map, and set the incremental strategy. According to the time dimension, automatically execute the task every two weeks, extract the relevant data of the quality problem record table and fill the customized quality defect analysis map. As new problems are continuously generated, the map is filled with data, and the data of the map is used in the quality defect problem analysis process.

[0083] In the process of quality problem analysis, the RCA root cause analysis map is used to help business personnel quickly identify the root causes of quality problems and formulate corresponding measures, thereby improving the efficiency and quality of problem handling.

[0084] In a preferred embodiment of the present invention, the method for designing and constructing a quality defect analysis map is:

[0085] According to the definition of root cause analysis, identify the ontology types: professional name, control standard, defect type, defect generation point, defect influencing factor, and solution measures;

[0086] According to the defined ontology types, the relationship between ontologies is confirmed to form an ontology-relationship-ontology structure:

[0087] Professional name - control standard - control standard, one major corresponds to multiple control standards;

[0088] Professional name-defect type-defect type. One professional corresponds to multiple defect types.

[0089] Defect type-defect generation point-defect generation point. One defect type contains multiple defect generation points.

[0090] Defect type-influencing factor-defect influencing factor. One defect type contains multiple defect influencing factors.

[0091] Defect generation point - influencing factor - defect influencing factor. One defect generation point corresponds to multiple influencing factors.

[0092] According to the structure of the structured data table, use the knowledge graph tool to create tasks, select data sets, map and configure the ontology types and relationships, and map the header fields of the structured table to the ontology one by one. For relationship mapping, select the entity structured header field corresponding to the start ontology and the entity structured header field corresponding to the end ontology according to the relationship structure.

[0093] According to the mapping configuration rules, start the construction task, fill the data of the structured data table into the quality analysis map, and set the incremental strategy. According to the time dimension, automatically execute the task every two weeks, extract the relevant data of the quality problem record table and fill it into the quality defect analysis map.

[0094] The quality defect analysis map is constantly accumulated with new problems and new solutions and added to the knowledge map as knowledge. When the same type of problems occur, relevant knowledge can be quickly retrieved and used as a reference.

[0095] The present invention also provides a quality problem management system based on a knowledge graph, including a processing unit, which executes the method of the present invention to manage quality problems in the automotive industry.

[0096] The present invention standardizes the unified management of quality problems. In the process of quality problem analysis, the RCA root cause analysis map is used to help business personnel quickly identify the root causes of quality problems and formulate corresponding measures, thereby improving the efficiency and quality of problem handling.

[0097] Secondly, as new problems and new solutions are constantly accumulated and added to the knowledge map, the quality defect analysis map can quickly retrieve relevant knowledge and provide reference when similar problems occur.

[0098] As knowledge is continuously added to the RCA root cause analysis map and the quality defect analysis map, a large amount of company-level knowledge assets will be formed, which can provide auxiliary support for other business scenarios and provide learning experience for new employees who have just joined the company, helping them to quickly understand the relevant knowledge of the quality problem sector.

[0099] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0100] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A quality problem management method based on knowledge graph, characterized in that: The steps include: All possible quality problems are coded and the quality problem records are automatically generated into a task management record table according to the early warning rules; Design and build RCA root cause analysis maps and quality defect analysis maps; In quality task management, for quality defects that occur, problem information is clarified, causes are analyzed, measures are formulated and verified according to the quality defect analysis map; When conducting cause analysis, the knowledge graph is used to directly call the RCA root cause analysis graph to automatically identify the root cause based on the defect type. New experience gained from quality problems can be transformed into knowledge and filled into the quality defect analysis graph to assist in supporting business.

2. The quality problem management method based on knowledge graph according to claim 1, characterized in that: When defining codes for all possible quality problems, all quality problems are standardized by unified coding according to defect name + defect location + defect description.

3. The quality problem management method based on knowledge graph according to claim 1, characterized in that: The early warning rules are: according to the various quality defect codes, process quality monitoring indicators, and eight abnormality judgment rules defined in the SPC monitoring early warning process: Any point greater than three standard deviations from the center line; 9 consecutive points are on the same side of the center line; All 6 consecutive points increase or decrease; 14 consecutive points staggered up and down; There are 2 points on the same side with a distance greater than 2 standard deviations from the center line); 4 of the 5 points are on the same side and are more than 1 standard deviation away from the center line; 15 consecutive points are within 1 standard deviation of the center line; 8 consecutive points are greater than 1 standard deviation from the center line; Follow the above rules to trigger quality task management.

4. The quality problem management method based on knowledge graph according to claim 1, characterized in that: Fill part of the original problem record information generated into the task management record table according to the early warning rules, including the problem number, defect type, problem location, problem area, problem time, problem level, importance, and number of times the problem occurred on that day; Problem information is clearly filled in by engineers according to the 8D process. Root cause identification comes from the data matching that already exists in the RCA root cause analysis map. Data that does not exist in the RCA root cause analysis map is filled in by engineers based on the 8D process identification. Definition / verification of corrective measures, implementation of corrective measures, and effect verification are filled in by engineers; control standards and control areas are derived from the quality defect analysis map or data updated by engineers.

5. The quality problem management method based on knowledge graph according to claim 1, characterized in that: Design and construct the RCA root cause analysis map. The specific method is as follows: According to the definition of quality RCA root cause analysis, the metadata in the root cause analysis process is summarized into five categories: defect type, cause analysis, root cause analysis, data source, and solution measures: Based on the defined ontology types, the relationship names are customized according to the actual relationships between ontologies to form an ontology-relationship-ontology structure: Defect type-correspondence-cause analysis, one defect type corresponds to multiple causes; Cause analysis - includes - root cause analysis, one cause includes multiple root causes; Root cause analysis-corresponding-data source, different root causes correspond to the same data source; Data source - corresponding - solution, different data sources correspond to multiple solutions; According to the structure of the structured data table, use the knowledge graph tool to create tasks, select data sets, map and configure the ontology types and relationships, and map the header fields of the structured table to the ontology one by one. For relationship mapping, select the entity structured header field corresponding to the start ontology and the entity structured header field corresponding to the end ontology according to the relationship structure. According to the mapping configuration rules, start the construction task, fill the data of the structured data table into the quality analysis map, and set the incremental strategy. According to the time dimension, automatically execute the task every two weeks, extract the relevant data of the quality problem record table and fill the customized quality defect analysis map. As new problems are continuously generated, the map is filled with data, and the data of the map is used in the quality defect problem analysis process.

6. The quality problem management method based on knowledge graph according to claim 1, characterized in that: The method for designing and constructing a quality defect analysis map is: According to the definition of root cause analysis, identify the ontology types: professional name, control standard, defect type, defect generation point, defect influencing factor, and solution measures; According to the defined ontology types, the relationship between ontologies is confirmed to form an ontology-relationship-ontology structure: Professional name - control standard - control standard, one major corresponds to multiple control standards; Professional name-defect type-defect type. One professional corresponds to multiple defect types. Defect type-defect generation point-defect generation point. One defect type contains multiple defect generation points. Defect type-influencing factor-defect influencing factor. One defect type contains multiple defect influencing factors. Defect generation point - influencing factor - defect influencing factor. One defect generation point corresponds to multiple influencing factors. According to the structure of the structured data table, use the knowledge graph tool to create tasks, select data sets, map and configure the ontology types and relationships, and map the header fields of the structured table to the ontology one by one. For relationship mapping, select the entity structured header field corresponding to the start ontology and the entity structured header field corresponding to the end ontology according to the relationship structure. According to the mapping configuration rules, start the construction task, fill the data of the structured data table into the quality analysis map, and set the incremental strategy. According to the time dimension, automatically execute the task every two weeks, extract the relevant data of the quality problem record table and fill it into the quality defect analysis map.

7. A quality problem management system based on knowledge graph, characterized in that: The invention comprises a processing unit, wherein the processing unit executes the method according to any one of claims 1 to 6 to manage quality issues in the automotive industry.

Images