JP4984580B2 - Defect countermeasure support device - Google Patents

Description

The present invention relates to one or more entities (parts or units such as) the troubleshooting support device is the supporting measures for troubles occurring in the production system according to the product produced from, in particular, be performed efficiently troubles measure support about the possible glitches measures support device.

  Conventionally, in a production system related to a product produced from one or more parts, various problems occur due to design deficiencies, human error, etc. at each stage of the manufacturing process, testing stage, etc. . Various methods have been considered to prevent such problems.

  Some of them view the production system as the main target for detecting defects, and others set the target for detecting defects in a wider range of systems, not limited to production systems. is doing.

  In the following description, various technologies will be described as background technologies. However, according to the above-mentioned categories, the scope of defense covers a broader range of systems, not only production systems, but also production systems. It is divided roughly into what is put on.

  Naturally, there are various methods for detecting defects in a production system, but as a basic concept used in those methods, there is a concept of bill of material (hereinafter referred to as BOM). . This parts table basically presents a list of parts constituting one product as a list, and the type is engineering-BOM used for design (hereinafter referred to as E-BOM). And Manufacturing-BOM (hereinafter referred to as M-BOM) used for manufacturing. E-BOM shows what parts (parts) the product is composed of, and M-BOM uses E-BOM information for actual production activities such as product plans and schedules. It holds information to make it available within.

  In addition, other concepts used in the method of detecting defects in the production system include a strength indicating the tolerance of the target object, and another target having an interface with the target. There is a stress strength model (hereinafter referred to as SSM) that analyzes the mechanism of the occurrence of a failure, depending on the stress that indicates the driving force that causes the failure in the target (see Non-Patent Document 1 below). The SSM is originally intended to prevent problems in the manufacturing industry.

  Then, as a method for detecting (analyzing) defects in the production system using the above-described concepts, a stepwise reliability design support system (hereinafter referred to as the Koga-Aoyama model) shown in Non-Patent Document 2 below. It has been known.

  This Koga-Aoyama model is targeted for reliability evaluation in a step-by-step design (top-down design), and represents a product by three elements of “substance”, “attribute”, and “interface”. And the information which shows what kind of parts a product is constituted (namely, the above-mentioned E-BOM) is used, and the malfunction which arises in the product is analyzed. At that time, based on the SSM system, the mechanism of failure (problem) is clarified. However, in the Koga-Aoyama model, referring to the analysis results of the mechanism, as a separate process, Fault Tree Analysis (hereinafter referred to as FTA), Failure Mode and Effects Analysis (hereinafter referred to as FMEA), etc. Reliability evaluation is performed, and the work is time-consuming twice.

  That is, in the Koga-Aoyama model, although a method for a quantitative approach in a narrow sense such as the SSM method can be successfully incorporated, it is applied to a broader range of objects conceptually such as the FTA and FMEA methods. It is difficult to incorporate methods that are not based on a general approach. This is considered to be the reason why the two times trouble of performing the analysis by the FTA or FMEA method in another process after the analysis by the SSM described above cannot be avoided in the Koga-Aoyama model. That is, it is not always clear what kind of cooperation (link) should be provided between the Koga-Aoyama model and the FTA or FMEA system.

  In addition, in the case of the FMEA method, the risk priority is ranked by focusing on a specific failure event independently of the E-BOM, and the failure that occurs in a product produced from a plurality of parts is determined. There is also a situation that is not a method that can be directly applied to analysis.

  Here, FTA is targeted for failure event cause analysis (including human causes and software factors) in systems (not just production systems, but a wider range). By placing each part of the tree structure (fault tree) via the gate, it is possible to point out the vulnerability of the related part and to recommend countermeasures (see Non-Patent Document 3 below).

  FMEA is also intended for systems that prevent failure in systems (not just production systems, but a wider range), and the analysis of effects of failure events and recommended countermeasures (single hardware failures and process design are mainly targeted). While clarifying the functional connection relationship of the components of, creating a worksheet (table) that considers how failure events at the lower level affect the upper level in a bottom-up manner, taking into consideration the failure grade and countermeasures It is a systematic study that evaluates the reliability incorporated into the design drawings. This technique can rank design vulnerabilities by ranking the risk priority of failure events based on the frequency of occurrence, the degree of impact, and the difficulty of detection. Can be recommended. In addition, it is possible to perform a study for preventing a malfunction in advance for all malfunction events (see Non-Patent Document 3 below).

  As described above, various methods such as FTA and FMEA are not limited to production systems, but are intended for a wider range of systems. For this reason, when trying to apply to a specific system such as a production system, it is necessary to drop the abstract concept of FTA, FMEA, etc. into the specific system. It can be considered that the Aoyama model is not linked to various systems such as FTA and FMEA to constitute one system.

The FMEA method is one of typical methods (logic) used when performing analysis taking manufacturing process information into consideration. As is clear from the above, the Koga-Aoyama model cannot take into account manufacturing process information as an integrated system, and thus, for example, manufacture of a part among individual parts constituting a product. It is not possible to deal with a case that causes a failure (failure) due to a process.
Yasuhiko Tamura and two others “Development of a system to utilize structured knowledge about defects” Satoshi Koga and 2 others “Proposal and construction of failure information management system that supports reliability design step by step” 2001, 10th Annual Conference of Japan Society of Mechanical Engineers Junjiro Suzuki et al. “FMEA-FTA Implementation Method” 1982, Nikka Giren

An object of the present invention is to provide a corrective measures supporting equipment capable of performing efficiently defects measure support.

  One aspect of the present invention deals with an apparatus that supports countermeasures against problems occurring in a production system related to products produced from one or more entities. When supporting countermeasures against defects using this apparatus, it is usually assumed that it is necessary to extract a comprehensive set of defect phenomena that can occur in the entity prior to countermeasure support.

  However, when supporting countermeasures against defects, extracting a comprehensive set of defect phenomena that may occur in the entity prior to countermeasure support does not necessarily provide effective countermeasure support. In other words, the malfunction phenomenon is raw data coming from a person in charge of the field in the mass production department, and from the viewpoint of expression, for example, various expressions can be made for the same phenomenon, which leads to confusion.

Features of the present invention, there is provided an apparatus for supporting a countermeasure against failure occurring in the production system according to the product to be produced from one or more entities, and stores information indicating an exhaustive set of failure symptoms which may occur in the entity A first defect coverage storage means and a second defect coverage for storing information indicating a comprehensive set of failure modes obtained by arranging the failure phenomena stored in the first defect coverage storage means according to form. Storage means, failure coverage storage means for storing information indicating a comprehensive set of failure modes obtained by organizing failure phenomena that may occur in the entity, and product structure information and manufacturing process information in advance Product / process model storage means for storing model information expressed in association as a product / process model, specification and standard data, knowledge on production methods and production system Includes a knowledge data storage means for storing the knowledge data consisting of knowledge concerning knowledge and know-how in Temu, in advance,
The user stores various data such as defect phenomena, defect modes, and knowledge / know-how stored in the knowledge data storage means indicating the comprehensive set stored in the first and second defect coverage storage means. Is read out and referred to, and the model information stored in the product / process model storage means is read out to have the model information as a framework, and the user is expanded into the model information for the framework. Logic that describes the attribute value corresponding to each attribute and makes a determination using various existing methods, and status information and status as attributes of the manufacturing process information stored in the product / process model storage means Means for constructing a quality knowledge database by adding information relating to the direction of transition;
When a predetermined malfunction occurs, the constructed quality knowledge database is read, and a plurality of attribute data is associated with an entity existing in the model information, and at least a part of the associated attribute data is used. First problem countermeasure support means for calculating the probability of occurrence of a defect phenomenon in the entity and providing countermeasure support by pursuing the cause and effect analysis of the defect;
Before the predetermined failure occurs, the constructed quality knowledge database is read out, and a plurality of attribute data is associated with an entity existing in the model information, and at least a part of the associated attribute data is used. A risk factor inherent in the product using the entity is calculated as a risk priority, a second failure countermeasure support means for using the risk priority for failure prevention support,
Probability candidate list for countermeasure support by pursuing the cause of the defect of the first defect countermeasure support means, and influence range and influence candidate list for countermeasure support by the impact analysis of the defect, respectively Sorting and outputting in order of (contribution rate), while outputting the countermeasure candidate list sorted in order of probability (contribution rate) for countermeasure support by the above-mentioned defect prevention support of the second defect countermeasure support means Output means .

  The defect countermeasure support apparatus of the present invention is an apparatus that supports countermeasures against defects occurring in a production system related to a product produced from one or more entities. Then, countermeasures are examined using an exhaustive set of failure modes obtained by organizing the failure phenomena that can occur in the entity by form. Then, by associating this failure mode with elements of various methods used for countermeasure support, countermeasure support is performed using the various systems. The various methods include not only production systems such as FTA and FMEA, but also methods for a wider range of systems. An exhaustive set of failure modes makes it easy to apply the abstract concept of FTA, FMEA, etc. to a specific system, which is a production system, and drop it to that specific system. . Therefore, when countermeasure support is performed, cooperation with various methods including FTA, FMEA, and the like is possible, and malfunction countermeasure support can be efficiently performed.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
The present invention is basically intended for a production system related to a product produced from one or more parts, and supports measures against problems occurring in the production system. More specific contents of countermeasure support include analysis of a defect and support for preventing the defect in advance.

  FIG. 1 is a diagram for explaining the flow of phases until a product is introduced into the market in the above-described production system. As shown in the figure, a product is put on the market through the phases of “product design”, “production preparation”, and “mass production”.

  In the product design phase, which is the first phase, design such as function, component configuration, shape, material, and reliability is performed based on, for example, market information and product planning information. As a result of the product design, product model data such as product drawings and materials is output to a database. The product model data may include E-BOM and design FMEA.

  In the production preparation phase, which is the next phase, in addition to the product model data (E-BOM, design FMEA, etc.), equipment capacity (C / T), process capacity (C / T), workability (inspection, tooling) (Replacement), tooling capability, equipment / tooling cost, and other constraints are read, and the process / equipment / tooling tool is designed. In other words, the flow of manufacturing process configuration is designed, M-BOM is created, equipment / tool selection, layout design, work drawing (working part / reference / dimension / method, tool path) creation, and manufacturing process Reliability design (referred to as process FMEA) is performed. As a result of this production preparation, equipment / tool specification, equipment / tool layout, machine drawing, layout drawing, program, parameters, manufacturing conditions, process flow, work procedure, quality standards, M-BOM, process FMEA, etc. Each data is output to the database.

  Based on the output data, in the mass production phase, which is a subsequent phase, parts and units are processed and assembled, product assembly, equipment maintenance, quality assurance, and the like are performed.

After these phases, a finished product is put on the market.
The defect information that occurs in this product includes defect information from the market after the product is introduced to the market (hereinafter referred to as market defect), and defect information that is generated at the factory and within each of the above phases (hereinafter referred to as production). Called a bug).

FIG. 2 is a diagram showing a work flow of failure processing performed when a failure occurs in the production system of FIG.
As shown in Fig. 2, when a market failure occurs, or when a production failure occurs (in the figure, a production failure occurs at the assembly plant in the mass production equipment phase), a failure notification form is issued. It is created by a complaint issuing department such as a field worker.

In step S1, the failure phenomenon is confirmed based on the failure notification form. In the confirmation of the failure phenomenon , identification of the primary cause, determination of importance (determination of whether to respond immediately or later), determination of priority, confirmation of occurrence frequency, determination of destination are performed. In the process of the next step S2, a measure is performed for the phenomenon confirmed in S1, and in step S3, the cause of the defect is pursued to analyze the true cause of the defect.

In the next step S4, the range of parts and processes affected by a defect occurring in a certain part is investigated, the degree of influence is grasped, and the destination is determined.
Note that the processing in steps S1 to S4 described above is processing performed in response to a problem that actually occurs. On the other hand, in the subsequent steps S5 to S7, measures are taken in advance to prevent problems that may cause serious damage to the production equipment and the like when they occur from the viewpoint of risk priority described later. This is a process that is performed in order to take a note.

  First, countermeasures are examined in step S5. Here, examination of alternatives (primary countermeasures, permanent countermeasures), estimation of the effects of the countermeasures, determination of the execution order of the countermeasures, and the like are performed. In the next step S6, a plan to be executed is selected from the alternatives studied in the previous step S5, determined to be executed, and notified to the execution department. This notification to the implementation department is, for example, a specification review for the product design department, a standard document / standard document, work procedure manual, and a review of the processing method for the process design department, For the facility design department, the equipment / tool specifications, work environment, and layout are reviewed, and for the mass production department, the countermeasures for pokayoke are reviewed.

  Upon receiving the above notification, for example, the specification drawing is corrected in the product design department, the reference document is corrected in the process design department, and the specification drawing is corrected in the equipment design department. Further, in the mass production department, for example, a part related to the content of the pokayoke countermeasure in the work procedure manual is corrected.

  In the subsequent step S7, a notification to the effect of execution from each department (each execution department) notified in the previous step S6 or feedback of the execution result is received. In step S8, the processing result or the recurrence prevention measure is reported to the above-mentioned complaint issuing department, and the business flow of the series of defect processing is terminated.

FIG. 3 is a block diagram showing the configuration of the failure countermeasure support system in the production system according to the embodiment of the present invention.
In the figure, the defect countermeasure support system 1 includes a registration unit 10, a knowledge / knowhow accumulation unit 20, a reference unit 30, and a defect countermeasure support unit 40.

  The knowledge and know-how storage unit 20 stores a defect coverage table storage unit 21 that stores various coverage tables related to defects, and a product that stores model information in which product structure information and manufacturing process information are associated with each other (ie, integrated). -Process model database (Product Process Model Data Base, hereinafter referred to as PPMDB) 22, design specification data, standard data, standard data, knowledge on manufacturing methods, knowledge on production systems, knowledge on know-how, etc. The model information of PPMDB 22 was created as a framework while referring to various data stored in the knowledge data storage unit 23 that stores the knowledge data, and the defect coverage table storage unit 21 and the knowledge data storage unit 23 described above. , Quality knowledge, which is a database used to support countermeasures for malfunctions described later A database (Quality Analysis / Question Answering Knowledge Data Base, hereinafter referred to as QA2-KDB) 24 is provided. The defect coverage table storage unit 21, PPMDB 22, and QA2-KDB 24 will be described later.

  In this defect countermeasure support system 1, for example, the registration unit 10, the reference unit 30, and the defect countermeasure support unit 40 provide functions that the client device has, and the knowledge / knowhow accumulation unit 20 is a server device. The function which has is provided.

  A user of this system performs maintenance such as registration / deletion of data held in the coverage table data storage unit 21 and the PPMDB 22 via the registration unit 10 based on design information examined in advance.

  In addition, the user of this system may access the knowledge / knowledge accumulation unit 20 via the reference unit 30 and refer to the knowledge / knowhow stored as information in the knowledge data storage unit 23 or the QA2-KDB 24. it can.

  In addition, the user of the present system accesses the knowledge / know-how storage unit 20 via the defect countermeasure support unit 40, thereby providing countermeasure support for a defect that has occurred in the product or a problem that may occur. More specifically, for the trouble that occurs in the product, the cause search and the impact analysis are performed, and for the trouble that may occur in the product, support for preventing the trouble is provided.

FIG. 4 is a diagram illustrating a conceptual structure of the PPMDB of FIG.
As shown in FIG. 4, PPMDB is product manufacturing information indicating what kind of entity (unit or part) a product is composed of, that is, how the product is composed of each unit, unit Holds information indicating how each part is constructed, and how the product is manufactured from each unit, how the unit is manufactured from each part, and each part It also holds manufacturing process information indicating whether it is manufactured by such a process. That is, the PPMDB stores model information that expresses product structure information and manufacturing process information in association with each other (that is, in an integrated manner).

  More specifically, in FIG. 4, the representation model 50 represents the following information. That is, the product structure information that the entity 51 (unit 1) is composed of the entity 52 (component 1-1) and the entity 53 (component 1-2) and the entity 52 (component 1-1) include the entity 53 ( Manufacturing process information that entity 51 is completed by combining parts 1-2), manufacturing process information that entity 52 (part 1-1) is completed from raw materials through intermediate products 1 to 3, and intermediate products from raw materials The manufacturing process information that the entity 53 (part 1-2) is completed through 1-2 is expressed.

More specifically, the entity 51 is represented by an arrow 71 from the entity 52 (component 1-1) to the entity 51 (unit 1) and an arrow 72 from the entity 53 (component 1-2) to the entity 51 (unit 1). The product structure information is expressed that (unit 1) is composed of an entity 52 (component 1-1) and an entity 53 (component 1-2).
For the entity 51 (unit 1), an arrow 66 going to state 61 (initial (preparation)), an arrow 67 going from state 61 (initial (preparation)) to state 62 (intermediate product 1), and state 62 ( The entity 52 (part 1-1) and the entity are represented by an arrow 68 from the intermediate product 1) to the state 63 (intermediate product 2) and an arrow 69 from the state 63 (intermediate product 2) to the state 64 (finished product). 53 (part 1-2) is received and a state 61 (initial (preparation)) is obtained, and then, for example, the entity 52 (part 1-1) is moved to a predetermined position where assembly is performed. (Intermediate product 1) is obtained, and by combining the entity 52 (component 1-2) with the state 62 (intermediate product 1), a state 63 (intermediate product 2) is obtained, and the state 63 (intermediate product 1) is obtained. Inspection of product 2), for example, whether it works correctly That entity 51 (Unit 1) is obtained by performing, and the production process information are shown.

  Further, for the entity 52 (part 1-1), an arrow 86 directed to the state 81 (raw material), an arrow 87 directed from the state 81 (raw material) to the state 82 (intermediate product 1), and a state 82 (intermediate product 1). From state 83 (intermediate product 2) to state 83 (intermediate product 2), from state 83 (intermediate product 2) to state 84 (intermediate product 3), and from state 84 (intermediate product 3) to state 85 (finished product) The raw material is received and the state 81 (raw material) is obtained by the arrow 90 heading, and then the state 81 (raw material) is cut to obtain the state 82 (intermediate product 1). 82 (intermediate product 1) is subjected to a polishing process to obtain a state 83 (intermediate product 2), and the state 83 (intermediate product 2) is subjected to a surface treatment to obtain a state 84 (intermediate product 3). Is necessary for the state 84 (intermediate product 3). Entity 52 a check is made (PVC, i.e., parts 1-1) that is obtained, and the production process information are shown.

  For the entity 53 (part 1-2), an arrow 96 toward the state 91 (raw material), an arrow 97 from the state 91 (raw material) to the state 92 (intermediate product 1), and a state 92 (intermediate product 1). An arrow 98 from state 93 (intermediate product 2) to an arrow 99 from state 93 (intermediate product 2) to state 94 (finished product) results in the arrival of raw material to obtain state 91 (raw material), Then, the state 91 (raw material) is subjected to a forging process to obtain a state 92 (intermediate product 1), and the state 92 (intermediate product 1) is subjected to heat treatment to obtain a state 93 (intermediate product 2). The manufacturing process information is shown in that the entity 53 (finished product, that is, the part 1-2) is obtained by performing necessary inspection on the state 93 (intermediate product 2).

FIG. 5 is a diagram showing a table 100 corresponding to the representation model 50 of FIG.
In FIG. 5, the table 100 is configured by items of an entity code 101, a state code 102, a previous state code 103, a state transition order 104, a state transition code 105, an input item code 106, an output item code 107, and a configuration flag 108. Is done. The definition of each item is as follows.

Entity code: A code that is uniquely assigned to an entity such as a unit or a part Status code: A code that is uniquely assigned to various states Previous state code: Unique to the state before transition to the state indicated by the state code Code to be assigned State transition order: Serial number indicating the order of state transition in the manufacturing process until the entity is completed within one entity such as a unit or part (usually assigned in order from 1).

State transition code: A code indicating the condition of the state transition (work type) Input item code: A code uniquely assigned to an item used (input) in each manufacturing process of each entity such as a unit or part Output item code: A code uniquely assigned to an item manufactured (output) in each manufacturing process of each entity such as a unit or a part. Configuration flag: Flag table 100 indicating a part constituting an entity such as a unit. It can be seen that each entity (unit, part) holds, as data, not only a completed state (completed product) but also an intermediate state in the manufacturing process.

  For the part 1-1 (entity 52), the raw material (state 81), the intermediate product 1 (state 82), and the intermediate product 2 (state 83) correspond to the entity code 101 “part 1-1”. The status code 102 of the intermediate product 3 (state 84) and the finished product (state 85) exists. The previous state code 103 corresponds to each state code 102 and indicates information indicating which state code 102 was before the transition to the state code 102. The combination of the state code 102 and the previous state code 103 corresponding to the state code 102 indicates how the state has changed due to the state transition.

Further, when the entity code 101 is “component 1-1”, referring to the state transition order 104, it is possible to know the order of state transition until the component 1-1 is completed.
Furthermore, by referring to the state transition code 105, it is possible to know the condition for transitioning to the state indicated by the state code 101 (that is, the type of work performed for transitioning to the state indicated by the state code 101). Such state transition conditions (work types) include “assembly (maintenance of form, dependent change)”, “processing”, “painting”, “heat treatment”, and the like. In processing, painting, and heat treatment, the shape is changed, but the dependency is maintained (morphological change, dependency maintenance).

  The input item code 106 indicates to which object (state) the work content shown in the state transition code 105 is to be performed. Therefore, the state transition code 105 is not set for the state where “raw material arrival”. Similarly, the output item code 107 indicates the state of the transition destination when the work content shown in the state transition code 105 is performed on the target (state) of the corresponding input item code 106, and basically, This matches the combination of the entity code 101 and the state code 102 corresponding to the state transition code 105.

The above description has been made on the part 1-1 (entity 52), but the same applies to the part 1-2 (entity 53), and thus the description thereof is omitted.
Also for the unit 1 (entity 51), the necessary items are set in the table 100 corresponding to the expression model 50 of FIG. 4 as in the case of the component 1-1 (entity 52). In the case of the unit 1, a flag indicating the component 1-1 and the component 1-2 is further set in the configuration flag 108. When this flag is set at two or more places for the same entity code 101, it is indicated that the entity of the entity code 101 is composed of two or more entities indicated by this flag. For example, FIG. 5 shows that “unit 1 is composed of component 1-1 and component 1-2”. This corresponds to the product structure information of the unit 1 in FIG.

  In addition, in a production system related to a product produced from one or more parts, it is considered that a malfunction is caused by various causes. The potential hazards that can cause the failure are referred to below as hazards. This hazard is merely conceptual and is used for convenience of explanation.

FIG. 6 is a diagram showing an example in which hazard and interface information are set for the representation model 50 of FIG.
In FIG. 6, the component 1-1 has a hazard 1 such as receiving a wrong type of raw material when the raw material is received, and when the raw material is cut to be processed into the intermediate product 1 And a hazard 2 such as that the equipment that is supposed to perform the predetermined procedure does not operate correctly.

  In addition, the component 1-1 has an interaction with another component having its own interface. Through this interface, the component 1-1 receives an interaction such as a substance, a signal, or an energy flow from another component. In this case, the part 1-1 has a hazard 10 that this interaction exceeds the allowable range of the part 1-1. Since other hazards are the same, the description thereof is omitted.

In the above description, the unit is composed of two parts, but it is needless to say that the unit can be more complicated.
FIG. 7 is a diagram showing an expression model of a unit composed of four parts.

  In FIG. 7, the unit 3 is composed of four parts, that is, a part 3-1, a part 3-2, a part 3-3, and a part 3-4. The unit 3 expression model has (has) more hazards corresponding to the complexity of the configuration.

  In the present embodiment, for such an expression model (or PPMDB), each entity included in the model has attribute data (generally a plurality of attribute data) belonging to the entity, and the attribute data. In this configuration, necessary information (for example, the probability of occurrence of a certain defect) is calculated by using at least a part thereof.

  However, as described above, when supporting countermeasures against defects, extracting a comprehensive set of defect phenomena that may occur in the entity prior to countermeasure support does not necessarily provide effective countermeasure support. In other words, the malfunction phenomenon is raw data coming from a person in charge of the field in the mass production department, and from the viewpoint of expression, for example, various expressions can be made for the same phenomenon, which leads to confusion.

  For this reason, in the present embodiment, in order to make it possible to handle the failure phenomenon quantitatively to some extent as an analysis target, a failure mode is provided in which the failure phenomena are arranged hierarchically according to form. By using this failure mode in various methods (logic) used for defect countermeasures, it is possible to quantitatively handle defect countermeasure support in the manufacturing process, which is difficult with the conventional quantitative approach in the narrow sense. It is possible.

  For example, in the conventional Koga-Aoyama model, when quantitatively analyzing the probability of occurrence of a failure phenomenon, a failure occurs when the condition of stress> strength is satisfied using the SSM method. However, such a quantitative approach can be applied only to the interaction propagating through the interface between the entities described above, and the analysis result by the SSM method or the like is provided at the latter stage of the analysis work. Based on this, as a separate process, we analyze the causal relationship and risk priority using various methods such as FTA and FMEA. The reason why the work is troublesome twice in the conventional example is that the quantitative approach in the narrow sense like this SSM method and the quantitative approach like the FTA and FMEA methods are not necessarily premised. This may be because there is no link between them.

  In the present embodiment, it is assumed that a method for a quantitative approach in a narrow sense such as the SSM method and a quantitative approach that is applied to a broader range of objects conceptually such as the FTA and FMEA methods are not necessarily used. As a data structure for establishing a link with a method that does not exist, a defect coverage table storage unit that comprehensively stores various defects shown in FIG. 3 is provided.

FIG. 8 is a diagram showing various coverage tables stored in the defect coverage table storage unit 21 of FIG.
As shown in FIG. 8, the coverage table stored in the failure coverage table storage unit 21 of FIG. 3 includes the failure (failure) part coverage table 21a, the failure (failure) mode coverage table 21b, and the failure (failure) phenomenon coverage table 21c. , A failure (failure) result coverage table 21d, a failure (failure) cause coverage table 21e, and a failure (failure) countermeasure coverage table 21f. Of these tables constituting the failure coverage table, the failure (failure) result coverage table 21d, the failure (failure) cause coverage table 21e, and the failure (failure) countermeasure coverage table 21f are described in Non-Patent Document 4 below. , Based on the well-known technique "Mandaru failure". In this “Failure Mandala”, all failure cases are recognized as human errors, and the context of failure is structured as a flow of human causes → human actions → results. Then, if each phase of “cause”, “behavior”, and “result” is muffled, those mumbles are classified into hierarchies, systematized and standardized, and knowledge of failures is thereby obtained.
General Yotaro Hatamura “Failure Knowledge Database Structure and Representation” December 2002, Science and Technology Promotion Foundation Failure Knowledge Database Development Project The following is an example of each coverage table stored in the defect coverage table storage unit 21 of FIG. This will be described with reference to FIGS. In the present embodiment, “failure” is interpreted as conceptually wider than “failure”. That is, “failure” refers to an object, a state, and things that are undesirable regarding the event, whereas “failure” refers to an object that is undesirable regarding the product (thing).

  FIG. 9 is a diagram showing an example of a defect (failure) part coverage table 21a corresponding to a certain product line. The defect (failure) part coverage table 21a comprehensively defines the part where the defect (failure) occurs. As shown in the figure, the parts are defined hierarchically.

  FIG. 10 and FIG. 11 are diagrams showing an example of a failure (failure) mode coverage table 21b corresponding to a manufacturing process of a certain product line or a product line. The failure (failure) mode coverage table 21b comprehensively defines failure (failure) modes that can occur in the product line. Here, the failure (failure) mode is a pattern obtained by organizing the failure (failure) phenomenon according to the form. The failure (failure) mode is organized hierarchically, and a failure (failure) mode coverage table is provided. can get. Examples of failure (failure) modes include disconnection, short circuit, breakage, wear, and the like. 10 shows an example of a failure (failure) mode based on a human error, while FIG. 11 shows an example of a failure (failure) mode based on a physical phenomenon.

  FIG. 12 and FIG. 13 are diagrams showing an example of a malfunction (failure) phenomenon coverage table 21c corresponding to a manufacturing process of a certain product line and a product line. The failure (failure) phenomenon coverage table 21c comprehensively defines failure (failure) phenomena that may occur in the product line. Here, in the malfunction (failure) phenomenon coverage table 21c, the object (elements of the table) is that a human being can sense with an organ or the like or can be measured with a meter or the like. That is, the malfunction (failure) phenomenon means a phenomenon itself recognized as a malfunction (failure). For example, it cuts, does not move, touches, breaks, wears out, makes a strange sound, the voltage exceeds xxV, it smells strange, it is hot, the surface is rough, etc. FIG. 12 shows an example of a malfunction (failure) phenomenon based on a human error, while FIG. 13 shows an example of a malfunction (failure) phenomenon based on a physical phenomenon.

  In FIG. 12 and FIG. 13, the failure (failure) phenomenon coverage table 21 c shows the failure (failure) phenomenon in association with the above-described failure (failure) mode. By using such a display format, if a failure (failure) mode corresponding to a failure (failure) phenomenon is used for subsequent analysis, a failure notification form is created for the same phenomenon (on-site). The person in charge can avoid the inconvenience of using the failure (failure) phenomenon in the analysis, which can be expressed in various ways and lead to confusion, and the proficiency level and proficiency level of the person in charge (user of this system) Human factors can be reduced.

  As shown in FIG. 12, particularly for human errors, the expression of error phenomenon is used instead of the expression of malfunction (failure) phenomenon, and the corresponding malfunction (failure) mode is also called error mode. You may make it distinguish from another malfunction (failure) phenomenon and malfunction (failure) mode.

  FIG. 14 is a diagram illustrating an example of a failure (failure) result coverage table 21d corresponding to a certain product line. Here, the failure (failure) result indicates damage that appears as a result of the occurrence of the failure (failure) due to the occurrence of the failure (failure) in the product of the product line or the manufacturing process thereof. The defect (failure) result coverage table 21d defines the damage that appears as a result of the occurrence of such a defect (failure) in a hierarchical manner and is comprehensively defined. In FIG. 14, for the upper items of this hierarchical classification, items are set in accordance with the above-mentioned “failed mandala”. In this “Failure Mandala”, the results are classified into human, physical, organizational / social results, results with external influences, results that must occur in the future, and results that may occur. But it is classified according to this classification. However, as a matter of course, items lower in the hierarchy differ depending on the individual system to which this “failure” is applied. In the production system related to the product produced from the entity of this embodiment, the lower items belonging to the “result to product” are “damage”, “defect phenomenon”, “function failure”, Furthermore, the subordinate items belonging to the item “defect phenomenon” include “electrical failure”, “chemical phenomenon”, “thermal fluid phenomenon”, “mechanical phenomenon”, and the like.

  FIG. 15 and FIG. 16 are diagrams showing an example of a failure (failure) cause coverage table 21e corresponding to a manufacturing process of a certain product line or a product line. Here, the cause of the malfunction (failure) refers to the cause of the malfunction (failure) occurring in the manufacturing process of the product system or the product. The failure (failure) cause coverage table 21e defines such failure (failure) causes hierarchically and comprehensively.

  FIG. 15 shows an example of the cause of failure (failure) based on a human error, while FIG. 16 shows an example of the cause of failure (failure) based on a physical phenomenon. Among these, in the example of the cause of failure (failure) based on the human error in FIG. 15, the upper item of the hierarchy (item located at the left end in the drawing) is set in accordance with the above “failure mandala”.

  17 and 18 are diagrams showing an example of a failure (fault) countermeasure coverage table 21f corresponding to a certain product system. If a defect (failure) occurs in a product in the product line or its manufacturing process, damage (corresponding to the failure (failure) result) shown as a result of the occurrence, or the cause of the damage (cause of the failure (failure)) In order to prevent such a failure (failure) from occurring, it is necessary to take measures against the failure (failure) (that is, measures against the failure (failure)). The trouble (fault) countermeasures coverage table 21f defines such trouble (fault) countermeasures in a hierarchical manner and is comprehensively defined.

  FIG. 17 shows an example of a countermeasure (failure) countermeasure based on a human error, while FIG. 18 shows an example of a countermeasure (failure) countermeasure based on a physical phenomenon. Of these, in the example of countermeasures against malfunctions (failures) based on human error in FIG. 17, items corresponding to “principle” and “problem” in the upper items in the hierarchy (the items on the left side in the figure correspond to “principle” and “problem”) This is an item for classifying items, and countermeasures correspond to items of realization methods).

  When the PPMDB of FIG. 4 or FIG. 5 is used to support countermeasures against defects that occur in the production system related to the product, the individual components constituting the framework of product structure information and manufacturing process information possessed by the PPMDB. An entity (unit or part) is provided with attribute data (generally a plurality) belonging to the entity and various logics (methods) using the attribute data. That is, based on PPMDB, QA2-KDB is created by further adding attribute data and various logics for each entity (unit or part) and each state of the entity. Then, information (for example, the most suspected cause of failure for a certain failure phenomenon) referred to when providing countermeasure support is calculated using the various logics of the QA2-KDB. In that case, the user of the trouble countermeasure support system according to the present embodiment refers to the various trouble coverage tables described above when setting values for trouble phenomena, trouble modes, trouble causes, and the like.

  The hazard as a potential danger source that can cause the above-described problem can be set on this QA2-KDB as well as PPMDB in FIG. FIG. 19 is a diagram illustrating an example of the hazard set on such QA2-KDB. In each process of manufacturing, etc., a place where a defect may occur is identified as a hazard and described in the figure, so that there is an advantage that a hazard source can be visualized and an image can be easily obtained. Note that this hazard is conceptual only, and is used for convenience of explanation as described above.

  As FIG. 4 is a conceptual diagram of PPMDB 22, FIG. 19 is also a conceptual diagram of QA2-KDB 24. For reference, the framework of QA2-KDB 24 (that is, based on PPMDB 22) will be described below. The process of automatically generating (QA2-KDB) in which no attribute value is registered in each item of attribute data and the logic is undefined will be described with reference to FIGS. .

FIG. 20 is a flowchart of the automatic generation process.
In FIG. 20, the PPMDB information 100 is acquired from the PPMDB 22 in step S11. An example of this information is shown in the table 100 of FIG.

  Subsequently, in step S12, using the acquired PPMDB information 100 (that is, referring to the configuration flag in the table 100 in FIG. 5), the product structure information and the elements (sub-substances) constituting each entity are hierarchized. (Blocks, units, parts, etc.). For example, for the table 100 in FIG. 5, information that the unit 1 is composed of the component 1-1 and the component 1-2 is extracted.

  In step S13, an entity data model template (template) corresponding to each extracted entity (including sub-entities) is acquired from a database (not shown). For example, in FIG. 5, the entity data model template corresponding to the unit 1, the component 1-1, and the component 1-2 is acquired. Note that this database holds entity data model templates corresponding to entities of all blocks, units, parts, and the like that constitute products of the target product line. FIG. 21 shows an example of the entity data model template 211 for the unit 1 and the entity data model template 212 for the component 1-1. In addition, although not shown in figure, naturally the entity data model template for components 1-2 also exists.

  In step S14, a data structure 214 corresponding to the product model is created based on the various entity data model templates and the product configuration (that is, product structure information). An example of the data structure 214 corresponding to the product model is shown in FIG. As shown in FIG. 21, in the data structure 214 corresponding to the product model, the various entity data model templates acquired in step S13 are the product structure information (in this case, the unit 1 is composed of the parts 1-1 and 1-2). Are arranged according to the information).

  In addition, a state data model template (template) corresponding to each state of each entity is created in advance and held in a database (not shown). In step S15, a state data model template 213 corresponding to each state of each entity is acquired from this database. In subsequent step S16, a state data model template 213 corresponding to each state of each entity acquired in step S15 is obtained. Based on the PPMDB information 100 (that is, referring to the state transition order 104 of the table 100 in FIG. 5), the data corresponding to the product / process model is sequentially added to each entity data model template acquired in step S14. A structure 215 is created.

  FIG. 22 shows an example of the data structure 215 corresponding to the product / process model. In the figure, for each unit 1, component 1-1, and component 1-2, corresponding state data model templates are added in the order based on the state transition order of FIG.

In the above description, various templates (entity data model template and state data model template) have been created in advance. The creation method will be briefly described below.
First, in the case of an entity data model template, it is organized and created from product design information (including knowledge and know-how). Product design information refers to information, knowledge, and know-how related to product design work. Various specifications such as functionality, performance, materials (materials), and geometric shapes of units or parts, business standards, quality standards, etc. Standards (these correspond to know-how), defect information (phenomenon, cause, countermeasures, etc.) due to past design defects (these correspond to knowledge / know-how), etc.

  In the case of a state data model template, it is organized and created from process design information and production manufacturing information (including knowledge and know-how). Process design information / production / manufacturing information is information / knowledge / know-how related to process design work and manufacturing work for manufacturing products. Unit / part manufacturing method, conditions, process order, equipment, jigs, machine specifications , Work specification information such as layout, various standards such as business standards and quality standards (these are equivalent to know-how), defect information (phenomenon, cause, countermeasures, etc.) due to past design defects (these are knowledge and know-how) Equivalent), various knowledges and know-how from production sites, countermeasures against pokayoke, etc.

  The framework of QA2-KDB24 of FIG. 3 is generated by the processing described above. Thereafter, various logics are added to the QA2-KDB24. For example, in FIG. 23, logic 1 and logic 2 are added to each of the unit 1, the component 1-1, and the component 1-2.

  The logic 1 compares whether the stress that the other entity gives to the entity having the logic 1 through the interface and the strength that the entity has, so that a failure based on the interaction occurs in this entity. Is a logic for determining The logic 1 includes, for example, SSM logic (a specific example will be described later). According to the SSM logic, a condition for causing a failure in the entity is given by stress> strength.

  Each entity has a manufacturing process (“assembly” is also included in the manufacturing process) as shown in FIGS. 4 and 5. The manufacturing process is constituted by a plurality of states and an arrow connecting these consecutive states. Each state (corresponding to the state code in FIG. 5) satisfies the state transition condition indicated by the state transition code 105 in FIG. 5 with respect to the previous state (corresponding to the previous state code in FIG. 5) (ie, By performing the type of work indicated in the code), the current state is entered. Therefore, the failure that occurs in the current state involves transition conditions including hazard mixing from the previous state and the previous state to the current state.

The logic 2 in FIG. 23 is a logic for calculating the causal relationship of defects in the manufacturing process of the entity in the entity having the logic 2.
Further, in the defect countermeasure support system of the present embodiment, the registration unit 10 in FIG. 3 can register (attribute) values for each of the above-described attributes, and can newly define or redefine logic. .

FIG. 24 is a diagram illustrating an example of functions of the registration unit 10 in FIG.
In FIG. 24, the user of the trouble countermeasure support system of the present embodiment sets an attribute value for each attribute set in the QA2-KDB 24 via the registration unit 10 in FIG.

  Registration of this attribute value is not an automatic process but is performed manually by a human. In the present embodiment, as shown in the figure, various master tables stored in the master table 241 based on the various coverage tables created in advance and stored in the defect coverage table storage unit 21, that is, the part master table By listing the contents of the phenomenon master table, result master table, failure mode master table, cause master table, countermeasure master table, etc. on the display of the user information processing terminal, for example, and letting the user select the necessary items , You can register attribute values. For example, when an attribute value is set for the attribute “phenomenon” in the illustrated entity / state data model, a phenomenon master table is displayed. The phenomenon master table stores specific phenomena, for example, heat generation, vibration, noise, and the like. By listing these and allowing the user to select them, attribute values for the attribute “phenomenon” can be registered. .

  Furthermore, the attribute values selected from the various master tables may be reflected in the knowledge / know-how sheet in the knowledge data storage unit 23 of FIG. When writing from the master table to the knowledge / know-how sheet, a comment or the like corresponding to the attribute value may be included.

In addition to the above-described attributes, each attribute value of the entity / state data model can be updated including other attributes (impact level, etc.).
Here, the update of the attribute value is, for example,
・ Update the attribute value related to occurrence frequency (attribute) from 2 times / month to 3 times / month ・ Update attribute value related to tolerance (strength) (attribute) from -30 ° C to + 50 ° C to -20 ° C to + 55 ° C The attribute value related to the influence degree (attribute) is updated from 2 to 5.

  Also, in FIG. 24, the user of the information processing terminal selects an arbitrary master table and displays it, and on this display screen, the user performs input operations such as changing, adding, or deleting the contents of the master table. By doing so, maintenance of the master table 241 can also be performed.

  The failure countermeasure support function realized by the failure countermeasure support unit 40 of FIG. 3 using various information stored in the knowledge / knowhow storage unit 20 includes a failure analysis support function and a failure prevention support function. The defect analysis support function includes a defect cause pursuit (support) function and a defect impact analysis function.

FIG. 25 is a diagram for explaining the overall processing flow performed by the failure countermeasure support unit 40 of FIG. 3 together with a simple flow.
In FIG. 25, when a defect occurs, a defect communication form indicating a part (part) where the defect has occurred and a defect phenomenon that has occurred in that part is created by, for example, a person in charge of the site. The user of this system (the person in charge at the site may also serve) takes in the contents described in the defect communication slip via the registration unit 10 in FIG. 3 (step S21). Thereby, the phenomenon of the malfunction corresponding to the site where the malfunction occurred can be understood. Then, in step S22, by performing the analysis, the entity (part) / process is known, and the cause candidates are listed. A search is performed using the part, phenomenon, and cause thus obtained as search keys, and candidate countermeasures are listed as search results. The above analysis results are organized as a set of parts, phenomena, causes, and countermeasures.

  In step S23, countermeasure consideration is performed. In this countermeasure examination, countermeasures to be taken are selected from the above-mentioned countermeasure candidates, and candidate countermeasures to be taken in order to prevent the occurrence of defects are listed from the viewpoint of risk priority.

FIG. 26 is a diagram for explaining the above-described defect cause pursuit support function.
In FIG. 26A, a common set of a comprehensive set A of failure modes corresponding to an entity (part) in which a failure has occurred and an exhaustive set B of failure modes corresponding to a failure phenomenon that has occurred in the entity. By taking the above, an exhaustive set C of failure modes associated with the occurrence of the failure phenomenon in the entity (part) is extracted. FIGS. 26B and 26C are diagrams showing the relationship between the defect coverage tables corresponding to the description of FIG. Referring to FIG. 26 (b), the failure mode K1, K2, K3 corresponds to the entity (for example, part) B1 where the failure has occurred as an exhaustive set, and FIG. ), It can be seen that the failure modes K1 and K3 correspond to the failure phenomenon G11 occurring in the entity B1 as an exhaustive set. Then, the failure cause pursuit support function causes the failure causes C1 and C3 corresponding to the failure modes K1 and K3 (which correspond to the set C) as the common set, and the failure phenomenon G11 occurs in the entity B1 ( Alternatively, it is estimated that this may occur.

  Here, the failure modes K1 and K3 as the common set relate the entity B1 in which the failure has occurred and the failure phenomenon G11 that has occurred in the entity B1. In the present embodiment, the probability that the failure phenomenon G11 occurs in the entity B1 is calculated through the failure modes K1 and K3 as the common set.

FIG. 27 is a diagram for more specifically explaining the defect (failure) cause pursuit (support) function of FIG.
In the expression model shown in FIG. 27, the unit 1 is composed of a part 1-1 and a part 1-2, which has the product structure information and the manufacturing process information as described above. Is omitted (since it is a repetition of the above).

  As described above, in the manufacturing process, a transition condition including a hazard mixture from the previous state and the previous state to the current state is involved in the malfunction occurring in the current state. For example, in the figure, in the component 1-1, the state 2 can be the cause of the failure for the state 3, and the state 1 can be the cause of the failure for the state 2. That is, the direction from the state 3 to the state 2 or the state 1 indicates a direction in which the cause of the malfunction that has occurred in the state 3 is pursued. On the other hand, the direction from state 2 or state 1 to state 3 is the direction in which a potential hazard (hazard) that can cause such a failure propagates, or the failure that occurred in states 1 to 3 It shows the direction of propagation to influence others. The direction of the arrow labeled “Propagation by state transition” in the figure matches the propagation direction of this hazard.

  On the other hand, an interface has been set for the component (the interface is set from the information processing terminal by the user), and the component is subjected to stress, and the stress is within the allowable range (resistance, strength) of the component. If it exceeds, a problem will occur. In addition, the figure assumes the case where a malfunction has occurred in the component 1-2. For this reason, the direction of the arrow indicated as “propagation by interaction (interface)” in the figure is the component 1-1 (that is, another component as viewed from the component 1-2, but as described above, it is determined by the user. An interface is set in advance between the component 1-1 and the component 1-2), and is directed to the component 1-2.

  Therefore, when a failure occurs in a certain part, it is necessary to first determine whether it is due to stress from other parts or whether there is a problem in the manufacturing process of the part itself. In the expression in the figure, it is first necessary to determine whether the interaction branch or the state transition branch.

The outline of the trouble (failure) cause pursuit support process will be described below.
First, the cause path of a possible failure (failure) mode is identified. In other words, for each entity, the state transition of the entity along the hazard propagation path and the failure (failure) modes that can occur due to the interaction given by other entities are comprehensively defined and given from each influence source. The degree of influence (the degree of influence (contribution rate)) and the possibility of the occurrence of the failure (fault) mode are identified probabilistically and systematically. As a result, a cause path is created.

Further, the frequency at which those possible failure modes described above actually occur is further considered.
Furthermore, systematically describe the effect on strength (what and how much) due to state transitions, and the effect on stress (what and how much) caused by interactions from other entities.

Next, cause analysis of the failure (failure) mode that occurs is performed. That is, the analysis is performed by tracing back the hazard propagation path in the reverse direction (backward) from the failure (failure) occurrence site.
First, the cause of failure (failure) is determined probabilistically as to whether the cause is only due to a decrease in strength, only due to a deterioration in stress, or due to both a decrease in strength and a deterioration in stress .

When the probability of strength reduction is high, an influence source in a state transition with a high probability that can occur in the manufacturing process of an entity is listed (in descending order of probability) with an emphasis on the state branch transition.
When there is a high probability of stress deterioration from other entities, emphasis is placed on interaction bifurcation, and sources of influence that are likely to occur in other entities are listed (in descending order of probability).

  In both cases, both the influence source in the state transition with high probability that can occur in the manufacturing process of the entity and the influence source with high probability that can occur in the other entity are listed (in descending order of probability).

  In the case of an assembly process in which a state transition occurs from an entity when the probability of stress deterioration from another entity is high, or when the probability of strength reduction is high, the other entity or the entity in which the state transition occurs is defective ( The failure (failure) mode is repeatedly analyzed assuming the failure mode.

Hereinafter, the above-described failure (failure) cause pursuit (support) function will be described with reference to the flowcharts of FIGS. 28 and 29.
First, prior to the defect cause pursuit process, the user of the present system receives a defect report form from the person in charge of the site. This defect communication form describes a product in which a defect has occurred, a part in which the defect has occurred in the product, a defect (fault) phenomenon that has occurred in that part, and the like. In step S101, the user inputs, from the information terminal device, the system name of the product in which the defect described in the defect notification form has occurred. In response to the input of the system name, in step S102, the master table group stored as data for all product systems, the master table of the product corresponding to the system name input from the QA2-KDB group, QA2 -KDB is selected (target product line is squeezed out).

  In the subsequent step S103, the part master table for the product system that is the current target is read, and the user selects a defective part described in the trouble communication slip. In step S104, the QA2-KDB (attribute / attribute value) of the target product line is read, and the part where the defect has occurred is selected from the QA2-KDB by the user. At this time, based on the link relationship between the defect coverage tables (this link is not an automatic process but manually performed by a human), a list of possible defect modes in the target defect part (FIG. 26B). ) Displays K1, K2, K3) and a list of malfunction phenomena corresponding thereto (G11, G12, G21, G22 in FIG. 26C).

  Subsequently, in step S105, the user determines whether or not there is a malfunction phenomenon corresponding to (matching) the malfunction phenomenon targeted at this time in the list. Taking FIG. 26 as an example, it is determined by the user whether the malfunction phenomenon targeted this time matches any of G11, G12, G21, and G22 in the list of FIG.

  As described above, if it is determined in step S105 that the displayed failure phenomenon list does not include the failure phenomenon corresponding to the target failure phenomenon, the phenomenon master table is opened by the user in step S106. In step S107, the user determines whether or not there is a target malfunction phenomenon in the phenomenon master table. If it is determined that there is not, the user maintains the phenomenon master table to add the target defect phenomenon as data in step S108, and then the user selects (registers) the added defect phenomenon in step S109. ) If it is determined that the event is in the phenomenon master table, the process of step S109 is immediately performed.

  After that, in step S110, using the defect phenomenon (attribute / attribute value) registered in step S109 as a key, the defect phenomenon (attribute / attribute value) from the defect mode (right side) (attribute / attribute value) of QA2-KDB. And the corresponding failure mode (attribute / attribute value). If it is determined in step S105 that the displayed failure mode list includes a failure phenomenon corresponding to the target failure phenomenon, step S110 is immediately executed.

  Subsequently, the process proceeds to FIG. 29. In step S111, the stress event (left side) related to the failure mode (attribute value) identified in step S110 (that is, the site where the failure occurred is from another site having an interface). Extract stress events received as input).

  Then, in step S112, the strength corresponding to each of the stress events extracted in step S111 is further extracted, and the stress status at the time of the failure is confirmed with the user (such as the on-site person who created the failure communication form). . As a result of this confirmation, the user inputs stress data or checks the presented stress strength comparison result (Yes / No / Unknown). The comparison result (unknown) is checked when the confirmation from the user cannot be obtained and the stress data at the time of the failure becomes unknown.

Then, the following steps S113 and S114 are executed for all the stress items extracted in this manner (related to the current defect).
That is, first, in step S113, it is determined whether the stress exceeds the allowable range indicated by the strength (that is, whether stress> strength). If the allowable range is exceeded, in step S114, the entity that leads to the stress event is set as a new defect site.

  If the processes in steps S113 and S114 are completed for all the stress items, the process proceeds to step S115, and it is determined whether there is a new defective part (entity) set in step S114. If there is, in step S116, the defective part set in step S114 is set as a new target (problem part) for cause analysis, the process returns to step S104, and the above steps S104 to S115 are repeated.

On the other hand, if no new defective part is set in step S114, the process proceeds to step S117.
In step S117, based on the failure mode identified in step S110, except for the failure mode related to the stress extracted in step S111, the remaining events (left side) that can be the cause are extracted, and each of the extracted The degree of influence (contribution rate) of the event to the target failure mode is calculated from the logic related to the target failure mode (such as FTA logic) and the occurrence probability of the end event.

  In step S118, the event extracted in step S117 is set as a cause (candidate) (target defect), and the corresponding influence (contribution rate) calculated (in step S117) is used as the cause (the cause candidates). ) Is true (probability that the cause is the cause), and the cause (factor) candidate list is displayed in a list on the display of the information processing terminal. . Alternatively, the contents corresponding to the list display result are stored in the QA2-KDB. This list display is an (analysis) result of the cause pursuit support process, and is referred to as a “cause (factor) candidate list”. In addition, since it is sorted and output in the order of probability, it is also called a “cause probability (contribution rate) order list”.

  FIG. 30 is a diagram for explaining how a cause (factor) candidate list is output in the cause seeking support process (for the target defect). As shown in the figure, an item “link” may be provided for each cause item on the cause (factor) candidate list. By referring to this link information (for example, specifying the link on the screen), it is possible to refer to materials and data related to the cause item.

Hereinafter, the cause seeking support process will be described more specifically with reference to FIGS. 31 to 37.
FIG. 31 is a diagram illustrating a construction example of QA2-KDB corresponding to a unit configured with a conductive plate and a base metal as parts.

  In FIG. 31, the unit 1 is composed of a component 1-1 (conductor plate) and a component 1-2 (base metal), and each entity (unit 1, component 1-1, component 1-2) is represented by an SSM logic. And FTA logic.

FIG. 32 is a diagram showing the entity / state data model of the part 1-1 (conductor plate) constituting the unit 1 of FIG. 31 together with the SSM method (logic) applied to the part 1-1.
In FIG. 32, a component 1-1 (conductive plate) has an interface, and stress 1 (stock temperature) from other components (not shown) due to heat generation of other adjacent components (not shown), Stress 2 (stock humidity) and stress 3 (stock period).

  28 and 29 described above, the failure mode 5 (as the failure mode associated with the failure phenomenon 5 ("rust is generated on the surface") occurring in the component 1-1 (conductive plate) is described. Rust) is acquired in step S110, and stress 1 (stock temperature), stress 2 (stock humidity), and stress 3 (stock period) are obtained in step S111 as stress events (left side) related to failure mode 5 (rust). Extracted. In the SSM method logic of failure mode 5 (rust) of this component 1-1, if the conditional expressions: (stress 1> strength 1) & (stress 2> strength 2) & (stress 3> strength 3) are true, Failure mode 5 (rust) is set to true. Strength 1, strength 2, and strength 3 are attributes indicating a rust temperature threshold, a rust humidity threshold, and a rust period threshold, respectively, and here, values of 30 ° C., 85%, and 10 hours are attributes. Stored as a value.

The user confirms the stress state at the time of occurrence of the malfunction with respect to the user, and as a result, it is determined whether or not the stress exceeds the above threshold values.
FIG. 33 is a diagram showing the entity / state data model of the part 1-1 (conductor plate) constituting the unit 1 of FIG. 31 together with the FTA method (logic) applied to the part 1-1.

  In FIG. 33, the entity / state data model of the component 1-1 (conductor plate) corresponds to each failure mode in each state of the manufacturing process of the component 1-1 (conductor plate), and outputs the failure mode. It has multiple FTA method logic.

In the figure, the FTA method in failure mode 2 (warp) has manufacturing conditions (component feed direction failure), manufacturing conditions (warp correction force failure), and environmental conditions (large curl) as end events. Set a value for failure mode 2 (warp).
Logical formula: (Component feed direction failure) + (Warpage correction force failure)-(Long curling)
Also, the failure mode 6 (warping) FTA method logic uses the stock condition (occurrence of external force due to superposition) and the failure mode output by the preceding FTA (not shown) as end events, and the failure mode 6 is expressed by the following logical expression. Set a value for (Warp).
Logical expression: (External force generation due to stacking) + (Failure mode output by FTA in the previous stage)
In the drawing, the failure mode 2 (warp) is a failure mode related to the press working process, and the failure mode 6 (warp) is a failure mode related to the stock process.

  In this embodiment, the method logic of the failure mode related to a certain state can also be configured to have the failure mode related to the previous state of the state on the left side (that is, the input side). Therefore, when several stages of method logic that inputs such a failure mode in the previous stage continue, the overall image is a large fault tree (the tree hierarchy from the top event to the end event is deep). Will have.

  28 and FIG. 29 described above, the failure mode associated with the failure phenomenon 6 (“the warpage is outside the range of 0 to 0.2 mm”) occurring in the component 1-1 (the conductive plate). The failure mode 6 (warp) is acquired in step S110, and the above-mentioned stock conditions (external force generation due to laying) are described as events related to the failure mode 6 (warp) (including the failure mode) (left side). As the failure mode output by the FTA of the previous stage goes back to the causal chain as an event (left side) related to the preceding FTA, the failure mode 2 (warp) ) FTA related events (left side) include manufacturing conditions (component feed direction failure), manufacturing conditions (warp correction force failure), and environmental conditions (large curl) at step S117. Is, by fault tree that includes each of these events, the degree of influence of each event for the failure mode 6 (warp) (contribution to the result event cause event) is (reverse) is calculated.

There are various methods for calculating the contribution rate of the cause event in the fault tree to the result event. These methods will be described below with reference to FIGS. 34 to 35.
FIG. 34 is a diagram illustrating an example of a fault tree as a target for calculating an occurrence probability (contribution rate).

In FIG. 34, the top event F is obtained from the end events F1, F2, and F3 by the following logical expression.
Logical expression: F = F1 + F2 · F3
A method for calculating the occurrence probability of an event in a fault tree is known.

For example, the logical sum P (E∪F) of the occurrence probabilities P (E) and P (F) of the events E and F that are not mutually exclusive is expressed by the following equation as described in Non-Patent Document 5 below. Given.
P (E∪F) = P (E) + P (F) −P (E∩F)
Nobuyuki Yamamoto “Statistics Summary” April 1964, Akibunbo The failure tree corresponding to the above equation, ie, top event A (occurrence probability P (A)) is end event E1 (occurrence probability P (E1) ) And terminal event E2 (occurrence probability P (E2)), and when events E1 and E2 are independent events, P (E∩F) = P (E) P (F) holds. The occurrence probability P (A) of the event A is as follows. P (A) = P (E1) + P (E2) −P (E1) P (E2) In the present embodiment, in particular, considering that the above-mentioned failure mode is not generally a rejection event, The contribution ratios PA (E1) and PA (E2) to the result events of the tree events E1 and E2 are calculated by the following equations. PA (E1) = (1-P (E2) / 2) P (E1) / P (A) PA (E2) = (1-P (E1) / 2) P (E2) / P (A) where Even if end events (cause events) E1 and E2 may occur simultaneously (that is, even if P (E1２E2) = P (E1) P (E2) ≠ 0), PA (E1) + PA In the present embodiment, (E2) = 1 always holds, that is, all end events (cause events) always have a total contribution rate to the top event (result event) of “1”. It is a feature of the structure and is easy to understand intuitively.

  In the present embodiment, as shown in the above equation, when the tree structure indicates a structure in which the result event (A) is derived from the logical sum of a plurality of cause events, which are not mutually exclusive events, the result A term (P (E1) P (E2)) composed of two or more products of the occurrence probabilities of the plurality of cause events in the expression expressing the occurrence probability of the event (A) by the occurrence probabilities of the plurality of cause events ) Is distributed to two or more of the occurrence probabilities of the plurality of cause events at a predetermined ratio (here, equally divided), and the contribution rate of the cause event to the result event is calculated.

  In general, the causal event Ei is not always an exclusion event, and when A = E1∪... En, the contribution rate PA (Ei) for the event Ei is the result event A occurs. In addition, the ratio of occurrence due to the contribution of the cause event Ei is formulated so that the total value for each cause event Ei is 1 (100%).

Furthermore, as an example, in the case of event A = E1∪E2 、 E3,
P (A) = P (E1) + P (E2) + P (E3) -P (E1) P (E2) -P (E1) P (E3) -P (E2) P (E3) + P (E1) P ( E2) P (E3)
PA (E1) = (1-P (E2) / 2-P (E3) / 2 + P (E2) P (E3) / 3) P (E1) / P (A)
PA (E2) = (1-P (E1) / 2-P (E3) / 2 + P (E1) P (E3) / 3) P (E2) / P (A)
PA (E3) = (1-P (E1) / 2-P (E2) / 2 + P (E1) P (E2) / 3) P (E3) / P (A)
PA (E1) + PA (E2) + PA (E3) = 1
It becomes.

In general, when A = E1∪ ... ∪En,
P (A) = Σ _{i = 1... N} P (Ei) + (− 1) ¹ Σ _{i, jε1... N, i <j} P (Ei) P (Ej) +. (-1) ^n-2 Σ _{i, j,..., Kε1... N, i <j <... <K} P (Ei) P (Ej). -1) ^n-1 _{ｉ i = 1... N} P (Ei)
PA (Ei) = (1-(− 1) ¹ Σ _{i, jε1... N, i <j} P (Ei) P (Ej) / 2 +... + (− 1) ⁿ⁻² Σ _{j ,..., Kε1... N, j <... <K} P (Ej)... P (Ek) / (n−1) + (− 1) ⁿ⁻¹ _{ｉ i = 1 · ..NP} (Ei) / n) P (Ei) / P (A)
Given in.

  According to the formula described above, if the occurrence probability of each terminal event (P (Ei)) is known, the contribution rate PA (Ei) of each terminal event to the result event is calculated by using the probability. be able to. However, the probability of occurrence may not be known for all terminal events. In such a case, a method of calculating the structural importance of the end event from the structure of the fault tree and a method of calculating the probability importance with the prior probabilities of all the end events being equal can be used.

First, the structural importance will be described.
FIG. 35 is a table corresponding to the fault tree of FIG. 34, and is a table used when calculating the structural importance.

In this table, the top event refers to the top event F in FIG. Then, the structural importance level Is (Fi) of the event Fi is obtained by the following equation.
Is (Fi) = ns (Fi) / 2 ^n-1 (n is the number of end events Fi)
Here, ns (Fi) is the number of times the top event F changes from 0 to 1 when the end event Fi changes from 0 to 1 (the other end events Fj (i ≠ j) do not change).

  For example, for event F1, state numbers 1 to 5, 2 to 6, 3 to 7 satisfy this condition, and are given by ns (F1) = 3. Therefore, structure importance level Is (F1) = 3/4 It becomes. Similarly, Is (F2) = Is (F3) = 1/4. The expected result is that the importance of the event F1 reaching the top event F without going through an AND gate is high.

Next, the probability importance will be described.
The probability importance is a quantified degree of influence of the i-th cause event Fi on the top event F. The probability importance ΔF (Fi) of the event Fi is given by the following equation.

ΔF (Fi) = F (Fi = 1) −F (Fi = 0)
Here, the occurrence probabilities of the end (cause) events Fj (i ≠ j) other than the end (cause) event Fi for calculating the importance are all set to predetermined values (here, 0.1). The importance ΔF (Fi) is calculated.

For example, for the event F1, F (F1 = 1) is the occurrence probability of the top event F when the occurrence probabilities of the end events F1, F2, and F3 are 1, 0.1 and 0.1, respectively. F (F1 = 0) is the occurrence probability of the top event F when the occurrence probabilities of the end events F1, F2, and F3 are 0, 0.1, and 0.1, respectively. Thus, the probability importance ΔF (F1) of the end event F1 is given by the following equation.
ΔF (F1) = F (F1 = 1) −F (F1 = 0) = 1− (0.1 × 0.1) = 0.99 (72.3%)
Similarly, ΔF (F2) and ΔF (F3) are given by the following equations.
ΔF (F2) = F (F2 = 1) −F (F2 = 0) = 0.1 + 0.1 × 1− (0.1 × (0.1 × 1)) = 0.19 (13.9%)
ΔF (F3) = F (F3 = 1) −F (F3 = 0) = 0.1 + 1 × 0.1− (0.1 × (1 × 0.1)) = 0.19 (13.9%)
Note that the cause (factor) candidate list is not necessarily output in the order of probability corresponding to each terminal event as a method of outputting the cause (factor) candidate list in the above-described defect cause search support process.

FIG. 36 is a diagram showing an output example of a cause (factor) candidate list in the cause seeking support process.
The fault tree shown in FIG. 36A has end events E1, E2, E3, E4 and a result event Y. These relationships are given by the following logical expression.
Y = (E1 + E2) · E3 + E4
= E1 / E3 + E2 / E3 + E4
Each term in the above expression in which the result event Y is represented by the logical sum of each term not including the logical sum (+) may be associated with each item in the cause (factor) candidate list. Referring to FIG. 36 (b), the degree of influence is listed in order of probability for each event E4, E1 · E3, E2 · E3.

FIG. 37 is a diagram for specifically explaining the failure (failure) influence analysis function.
In the expression model shown in FIG. 37, the unit 1 is composed of a part 1-1 and a part 1-2, which has product structure information and manufacturing process information as described above. Is omitted (since it is a repetition of the above).

  As described above, in the manufacturing process, a transition condition including a hazard mixture from the previous state and the previous state to the current state is involved in the malfunction occurring in the current state. For example, in the figure, in the component 1-1, the state 2 can be the cause of the failure for the state 3, and the state 1 can be the cause of the failure for the state 2. That is, the direction from the state 3 to the state 2 or the state 1 indicates a direction in which the cause of the malfunction that has occurred in the state 3 is pursued. On the other hand, the direction from state 2 or state 1 to state 3 is the direction in which a potential hazard (hazard) that can cause such a failure propagates, or the failure that occurred in states 1 to 3 It shows the direction of propagation to influence others. The direction of the arrow labeled “Propagation by state transition” in the figure matches the propagation direction of this hazard.

  On the other hand, an interface has been set for the component (the interface is set from the information processing terminal by the user), and the component is subjected to stress, and the stress is within the allowable range (resistance, strength) of the component. If it exceeds, a problem will occur. In the figure, it is assumed that a failure has occurred in the component 1-1. For this reason, the direction of the arrow written as “propagation by interaction (interface)” in the figure is from the component 1-1 to the component 1-2 (that is, the component affected by the failure occurring in the component 1-1). Heading to

  Therefore, when a failure occurs in a certain part, the influence on the stress of other parts that have the interface with that part is dominant, or the influence on the next manufacturing process of the part (using) that the failure has occurred. It is necessary to decide first whether it is dominant. In the expression in the figure, it is first necessary to determine whether the interaction branch or the state transition branch.

The outline of the failure (failure) impact analysis (support) process will be described below.
First, an influence path of a possible failure (failure) mode is identified. In other words, for each entity, the state transition of the entity along the hazard propagation path and the failure (failure) modes that can occur due to the interaction given by other entities are comprehensively defined and given from each influence source. The degree of influence (impact degree) and the possibility of the occurrence of the failure (failure) mode are identified probabilistically and systematically. As a result, an influence path is created.

Then, the importance of these possible failure modes and the actual occurrence frequency are further taken into consideration.
Furthermore, systematically describe the effect on the strength (what and how much) due to the state transition, and the effect (what and how much) on the stress due to the interaction given by other entities.

Next, the influence analysis of the failure (failure) mode that occurs is performed. That is, the hazard propagation path is analyzed in the forward direction (forward) from the failure (failure) occurrence site.
First, it is probabilistically determined whether the influence is caused by a decrease in strength, only by a deterioration in stress, or by both a decrease in strength and a deterioration in stress from a malfunction (failure) phenomenon.

When the probability of strength reduction is high, an emphasis is placed on the state transition branch, and the effects (results) with high probability that can occur in the next manufacturing process of the entity are listed (in descending order of probability).
When the probability of worsening stress to other entities is high, emphasis is placed on interaction bifurcation, and the effects (results) with high probability that can occur in other entities are listed (in descending order of probability).

  In both cases, a list of both high-probability effects (results) that can occur in the next manufacturing process of an entity and high-probability effects (results) that can occur in other entities (in descending order of probability). Up.

In addition, in the case of an assembly process in which state transition occurs from an entity when the probability of worsening stress to other entities is high, or when the probability of strength reduction is high, other entities or entities where state transitions occur are defective ( Assuming the failure mode, the above-described failure (failure) mode effect analysis is repeated.

  The failure (failure) influence analysis (support) function will be described with reference to the flowchart of FIG. This flow corresponds to the flow of FIG. 29 describing the defect (failure) cause pursuit function and describes differences in the impact analysis. Note that the processing corresponding to the flow of FIG. 28 that describes the defect (failure) cause pursuit function is the same in the impact analysis, and is therefore omitted. In the following description, differences from FIG. 29 will be mainly described.

First, the changes in the interaction branch are as follows.
Even when the defect influence analysis process is performed, the processing from step S101 to step S110 is the same as the defect cause pursuit process of FIG.

  In step S211 following step S110, the stress event (right side) related to the failure mode (attribute value) specified in step S110 (that is, the stress event that the portion where the failure occurred is output to another portion having an interface) To extract.

  For this reason, when the strength corresponding to each of the stress events extracted in step S211 is extracted in step S112, which is the next step after step S211, the strength is stressed through the interface from the site where the failure occurred. It becomes the attribute of the other part that receives.

Next, the changes in the state transition branch are as follows.
In step S217, which is the next step after step S115 in FIG. 29, an influence result can be obtained except for the failure mode related to the stress (right side) extracted in step S211 based on the failure mode identified in step S110. The remaining events (right side) are extracted, and the degree of influence (probability) of each extracted event from the target failure mode is calculated from the logic related to the target failure mode (such as FTA logic) and the occurrence probability of the end event. To do.

  Then, in step S218, the event extracted in step S217 is set as the influence result (candidate) (target defect), and the corresponding influence degree (probability) calculated (in step S217) is used as the influence result (the influence). The influence candidate list is displayed as a list on the display of the information processing terminal so as to be associated with the influence (the influence candidates) as the probability that the candidate is true. Alternatively, the contents corresponding to the list display result are stored in the QA2-KDB. Note that this list display also indicates the range of influence, and is also referred to as “a list of influence ranges and influence degree candidates (in order of probability)”.

  In the case of performing the impact analysis, the fault tree is advanced in the opposite direction from the case of performing the cause search. That is, in the impact analysis, the failure tree is analyzed from the end event to the top event. In this case, the failure mode corresponding to the entity in which the failure has occurred is set to an occurrence probability of 1, and analysis is performed on the failure tree. For example, if the failure mode is one of the inputs of the OR gate, the output of the OR gate is 1, and if the failure mode is one of the inputs of the 2-input AND gate, the AND gate The output is the other input value.

  In the above description, the logic added to the entity / state data model (QA2-KDB) is the SSM or FTA logic, but other logics may be added.

  In the following, FMEA logic is considered as this additional logic. The FMEA logic is logic used for failure prevention support. Therefore, the risk inherent in the product (more broadly, the entire production system related to the product) before the failure occurs, rather than thinking about countermeasures for the failure after the actual failure has occurred. Factors are calculated as risk priorities, and the risk priorities are used to support failure prevention.

FIG. 39 is a flowchart of the failure prevention support process.
In FIG. 39, in step S301, the user inputs the product line name targeted at this time from the information terminal device and starts FMEA. In response to the input of the system name, in step S302, a master table group stored as data for all product systems, a master table of products corresponding to the system name input from the QA2-KDB group, QA2-KDB is selected (target product line is squeezed out).

Subsequently, the following steps S303 to S305 are repeated for all entities / processes of the product selected in step S302.
That is, in step S303, the risk priority number (risk priority) of each failure mode in the target entity / process is calculated from the FMEA method of the target entity / process. FIG. 40 is a diagram for explaining this situation. In the figure, n failure modes (failure mode 1 to failure mode n) are included in the entity / process targeted at this time. In each failure mode, the risk priority number of the failure mode is calculated based on the following formula.
Risk priority number = occurrence frequency x influence degree x detection difficulty level Here, the definition of each item is
・ Occurrence frequency: Number of times that the failure occurs within a predetermined period (for example, within one month) ・ Influence level: Degree of influence that the production system receives when the failure occurs ・ Detection difficulty level… Given by the difficulty level of detecting a defect.

  In step S304, the risk priority numbers calculated in step S303 are sorted in order of magnitude and stored in a table. This table lists the risk priority numbers of each failure mode for each entity / process in order of magnitude, and is called a risk priority number (larger order) list of each failure mode in the entity / process.

  Subsequently, in step S305, each failure mode of the substance / process processed in step S304 is sorted in order of magnitude according to the corresponding risk priority number and stored in a table. This table lists the risk priority numbers of each failure mode in all systems in order of magnitude, and is called a risk priority number (large order) list of each failure mode in all systems.

  When the processing of step S303 to step S305 is performed for all entities / processes of the target product, the process proceeds to step S306. In step S306, one of the items is selected by the user from the items "substance / process" and "system batch" displayed on the display of the information processing terminal.

  In subsequent step S307, it is determined whether or not the item selected in step S306 is "system batch". If “system batch” is selected, the process proceeds to step S308, and the risk priority number list of each failure mode in all the systems described above is output to the output device. On the other hand, if “substance / by process” is selected, the process proceeds to step S309.

  In step S309, the user determines whether to display by entity / process. If not, a series of failure prevention support processing ends. On the other hand, in the case of performing, in step S310, a target entity / process on the display is selected by the user from the part master table of the target product line.

  In subsequent step S311, the risk priority number list of each failure mode in the entity / process corresponding to the selected entity / process is output to the output device based on QA2-KDB of the target product line. In step S312, the user selects whether to display another entity / process. If so, the process returns to step S310 and the above process is repeated. If not displayed, the series of failure prevention support processing is terminated.

Note that the list output in steps S308 and S311 is also called a countermeasure candidate list.
FIG. 41 is a diagram for explaining the maintenance function of the countermeasure candidate list described above.

41, countermeasure candidate lists are listed in the order of probability as output results in steps S308 and S311 of FIG.
In step S31, the user determines whether there is an effective measure corresponding to the listed measures. If there is an effective measure (YES in S31), the effective measure is selected in step S32 and registered in the attribute “measure” in QA2-KDB24. If there is no effective countermeasure (NO in S31), it is determined in step S33 by the user whether or not the effective countermeasure is in the master table. If it is in the master table (YES in S33), the process proceeds to S34, and an effective countermeasure is selected from the master table and registered in the attribute “Countermeasure” in the QA2-KDB 24. If it is not in the master table (NO in S33), the process proceeds to S35, a necessary countermeasure (sentence) is added to the master table, and then the process proceeds to step S34. An effective countermeasure is selected from the master table, and QA2 -Register in the attribute "Countermeasure" in KDB24.

  FIG. 42 is a diagram illustrating an output example of a candidate list as a result of the above-described defect cause pursuit process, defect effect analysis process, and defect prevention support process. In the figure, a cause (factor) candidate list is used for investigating the cause of a defect, and an influence range (result) and influence candidate list are used for grasping the influence of a defect. Each countermeasure candidate list is output (sorted in the order of probability (contribution rate)).

FIG. 43 is a diagram showing an example of an entity / state data model (corresponding to a certain entity).
FIG. 44 is a diagram showing a defect cause coverage table used for description of the entity / state data model of FIG. In the present embodiment, among the attributes of the entity / state data model, the cause attribute describes an attribute value using a defect cause coverage table.

  For example, among the attributes on the left side of the entity / state data model in FIG. 43, the processing manufacturing method (defect) is a cause attribute, and the attribute value is described using the defect cause coverage table in FIG. Note that the expression “processing and manufacturing method (defect)” has two pieces of information. One is information that the attribute “processing and manufacturing method” is the attribute value “...”, And the other is information that the cause attribute is the attribute value “processing and manufacturing method (defective)”.

  FIG. 45 is a diagram showing a defective part coverage table used for description of the entity / state data model of FIG. For example, the attributes “rim part”, “bubb part”, and “spoke part” of the entity / state data model in FIG. 43 use the items “rim part”, “bubb part”, and “spoke part” in the defective part coverage table of FIG. is described.

  Note that the attributes in the entity / state data model (for example, as shown in FIG. 43) other than those described using the various defect coverage tables described above are used for design and manufacturing as reference and supplementary information. There are things that describe know-how information using design documents, standards, standards, etc., and link information to documents that have knowledge and know-how.

  The above description has been made using various data, but here, for reference, a mode in which these various data exist is shown. FIG. 46 is a diagram showing how the various data actually exist. As shown in the figure, the various types of data are broadly classified into data related to product design information, data related to process design information, and data related to production / manufacturing information. And, as for the form in which any data exists, any knowledge / know-how in human brain, computerized / non-electronic drawing / form information, database information, or any combination thereof Exist as.

  The present invention can be applied to design support for a production system such as a production assembly type production system.

It is a figure explaining the flow of the phase until a product is marketed in a production system. It is a figure which shows the work flow of the malfunction process corresponding to the production system of FIG. It is a block diagram which shows the structure of the malfunction countermeasure assistance system in the production system of one Embodiment of this invention. It is a figure which shows the conceptual structure of PPMDB. It is a figure which shows the table corresponding to the representation model (PPMDB) of FIG. It is a figure which shows an example which set the hazard and the interface information with respect to the representation model (PPMDB) of FIG. It is a figure which shows the representation model of the unit comprised from four components. It is a figure which shows the various coverage tables memorize | stored in the malfunction coverage table storage part of FIG. It is a figure which shows an example of the malfunction (failure) site | part coverage table corresponding to a certain product type | system | group. It is a figure which shows an example of the malfunction (failure) mode coverage table (human error base) corresponding to a certain product type | system | group. It is a figure which shows an example of the malfunction (failure) mode coverage table (physical phenomenon base) corresponding to a certain product type | system | group. It is a figure which shows an example of the malfunction (failure) phenomenon coverage table (human error base) corresponding to a certain product type | system | group. It is a figure which shows an example of the malfunction (failure) phenomenon coverage table (physical phenomenon base) corresponding to a certain product system. It is a figure which shows an example of the malfunction (failure) result coverage table corresponding to a certain product type | system | group. It is a figure which shows an example of the malfunction (failure) cause coverage table (human error base) corresponding to a certain product type | system | group. It is a figure which shows an example of the malfunction (failure) cause coverage table (physical phenomenon base) corresponding to a certain product system. It is a figure which shows an example of the malfunction (failure) countermeasure coverage table (human error base) corresponding to a certain product system. It is a figure which shows an example of the malfunction (fault) countermeasure coverage table (physical phenomenon base) corresponding to a certain product system. It is a figure which shows an example which set the hazard on QA2-KDB24 produced | generated based on PPMDB of FIG. It is a flowchart which shows the process which produces | generates the framework of QA2-KDB24 based on PPMDB22 of FIG. It is a figure which shows an example of the data structure corresponding to a product model. It is a figure which shows an example of the data structure corresponding to a product process model. It is a figure which shows the place which added the logic with respect to the frame | skeleton of QA2-KDB24 of the produced | generated FIG. It is a figure explaining an example of the function of the registration part of FIG. It is a figure explaining the flow of the whole process which the malfunction countermeasure assistance part of FIG. 3 performs with a simple flow. It is a figure explaining a defect cause pursuit support function. It is a figure explaining a malfunction (failure) cause pursuit (support) function more concretely. It is a flowchart (the 1) of the cause pursuit (support) process of a malfunction (failure). It is a flowchart (the 2) of the cause pursuit (support) process of a malfunction (failure). It is a figure explaining a mode that a cause (factor) candidate list | wrist is output in the cause pursuit assistance process (target defect). It is a figure which shows the construction example of QA2-KDB corresponding to the unit which comprises a conducting plate and a base metal as components. It is the figure which showed the entity / state data model of the component 1-1 (conductor plate) which comprises the unit 1 of FIG. 31 with the SSM method (logic) applied to the component 1-1. It is the figure which showed the entity / state data model of the component 1-1 (conductor plate) which comprises the unit 1 of FIG. 31 with the FTA method (logic) applied to the component 1-1. It is a figure which shows an example of the failure tree as a target which calculates occurrence probability (contribution rate). It is a table corresponding to the fault tree of FIG. 34, and is used when calculating the structure importance. It is a figure which shows the example of an output of the cause (factor) candidate list | wrist in a cause pursuit assistance process. It is a figure explaining a malfunction (failure) influence analysis function concretely. It is a flowchart of a malfunction (failure) influence analysis process. It is a flowchart of a trouble prevention support process. It is a figure which shows the FMEA method (logic) applied to a certain target product system | strain. It is a figure explaining the maintenance function of a countermeasure candidate list. It is a figure which shows the candidate list output as a result of various malfunction countermeasure assistance processing. It is a figure which shows an example of an entity / state data model. FIG. 44 is a diagram showing a defect cause coverage table used for description of the entity / state data model of FIG. 43. FIG. 44 is a diagram showing a defective part coverage table used for description of the entity / state data model of FIG. 43. It is a figure which shows the method of the actual existence of various data in this embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Registration part 20 Knowledge and know-how accumulation part 21 Defect coverage table storage part 21a Defect site coverage table 21b Failure mode coverage table 21c Failure phenomenon coverage table 21d Failure result coverage table 21e Failure cause coverage table 21f Failure countermeasure coverage table 22 PPMDB
23 Knowledge Data Storage Unit 24 QA2-KDB
30 Reference section 40 Defect countermeasure support section

Claims (7)

In a device that supports countermeasures against malfunctions that occur in a production system related to products produced from one or more entities,
First fault coverage storage means for storing information indicating a comprehensive set of fault phenomena that may occur in the entity ;
A second fault coverage storage means for storing information indicating an exhaustive set of failure modes obtained by organizing a defect phenomena stored in the first fault coverage storage means for each form,
Product / process model storage means for storing model information representing product structure information and manufacturing process information associated with each other as a product / process model ;
Knowledge data storage means for storing knowledge data consisting of data on specifications and standards, knowledge on manufacturing methods, knowledge on production systems, knowledge on know-how, etc.
In advance,
Defect phenomenon indicating the comprehensive set stored in the first and second fault coverage storage means, failure mode, and the user various data such knowledge and know-how stored in the knowledge data storage means Is read out and referred to, and the model information stored in the product / process model storage means is read out to have the model information as a framework, and the user is expanded into the model information for the framework. Logic that describes the attribute value corresponding to each attribute and makes a determination using various existing methods, and status information and status as attributes of the manufacturing process information stored in the product / process model storage means Means for constructing a quality knowledge database by adding information relating to the direction of transition ;
When a predetermined malfunction occurs, the constructed quality knowledge database is read, and a plurality of attribute data is associated with an entity existing in the model information, and at least a part of the associated attribute data is used. First problem countermeasure support means for calculating the probability of occurrence of a defect phenomenon in the entity and providing countermeasure support by pursuing the cause and effect analysis of the defect;
Before the predetermined failure occurs, the constructed quality knowledge database is read out, and a plurality of attribute data is associated with an entity existing in the model information, and at least a part of the associated attribute data is used. A risk factor inherent in the product using the entity is calculated as a risk priority, a second failure countermeasure support means for using the risk priority for failure prevention support,
Probability candidate list for countermeasure support by pursuing the cause of the defect of the first defect countermeasure support means, and influence range and influence candidate list for countermeasure support by the impact analysis of the defect, respectively Sorting and outputting in order of (contribution rate), while outputting the countermeasure candidate list sorted in order of probability (contribution rate) for countermeasure support by the above-mentioned defect prevention support of the second defect countermeasure support means Output means;
A trouble countermeasure support apparatus in a production system, comprising: The existing method is an FTA method,
The first defect countermeasure support means is:
Means for storing FTA information in which the failure mode associated with the predetermined failure phenomenon is arranged in each part of the tree structure via a logic gate for an entity in which the predetermined failure phenomenon has occurred;
The tree structure of the FTA information is obtained by proceeding in a predetermined direction from the start point of the analysis using the entity where the predetermined defect phenomenon has occurred as a start point of the analysis, thereby generating a probability of occurrence of a failure mode located in each part of the tree ( The failure countermeasure support apparatus according to claim 1 , further comprising means for calculating a contribution ratio) based on the FTA method . Based on the FTA information, when the existing FTA method calculates the occurrence probability (contribution rate) of the event,
When the tree structure indicates a structure in which a result event is derived from a logical sum of a plurality of cause events that are not mutually exclusive events, the occurrence probability of the result event is represented by the occurrence probability of the plurality of cause events. In the formula, the term constituted by the product of two or more occurrence probabilities of the plurality of cause events is distributed to the two or more occurrence probabilities of the plurality of cause events at a predetermined ratio, and the cause events the troubleshooting support device according to claim 2 Symbol mounting, characterized in that it comprises means for calculating a contribution to the result events. 3. The trouble countermeasure support apparatus according to claim 2, wherein the starting point of the analysis is an entity in which a defect phenomenon has occurred, and the entity has an interaction with another entity through an interface. Which of the other entities that influence the entity through the interface and the manufacturing process corresponding to the entity has a larger contribution as a cause of the occurrence of a predetermined malfunction occurring in the entity Further comprising means for calculating,
When the cause contribution calculation unit calculates that the influence from the manufacturing process corresponding to the entity is larger, the occurrence probability (contribution rate) calculation unit calculates the occurrence of the manufacturing process corresponding to the entity. 5. The failure countermeasure support apparatus according to claim 4, wherein an occurrence probability (contribution rate) is calculated for a manufacturing process corresponding to the entity based on FTA information. Calculate which of the contributions to the given malfunction phenomenon that occurred in the entity is larger in the other entities that the entity influences through the interface and the assembly process in which the entity is incorporated. Further comprising means for
When the influence contribution calculating means calculates that the influence on the other entity is larger, the occurrence probability (contribution rate) calculating means is based on the FTA information corresponding to the other entity. The failure countermeasure support apparatus according to claim 4, wherein an occurrence probability (contribution rate) is calculated for the other entity. The existing method is FMEA,
The second problem countermeasure support means is:
The failure countermeasure support apparatus according to claim 1, wherein the risk priority of the failure mode associated with the entity is calculated based on the FMEA method for an arbitrary entity constituting the product.