Failure mode effect analysis management system, failure mode effect analysis management method and information processing system

Abstract

To provide a failure mode effect analysis management system or the like capable of managing a risk in business that changes from moment to moment by enabling an FMEA to be calculated even without the help of a skilled person or the like.SOLUTION: A failure mode effect analysis management system 210 include: a failure mode analysis part 310 for discriminating a failure mode representing a content of defect from a past failure mode effect analysis result and a defect factor list; an occurrence degree calculation part 320 for calculating an occurrence degree of defect for the failure mode from the past failure mode effect analysis result and the defect factor list; an effect degree calculation part 330 for calculating an effect degree when defect occurs for the failure mode from the past failure mode effect analysis result and work result data; a detection degree calculation part 340 for calculating a detection degree of the defect for the failure mode from the past failure mode effect analysis result and the defect factor list; and an FMEA update part 350 for updating a failure mode effect analysis result for the failure mode.SELECTED DRAWING: Figure 9

Description

Technical Field

[0001] The present invention relates to a failure mode effect analysis management system, a failure mode effect analysis management method, and an information processing system. In particular, the present invention relates to a failure mode effect analysis management system and the like that perform analysis when a defect occurs in a product produced in the manufacturing industry.

Background Art

[0002] For example, in each operation in the manufacturing industry, when a defect occurs within the operation, there is a significant risk that the production line will stop. In the manufacturing industry, failure mode and effects analysis (FMEA) is one method for reducing this risk. FMEA classifies the components of a process according to failure modes and calculates the magnitude of the risk of the classified failure modes using the Risk Priority Number (RPN). Generally, RPN is calculated as Severity × Occurrence × Detection, and failure modes with larger numerical values are dealt with preferentially. However, the definition of RPN may add a certain variable or delete a certain value depending on the purpose.

[0003] Patent Document 1 discloses an example of automatically performing a risk assessment of a certain production line using FMEA. Here, RPN as a risk assessment index is defined as Severity × Occurrence.

[0004] Patent Document 2 discloses a risk assessment device that outputs an expected non-loss value as information used for risk assessment. This risk assessment device has an input unit for inputting frequency, potentiality, severity, and impact for each failure mode. It also has a first calculation unit that reads out from a database the frequency probability corresponding to the input frequency, the potential probability corresponding to the potentiality, and the severity probability corresponding to the severity, and calculates the probability of an accident occurring per unit period. Furthermore, based on the probability of an accident occurring per unit period, it has a second calculation unit that calculates the unexpected loss from the value indicated by the probability distribution of an accident occurring during the measurement period and the value of the loss amount. And it has an output unit that outputs the calculated unexpected loss as information used for risk assessment.

[0005] Patent Document 3 discloses an earthquake risk assessment system for assessing the earthquake risk of a predetermined assessment target. In this earthquake risk assessment system, the assessment of the earthquake risk of a predetermined assessment target reads information about the analysis object from the storage unit, and calculates the response output when the design input waveform simulating the earthquake motion input to the analysis object is input to the object according to a predetermined calculation formula stored in the storage unit. Next, based on the calculated response output and the characteristics related to the seismic resistance of the analysis object, with at least the response output among the response output and the seismic resistance related characteristics as a random variable, the probability of a damage mode occurring in the analysis object is calculated according to a predetermined arithmetic formula stored in the storage unit. Furthermore, based on the calculated damage probability and the information about the impact degree stored in the storage unit, a risk assessment result is created, and the created risk assessment result is output through the output unit.

Prior Art Documents

Patent Documents

[0006]

Patent Document 1

Patent Document 2

Patent Document 3

Summary of the Invention

Problems to be Solved by the Invention

[0007] Conventionally, FMEA has generally been created based on the knowledge of experts at the time of creation. However, with this method, it is difficult to sequentially update the FMEA for the risks within the business that are changing moment by moment. Also, with the conventional method, when a new failure mode (a failure mode not considered at the time of initial creation) occurs, it is difficult to detect, and it is difficult to respond quickly when a new failure mode occurs. An object of the present invention is to provide a failure mode effect analysis management system, a failure mode effect analysis management method, and an information processing system that can calculate FMEA without the intervention of experts or the like and can manage the risks within the business that are changing moment by moment.

Means for Solving the Problems

[0008] To solve the above problems, the present invention provides a failure mode analysis unit that discriminates a failure mode representing the content of a defect from the results of past failure mode effect analysis and a defect cause list that is a history when a defect occurs, a frequency calculation unit that calculates the frequency of occurrence of a defect for a failure mode from the results of past failure mode effect analysis and the defect cause list, an impact calculation unit that calculates the impact when a defect occurs for a failure mode from the results of past failure mode effect analysis and work performance data representing the work performance, a detectability calculation unit that calculates the detectability of a defect for a failure mode from the results of past failure mode effect analysis and the defect cause list, and an update unit that updates the results of the failure mode effect analysis for a failure mode. In this case, a failure mode effect analysis management system that can calculate FMEA without the intervention of experts or the like and can manage the risks within the business that are changing moment by moment can be provided.

[0009] Here, for example, the failure mode analysis unit discriminates the failure mode by the distance between vectors after converting the defect content described in the defect cause list for the failure mode into vectors. In this case, the determination of the failure mode can be performed more accurately. Also, for example, the occurrence degree calculation unit calculates the occurrence degree from the defect occurrence date and time information described in the defect cause list for the failure mode. In this case, the calculation of the occurrence degree becomes easy. Furthermore, for example, the occurrence degree calculation unit calculates the occurrence degree by either a method of updating based on a predefined occurrence degree table based on the defect occurrence date and time information, or a method of relatively updating from the defect occurrence frequencies of all failure modes. In this case, the user can select the method for calculating the occurrence degree. Moreover, for example, the impact degree calculation unit calculates the contribution rate of each process regarding the target variable using machine learning from the work performance data, and sets the calculated contribution rate as the impact degree of each process. In this case, the impact degree can be obtained by calculating the contribution rate. And, for example, the impact degree calculation unit calculates the impact degree based on at least one target variable determined by the user. In this case, the FMEA can be calculated according to the target variable selected by the user. Also, for example, the detectability calculation unit calculates the detectability based on the defect detection method information described in the defect cause list for the failure mode. In this case, the detectability can be calculated based on the information on the actual detection of the defect, and the accuracy of the detectability is improved. Furthermore, for example, the detectability calculation unit calculates the detectability by either a method of referring to and updating the most recent defect cause list based on a predefined detectability table based on the defect detection method information, or a method of referring to all defect cause lists and updating with the average value. In this case, the FMEA can be calculated according to the method selected by the user. Furthermore, for example, the update unit notifies the user of the failure mode determined by the failure mode analysis unit, the occurrence degree calculated by the occurrence degree calculation unit, the impact degree calculated by the impact degree calculation unit, and the detection degree calculated by the detection degree calculation unit. When the user selects to update, the update unit updates the result of the failure mode impact analysis for the failure mode. In this case, by involving the user, the accuracy of FMEA can be ensured. Also, for example, the failure mode impact analysis is updated each time a list of root causes of defects is created. In this case, each time a defect occurs, FEMA is updated, enabling countermeasures against risks within the ever-changing business operations at all times.

[0010] In addition, in the present invention, by a processor executing a program recorded in a memory, from the results of past failure mode impact analysis and a list of root causes of defects, which is a history when a defect occurs, a failure mode representing the content of the defect is discriminated. From the results of past failure mode impact analysis and the list of root causes of defects, the occurrence degree of the defect for the failure mode is calculated. From the results of past failure mode impact analysis and work performance data representing the work performance, the impact degree when a defect occurs for the failure mode is calculated. From the results of past failure mode impact analysis and the list of root causes of defects, the detection degree of the defect for the failure mode is calculated, and the result of the failure mode impact analysis for the failure mode is updated. In this case, FMEA can be calculated without the intervention of experts, etc., and a failure mode impact analysis management method that can manage risks within the ever-changing business operations at all times can be provided.

[0011] Furthermore, the present invention includes a failure mode and effects analysis management system that performs failure mode and effects analysis, a failure mode and effects analysis generation device that sends the results of past failure mode and effects analysis to the failure mode and effects analysis management system, a root cause list generation device that sends a root cause list, which is a history when a defect occurs, and a work performance data generation device that sends work performance data representing work performance. The failure mode and effects analysis management system includes a failure mode analysis unit that discriminates a failure mode representing the content of a defect from the results of past failure mode and effects analysis and the root cause list, an occurrence degree calculation unit that calculates the occurrence degree of a defect for the failure mode from the results of past failure mode and effects analysis and the root cause list, an impact degree calculation unit that calculates the impact degree when a defect occurs for the failure mode from the results of past failure mode and effects analysis and the work performance data, a detection degree calculation unit that calculates the detection degree of a defect for the failure mode from the results of past failure mode and effects analysis and the root cause list, and an update unit that updates the results of failure mode and effects analysis for the failure mode. In this case, it is possible to provide an information processing system that makes it easy to manage the objects of failure mode and effects analysis by FMEA.

Advantages of the Invention

[0012] According to the present invention, it is possible to provide a failure mode and effects analysis management system, a failure mode and effects analysis management method, and an information processing system that can calculate FMEA without the intervention of a skilled person or the like and can manage the risks within the business that change every moment.

Brief Description of the Drawings

[0013]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Embodiments for Carrying Out the Invention

[0014] Hereinafter, with reference to the accompanying drawings, embodiments of the present invention will be described in detail. Here, the present invention will be described with reference to the first embodiment to the third embodiment. Note that the embodiments described below do not limit the invention related to the claims, and not all of the various elements and combinations thereof described in the embodiments are essential for the solution means of the invention.

[0015] 〔First Embodiment〕 <Overall Explanation of the Operation of the Information Processing System 1> FIG. 1 is a diagram showing an overview of an information processing system 1 according to the first embodiment. Here, the information processing system 1 includes a manufacturing site 100 and a computer 110, and shows that the computer 110 performs a failure mode effect analysis based on the information obtained from the manufacturing site 100. At the manufacturing site 100, products to be subject to FMEA are manufactured. The products manufactured at the manufacturing site 100 are not particularly limited, and examples include automobiles, machine tools, household electrical appliances and their parts, as well as chemical products, foods, and the like. The computer 110 acquires FMEA data 120 calculated so far, a list of defect causes 130 that describes defect content, defect causes, and improvement measures and is created as a defect log when a defect occurs in the business, and work performance data 140 accumulated at the manufacturing site and the like. Then, each time a defect occurs in the product and the list of defect causes 130 is created, the failure mode effect analysis management system 210 calculates new FMEA data 120 using the FMEA data 120, the list of defect causes 130, and the work performance data 140, and performs automatic update. Although details will be described later, the failure mode effect analysis management system 210 calculates and updates the FMEA data 120 by each functional unit of a failure mode analysis unit 310, an occurrence degree calculation unit 320, an impact degree calculation unit 330, a detection degree calculation unit 340, and an FMEA update unit 350. Then, the calculated FMEA data 120 is sent to the manufacturing site 100, and based on this, the FMEA administrator manages the risks that may occur in the product and takes measures against the occurrence of defects.

[0016] <Description of the functional configuration of the information processing system 1> FIG. 2 is a block diagram showing an example of the functional configuration of the information processing system 1 in the first embodiment. Here, the information processing system 1 is divided into the above-described manufacturing site 100 and the contractor 200. The contractor 200 operates and manages the information processing system 1. Here, the failure mode effect analysis management system 210 corresponds to the core part that calculates the FMEA by the computer 110 shown in FIG. 1. In FIG. 2, the failure mode effect analysis management system 210 is connected to the network 220. Further, connected to the network 220 are a work performance data generation device 230, an FMEA generation device (failure mode effect analysis generation device) 250, and a root cause list generation device 270, which are arranged at the manufacturing site 100. In addition, at the contractor 200, in addition to the failure mode effect analysis management system 210, a work performance data storage unit 240, an FMEA management unit 260, and a root cause list storage unit 280 are connected to the information collection system 290 via the network 220. Furthermore, connected to the failure mode effect analysis management system 210 is an operation terminal 291 for operating the failure mode effect analysis management system 210. The operation terminal 291 is used by various users such as an administrator who manages the failure mode effect analysis management system 210, a monitor who monitors, and a user who uses it, and is one or more information terminal devices. In the example of FIG. 2, it is illustrated as a configuration different from that of the contractor 200 by arranging the operation terminal 291 within the manufacturing site 100, but a part or all of the operation terminal 291 may be provided within the contractor 200. Also, by connecting the failure mode effect analysis management system 210 and the operation terminal 291 via the network 220, it may be configured to access the failure mode effect analysis management system 210 from an arbitrary location outside the contractor 200.

[0017] The failure mode effect analysis management system 210 includes a processor 211 such as a central processing unit (CPU) that performs overall control of the failure mode effect analysis management system 210, a storage device 212 that stores various processing programs and the like for realizing the functions of the failure mode effect analysis management system 210, a network interface (I / F) 213, and the like. The storage device 212 is realized using a known storage device such as a ROM (Read Only Memory) that stores various processing programs and the like, a RAM (Random Access Memory) that temporarily stores information, and a hard disk drive (HDD). By the processor 211 executing various processing programs stored in the storage device, each function of the failure mode effect analysis management system 210 is realized. Note that the configuration of the failure mode effect analysis management system 210 is not limited to the example shown in the figure, and a part or all of the program may be configured to be introduced from another device via a non-temporary storage medium or a communication line.

[0018] The work performance data generation device 230 is a device that monitors the work in the manufacturing site 100 and generates work performance data 140 indicating the progress status and progress results of the work. The work performance data 140 may be, for example, a barcode reader, a PC (Personal Computer), or a server that acquires the work log of the worker, or a machine that processes parts or assembles finished products, or a sensor that collects inspection information of RFID (Radio Frequency IDentifier) attached to parts or finished products. A large number of these work performance data generation devices 230 are provided at the manufacturing site, and the collected work performance data 140 is accumulated in a system that manages the manufacturing site 100. The accumulated work performance data 140 is transmitted to and managed by the work performance data storage unit 240 via the network 220.

[0019] The work performance data storage unit 240 is a storage device such as a server or a memory, and stores the work performance data 140 received from the work performance data generation device 230 via the network 220.

[0020] FIG. 3 is a diagram showing an example of the work performance data 140. Here, the work performance data 140 used in the automobile manufacturing industry is shown. The illustrated work performance data 140 has, as columns, the production number 141, and as the work performance of each process, press 142, welding 143, painting 144, engine manufacturing 145, assembly 146, final pre-shipment inspection result (inspection) 147, and lead time 148.

[0021] Returning to FIG. 2, the FMEA generation device 250 is a device that generates one or more pieces of FMEA data 120 managed in files in EXCEL or CSV format. The FMEA data 120 is transmitted via the network 220 to the FMEA management unit 260 for management.

[0022] The FMEA management unit 260 is a storage device such as a server or a memory, and centrally manages the FMEA data 120 received from the FMEA generation device 250 via the network 220.

[0023] FIG. 4 is a diagram showing an example of the FMEA data 120. Here, the FMEA data 120 used in the automobile manufacturing industry is shown. The illustrated FMEA data 120 has, as columns, process 121, part name 122, failure mode 123, severity 124, occurrence 125, detectability 126, RPN 127, and need for countermeasure 128.

[0024] In the automotive manufacturing industry, RPN127 is defined in a form such as an impact table that defines impact levels, an occurrence table that defines occurrence levels, and a detection table that defines detection levels. When initially constructing the FMEA, after subjecting the identified failure modes to brainstorming by experts, the values of each variable are determined while referring to the impact table, occurrence table, and detection table, and RPN127 is calculated.

[0025] Figure 5 is a diagram showing the impact table 500 that defines impact levels. The illustrated impact table 500 is premised on being used in the automotive manufacturing industry. And the impact table 500 defines, on a 10-point scale from 1 to 10, the phenomena that can occur due to a defect when a defect occurs in the vehicle as the product. In this case, the impact level indicates that 1, "no recognizable impact," is the lowest, and 10, "affects the safe operation of the vehicle," is the highest.

[0026] Figure 6 is a diagram showing the occurrence table 600 that defines occurrence levels. The illustrated occurrence table 600 is premised on being used in the automotive manufacturing industry. And the occurrence table 600 defines the frequency of occurrence of defects on a 10-point scale from 1 to 10. In this case, the occurrence level indicates that 1, "does not occur," is the lowest, and 10, "occurs every time," is the highest.

[0027] Figure 7 is a diagram showing the detection table 700 that defines detection levels. The illustrated detection table 700 is premised on being used in the automotive manufacturing industry. And the detection table 700 defines, on a 10-point scale from 1 to 10, the difficulty of detecting whether a defect can be detected during the manufacture of the vehicle as the product. In this case, the detection level indicates that 1, "always detectable," is the lowest, and 10, "undetectable," is the highest.

[0028] Returning to FIG. 2 again, the defect cause list generation device 270 is a device that acquires one or more defect cause lists 130 managed in files in EXCEL or CSV format. This defect cause list 130 is created as a defect log when a defect occurs within the business. The defect cause list 130 is, for example, a group of defect cause lists that describe the defect content, the defect cause, and the defect countermeasure. The created defect cause list generation device 270 is stored in the defect cause list storage unit 280 via the network 220.

[0029] The defect cause list storage unit 280 is, for example, a storage device such as a server or a memory, and stores the defect cause list 130 received from the defect cause list generation device 270 via the network 220.

[0030] FIG. 8 is a diagram showing an example of the defect cause list 130. Here, the defect cause list 130 used in the automobile manufacturing industry is shown. The illustrated defect cause list 130 has, as columns, the defect occurrence date and time 131, the defect discovery method 132, the defect occurrence process 133, the part name 134, the defect content details 135, the defect cause 136, and the defect treatment method 137.

[0031] <Detailed description of the functions and operations of the failure mode effect analysis management system 210> Next, the functions and operations of the failure mode effect analysis management system 210 will be described in detail. FIG. 9 is a block diagram for explaining the functions of the failure mode effect analysis management system 210. The failure mode effect analysis management system 210 realizes the functions of the failure mode analysis unit 310, the occurrence degree calculation unit 320, the influence degree calculation unit 330, the detection degree calculation unit 340, and the FMEA update unit 350 by a computer program. As described above, the failure mode effect analysis management system 210 is connected to the work performance data storage unit 240, the FMEA management unit 260, and the defect cause list storage unit 280 via the network 220.

[0032] Also, as described above, the failure mode effect analysis management system 210 is connected to the operation terminal 291. Here, the flow of information through the connection path is indicated by the direction of the arrow. The operation terminal 291 provides a function that allows the user to arbitrarily determine whether to update the FMEA data 120 from the notification screen described later, referring to the failure mode transmitted from the FMEA update unit 350 and the updated value of the RPN 127.

[0033] First, the failure mode analysis unit 310 of the failure mode effect analysis management system 210 refers to the failure mode 123 of the FMEA data 120 managed by the FMEA management unit 260 and the detailed defect content 135 of the defect factor list 130 managed by the defect factor list accumulation unit 280. Then, the failure mode analysis unit 310 determines whether the newly created failure mode of the detailed defect content 135 is a known failure mode or an unknown failure mode.

[0034] The occurrence degree calculation unit 320 calculates the occurrence degree by referring to the defect occurrence date and time 131 of the defect factor list 130 managed by the defect factor list accumulation unit 280 for the failure mode specified by the failure mode analysis unit 310. The occurrence degree calculation unit 320 allows the user to arbitrarily select a method of updating based on the predefined occurrence degree table 600 as shown in FIG. 6 and a method of relatively updating from the defect occurrence frequencies of all failure modes. Here, when the method of relatively updating from the defect occurrence frequencies of all failure modes is used, the user can arbitrarily select the calculation method from simple moving average, weighted moving average, and exponential smoothing moving average.

[0035] The severity calculation unit 330 calculates the severity by referring to the work performance data 140 managed by the work performance data accumulation unit 240 for the failure mode specified by the failure mode analysis unit 310. The severity calculation unit 330 adopts, for example, a method of updating based on the predefined severity table 500 as shown in FIG. 5. Here, the target variable of the severity can be arbitrarily set by the user. For example, the user can arbitrarily set columns such as the inspection 147 and lead time 148 in FIG. 3 as the target variable. It is also possible to set a plurality of columns as the target variable.

[0036] The detection degree calculation unit 340 calculates the detection degree by referring to the defect detection method 132 of the defect cause list 130 managed by the defect cause list storage unit 280 for the failure mode specified by the failure mode analysis unit 310. Since the detection degree calculation unit 340 adopts a method of updating based on the detection degree table 700 defined in advance as shown in FIG. 7, the user shall select the defect detection method from among the detection degree tables 700 defined in advance when describing the defect cause list. Note that the execution order of the occurrence degree calculation unit 320, the impact degree calculation unit 330, and the detection degree calculation unit 340 can be arbitrarily changed by the user.

[0037] The FMEA update unit 350 notifies the operation terminal 291 of the failure mode specified by the failure mode analysis unit 310, the occurrence degree calculated by the occurrence degree calculation unit 320, the impact degree calculated by the impact degree calculation unit 330, and the detection degree calculated by the detection degree calculation unit 340. Then, if the user operates on the notification screen described later on the operation terminal 291 and permits the update, the FMEA update unit 350 updates each variable value of the FMEA data 120 for the corresponding failure mode.

[0038] FIG. 10 is a sequence diagram showing an example of the operation procedure until the failure mode impact analysis management system 210 automatically updates the FMEA data 120. First, the user who uses the failure mode impact analysis management system 210 issues an instruction to add a defect cause list to the failure mode impact analysis management system 210 using the operation terminal 291 (S1001). The operation terminal 291 transmits the defect cause list 130 to the failure mode impact analysis management system 210 (S1002).

[0039] Next, the failure mode analysis unit 310 refers to the defect cause list 130 added by S1001 to S1002 and the FMEA data 120 transmitted from the FMEA management unit 260 (S1003). Then, the failure mode analysis unit 310 determines whether the added defect cause list 130 is a known failure mode or an unknown failure mode (S1004). It can also be said that the failure mode analysis unit 310 discriminates a failure mode representing the content of a defect from the results of past failure mode effect analysis (FMEA data 120) and the defect cause list 130 which is the history when a defect occurred.

[0040] Next, the failure mode effect analysis management system 210 refers to the defect cause list 130 added by the defect cause list addition instruction 1010 and the FMEA data 120 transmitted from the FMEA management unit 260 in the occurrence degree calculation unit 320 (S1005), and also refers to the defect cause list 130 transmitted from the defect cause list storage unit 280 (S1006). Then, the occurrence degree calculation unit 320 calculates the occurrence degree for the failure mode specified by the failure mode analysis unit 310 (S1007). It can also be said that the occurrence degree calculation unit 320 calculates the occurrence degree of a defect for a failure mode from the results of past failure mode effect analysis (FMEA data 120) and the defect cause list 130.

[0041] Next, the failure mode effect analysis management system 210 refers to the defect cause list 130 added by the defect cause list addition instruction 1010 and the FMEA data 120 transmitted from the FMEA management unit 260 in the impact degree calculation unit 330 (S1008), and also refers to the work performance data 140 transmitted from the work performance data storage unit 240 (S1009). Then, the impact degree calculation unit 330 updates the impact degree for the failure mode specified by the failure mode analysis unit 310 (S1010). It can also be said that the impact degree calculation unit 330 calculates the impact degree when a defect occurs for a failure mode from the results of past failure mode effect analysis (FMEA data 120) and the work performance data 140 representing the work performance.

[0042] Next, the failure mode effect analysis management system 210, in the detection degree calculation unit 340, refers to the defect cause list 130 added by the defect cause list addition instruction 1010 and the FMEA data 120 transmitted from the FMEA management unit 260 (S1011), and also refers to the defect cause list 130 transmitted from the defect cause list storage unit 280 (S1012). Then, the detection degree calculation unit 340 updates the detection degree for the failure mode specified by the failure mode analysis unit 310 (S1013). This can also be said that the detection degree calculation unit 340 calculates the detection degree of defects for the failure mode from the results of past failure mode effect analysis (FMEA data 120) and the defect cause list 130.

[0043] Next, the failure mode effect analysis management system 210, in the FMEA update unit 350, transmits the failure mode specified by the failure mode analysis unit 310, the occurrence degree calculated by the occurrence degree calculation unit 320, the impact degree calculated by the impact degree calculation unit 330, and the detection degree calculated by the detection degree calculation unit 340 to the operation terminal 291 (S1014).

[0044] The operation terminal 291 notifies the user of whether the defect cause list 130 added by the defect cause list addition instruction 1010 is a known failure mode, the occurrence degree calculated by the occurrence degree calculation unit 320, the impact degree calculated by the impact degree calculation unit 330, the detection degree calculated by the detection degree calculation unit 340, and the update content of the RPN127 (S1015). Actually, these pieces of information are displayed on the operation terminal 291. In this case, it can also be said that the FMEA update unit 350 notifies the user of the failure mode discriminated by the failure mode analysis unit 310, the occurrence degree calculated by the occurrence degree calculation unit 320, the impact degree calculated by the impact degree calculation unit 330, and the detection degree calculated by the detection degree calculation unit 340.

[0045] Next, the user determines whether to update the FMEA data 120 managed by the FMEA management unit 260 by operating the operation terminal 291 (S1016). As a result, when the user selects an update, an instruction to that effect is sent from the operation terminal 291 to the failure mode impact analysis management system 210 (S1017). Then, the failure mode impact analysis management system 210 updates the FMEA data 120 of the FMEA management unit 260 with the notified content in the FMEA update unit 350 (S1018). In this case, the FMEA update unit 350 functions as an update unit that updates the result of the failure mode impact analysis (FMEA data 120) for the failure mode. Also, it can be said that the FMEA update unit 350 updates the result of the failure mode impact analysis for the failure mode when the user selects to update. Note that when the user selects not to update, the FMEA data 120 is not updated. In this embodiment, although the FMEA can be automatically generated in principle, the accuracy of the FMEA can be ensured by involving the user's judgment during the update of the FMEA.

[0046] Next, the operations of the failure mode analysis unit 310, the occurrence degree calculation unit 320, the impact degree calculation unit 330, and the detection degree calculation unit 340 will be described in detail. <Detailed description of the operation of the failure mode analysis unit 310> FIG. 11 is a flowchart for explaining the operation of the failure mode analysis unit 310. First, the failure mode analysis unit 310 compares the defect occurrence process 133, the part name 134, and the defect content details 135 of the defect cause list 130 with the process 121, the part name 122, and the failure mode 123 of the FMEA data 120 to identify which process and part name failure mode in the FMEA data 120 the failure mode of the newly created defect cause list 130 is (S1110). Next, the failure mode analysis unit 310 converts the defect content details 135 of the newly created defect cause list 130 from natural language to a vector for the defect content details 135 of the defect cause list 130 accumulated in the defect cause list storage unit 280 for the failure mode identified in S1110 (S1120). Next, the failure mode analysis unit 310 categorizes the natural language vectorized in S1120 using the elbow method or the like to determine the number of categories (S1130). Then, the failure mode analysis unit 310 determines whether a new cluster has appeared (S1140). As a result, if a new cluster has appeared (Yes in S1140), the new cluster is determined to be a new failure mode (S1150). On the other hand, if a new cluster has not appeared (No in S1140), the new cluster is determined not to be a new failure mode (S1160).

[0047] FIG. 12 is a diagram for explaining the process in which the failure mode analysis unit 310 vectorizes natural language in S1130 of FIG. 11. Here, a case where the defect content detail 135 is the sentence "There is a scratch on the processed product" will be described. First, the failure mode analysis unit 310 tokenizes the sentence into sub-words (step 1). In this case, the failure mode analysis unit 310 shows that by tokenizing the above sentence, it has been sub-worded into 'processing', 'product', 'to','scratch', 'is', 'there', 'yes'. Next, the failure mode analysis unit 310 converts the sub-words into vectors (step 2). In this case, for example, it shows that 'processing' = array([-1.52316585e-01, -3.27842422e-02, ~ ]) shape=(300,), 'product' = array([-9.41298604e-02, 2.89820246e-02, ~]) shape=(300,) have been converted into such vectors. Furthermore, for all tokens, the failure mode analysis unit 310 defines the average value as the vector of the sentence, and aligns the dimensions of the defect content details 135 and the FMEA (step 3). In this case, the average values of 'processing', 'product', 'to', 'damage', 'exists', 'yes' are shown as ['processing', 'product', 'to', 'damage', 'exists', 'yes'] = array([ 3.00453138e-02, -1.01010511e-02, ~ ]) shape=(300,). Finally, the failure mode analysis unit 310 detects the node of the nearest past defect cause list based on the Cosine similarity, and discriminates the category based on the distance (step 4). In this case, it can be said that the failure mode analysis unit 310 discriminates the failure mode by the distance between vectors after converting the defect content described in the defect cause list 130 for the failure mode into a vector.

[0048] Also, in S1140 of FIG. 11, determining whether a new cluster has appeared in the failure mode analysis unit 310 is performed, for example, as follows. This is also the process of Step4 in FIG. 11. First, for each process and part name, map the Cosine similarity two-dimensionally (step 4-1). Next, detect which failure mode the newly created defect cause list 130 belongs to. This is determined by the distance index between the mapped nodes (step 4-2). Thus, it can be known whether it is an existing failure mode or a new failure mode. Then, keyword extraction is performed for each cluster for each failure mode, and the most frequently occurring keyword is set as the failure mode name (step 4-3). Note that the failure mode name can be arbitrarily changed by the user.

[0049] FIGS. 13(a) to (b) are diagrams showing the results of categorizing natural language in S1130 of FIG. 11. Among these, Fig. 13(a) shows the case where a new cluster appears (Yes in S1140). This illustrates the result of categorizing the failure modes when a new failure mode appears. In Fig. 13(a), it shows that there are existing failure modes 1310, 1320, 1330, and 1340 for "simulation", "scratch", "initial defect", and "clogging" respectively as existing failure modes, and represents that a failure mode 1350 of "deterioration" has newly appeared. Here, the failure mode analysis unit 310 may perform categorization using a topic modeling method such as BERTopic. Since BERTopic performs the processing of the flowchart in Fig. 11 all at once within the model, there is an advantage that the implementation becomes rapid. On the other hand, Fig. 13(b) shows the case where a new cluster does not appear (No in S1140). Here, the detailed defect content 135 is "There is a scratch on the processed product", indicating that this belongs to the existing failure mode 1320 having the failure mode name of "scratch".

[0050] <Detailed description of the operation of the occurrence degree calculation unit 320> Fig. 14 is a diagram for explaining the operation of the occurrence degree calculation unit 320. The occurrence degree calculation unit 320 calculates the occurrence degree by referring to the failure occurrence date and time 131 of the same defect cause list 130 as the failure mode specified by the failure mode analysis unit 310. In this case, it can also be said that the occurrence degree calculation unit 320 calculates the occurrence degree from the failure occurrence date and time information (in this case, the failure occurrence date and time 131) described in the defect cause list 130 for the failure mode. Here, the method for calculating the occurrence degree allows the user to arbitrarily select between the method of updating based on the predefined occurrence degree table 600 described in Fig. 6 and the method of updating relatively from the failure occurrence frequencies of all failure modes. In this case, it can also be said that the occurrence degree calculation unit 320 calculates the occurrence degree by either one of the method of updating based on the predefined occurrence degree table 600 based on the failure occurrence date and time information and the method of updating relatively from the failure occurrence frequencies of all failure modes.

[0051] In Fig. 14, an example of updating the occurrence degree of "Failure mode: Scratch" for "Defect occurrence process: Welding" and "Component name: Component 1" based on the occurrence degree table 600 is shown. Here, at the initial construction (January 1, 2023), the occurrence degree was set to 3, and the case where the first defect occurred 4 days later (January 5, 2023) and the occurrence degree was updated to 7 is shown. Also, the case where the second defect occurred 54 days later (March 1, 2023) and the occurrence degree was updated to 4, and the third defect occurred 31 days later (April 1, 2023) and the occurrence degree was updated to 5 is shown. Also, when using the method of updating relatively from the defect occurrence frequencies of all failure modes, the user can arbitrarily select the calculation method from the simple moving average, weighted moving average, and exponential smoothing moving average.

[0052] <Detailed description of the operation of the impact degree calculation unit 330> Fig. 15 is a flowchart explaining the operation of the impact degree calculation unit 330. The impact degree calculation unit 330 inputs the work performance data 140 of all processes into XGBOOST (Gradient Boosting Regression Tree) of the machine learning model (S1510). Next, the user selects one column as the target variable from the work performance data 140 input in step 1510 (step 1520). Next, the impact degree calculation unit 330 calculates the contribution rate of the explanatory variable with respect to the target variable set in step 1420 (step 1530). Here, since the explanatory variable refers to any process, the value calculated in step 1530 is set as the impact degree of the failure mode specified by the failure mode analysis unit 310 (step 1540). In this case, it can also be said that the impact degree calculation unit 330 calculates the contribution rate of each process related to the target variable using machine learning from the work performance data 140, and sets the calculated contribution rate as the impact degree of each process. Also, it can be said that the impact degree calculation unit 330 calculates the impact degree based on at least one target variable determined by the user.

[0053] <Detailed description of the operation of the detectability calculation unit 340> Fig. 16 is a diagram explaining the operation of the detectability calculation unit 340. The detection degree calculation unit 340 calculates the detection degree by referring to the defect discovery method 132 in the same list of defect causes 130 as the failure mode specified by the failure mode analysis unit 310. In this case, it can also be said that the detection degree calculation unit 340 calculates the detection degree based on the defect discovery method information (in this case, the defect discovery method 132) described in the list of defect causes 130 for the failure mode. Here, as the method for calculating the detection degree, the user can arbitrarily select a method of referring to and updating the most recent list of defect causes based on the predefined detection degree table 700 described in FIG. 7, and a method of referring to all the lists of defect causes 130 and updating with the average value. In this case, it can also be said that the detection degree calculation unit 340 calculates the detection degree by either one of the method of referring to and updating the most recent list of defect causes 130 based on the predefined detection degree table 700 and the method of referring to all the lists of defect causes 130 and updating with the average value, based on the defect discovery method information (in this case, the defect discovery method 132).

[0054] FIG. 16 shows an example of updating the detection degree for the "failure mode: scratch" of "defect occurrence process: welding" and "part name: part 1" similar to FIG. 14, based on the detection degree table 700. Here, at the initial construction (January 1, 2023), the detection degree was set to 10, and it shows the case where after the first occurrence of a defect on the 4th day (January 5, 2023) and it was detected in the subsequent process, the detection degree was updated to 4. Also, after the second occurrence of a defect on the 54th day (March 1, 2023) and it was detected in the next process, the detection degree was updated to 3, and after the third occurrence of a defect on the 31st day (April 1, 2023) and it was detected during the work, the detection degree was updated to 2.

[0055] <Description of the life cycle of the failure mode effect analysis management system 210> FIG. 17 is a diagram showing the life cycle of the failure mode effect analysis management system 210. The failure mode effect analysis management system 210 may manually register the FMEA at the initial operation, or may not register the FMEA (S1610). Next, when a defect occurs (S1620), a list of defect causes 130 is manually created (S1630). Next, the RPN 127 of the FMEA data 120 is updated (S1640) using the work performance data stored in the work performance data storage unit 240, the FMEA data 120 stored in the FMEA management unit 260, and the defect cause list 130 stored in the defect cause list storage unit 280. Thereafter, every time a defect occurs, the same process is performed to update the RPN 127 of the FMEA data 120 at all times (S1650). In this case, it can also be said that the FMEA (Failure Mode and Effects Analysis) is updated every time the defect cause list 130 is created.

[0056] <Explanation of the notification screen> FIG. 18 is a diagram showing a notification screen 1800 when notifying the user in S1015 of FIG. 10. Here, the notification screen 1800 to the user when a new failure mode occurs is shown. The user can select whether to register the new failure mode in the FMEA management unit 260 by operating the failure mode notification screen 1800. The user refers to the notified content 1850 and presses the add button 1820 when adding a new failure mode. On the other hand, when not adding a new failure mode, the delete button 1810 is pressed. Also, when added, it is immediately reflected in the FMEA view 1860, and the severity, occurrence, and detectability of the new failure mode can be changed by pressing the edit mode button 1830. When the user only wants to view, it is possible to switch to the view mode by pressing the view mode button 1840.

[0057] 〔Second Embodiment〕 In the second embodiment, the case of automatically updating the HFMEA (Health care FMEA) used in the medical field will be described. FIG. 19 is a diagram showing HFMEA data 120A as FMEA data 120 used in the medical field. The illustrated HFMEA data 120A has, as columns, a medical process 121A, a medical failure mode 123A, severity 124, occurrence 125, detectability 126, RPN 127, and need for countermeasure 1208. Comparing FIG. 19 with FIG. 4, the medical process 121A and process 121, and the medical failure mode 123A and failure mode 123 are similar, although they have different names. On the other hand, in FIG. 19, there is no column corresponding to the part name 122 in FIG. 4.

[0058] FIG. 20 is a diagram showing a cause-of-defect list 130A as the cause-of-defect list 130 of the second embodiment. The cause-of-defect list 130A of the second embodiment has a date of occurrence 131A, a method of defect discovery 132, a process of defect occurrence 133, detailed defect content 135, cause of defect 136, and method of defect treatment 137. Comparing FIG. 20 with FIG. 8, the date of occurrence 131A and the defect occurrence date and time 131 are similar, although they have different names. On the other hand, in FIG. 20, there is no corresponding item to the part name 134 in FIG. 8. However, it is possible to automatically update the HFMEA in the medical field with the same functional configuration and flowchart as in the first embodiment.

[0059] 〔Third Embodiment〕 In the third embodiment, a case of automatically updating FMEA (hereinafter referred to as SCFMEA to distinguish it from the FMEA used in the first example) used in the risk management of the supply chain will be described. FIG. 21 is a diagram showing SCFMEA data 120B as FMEA data 120 used in the risk management of the supply chain. The illustrated SCFMEA data 120B has, as columns, company name 2101, SCOR model 2102, process name 121B, risk event 123B, severity 124, occurrence 125, detection 126, RPN 127, and need for countermeasure 1208. Comparing FIG. 21 with FIG. 4, columns for the company name 2101 and the SCOR model 2102, which is a reference model for supply chain processes proposed by the U.S. Supply Chain Council (SCC), are added. Also, the process name 121B and process 121, and the risk event 123B and failure mode 123 have different names but are similar. On the other hand, in FIG. 21, there is no column corresponding to the part name 122 in FIG. 4.

[0060] FIG. 22 is a diagram showing the cause of defect list 130B as the cause of defect list 130 of the third embodiment. The cause of defect list 130B of the third embodiment has company 2201, occurrence date 131B, SCOR model 2202, incident defect discovery method 132B, process name 133B, incident content details 135B, cause of defect 136, and defect treatment method 137. Comparing FIG. 22 with FIG. 8, the company 2201 and the SCOR model 2202 are added. Also, the occurrence date 131B and the defect occurrence date and time 131, the incident defect discovery method 132B and the defect discovery method 132, the process name 133B and the defect occurrence process 133, and the incident content details 135B and the defect content details 135 have different names but are similar. On the other hand, in FIG. 22, there is nothing corresponding to the part name 134 in FIG. 8. However, SCFMEA automatic update in the supply chain is possible with the same functional configuration and flowchart as in the first embodiment.

[0061] According to the embodiments described in detail above, it is possible to calculate FMEA without the intervention of a skilled person or the like, and since automatic update is possible, it is possible to provide a failure mode effect analysis management system 210 capable of managing risks within the business that change moment by moment. In addition, in the manufacturing industry, for operations with high potential risks, the RPN127 may be underestimated, and the risks may be underestimated. Also, there were problems such as labor and cost. For example, it was necessary to manually update each variable of the RPN in FMEA. Also, when a new failure mode occurred, it was necessary to visually confirm whether it was an existing failure mode or an unknown failure mode. Furthermore, in the case of a method specialized for a specific field such as an earthquake, it cannot be applied to the manufacturing industry.

[0062] According to the failure mode and effects analysis management system 210 of this embodiment, it is unlikely that the RPN127 will be underestimated and the risks will be underestimated. Also, each variable of the RPN127 in FMEA can be automatically updated, and it is unlikely that problems such as labor and cost will occur. Furthermore, even when a new failure mode occurs, it can be automatically discriminated, and prompt countermeasures can be taken. And FMEA can be calculated by a method suitable for the manufacturing industry.

[0063] <Explanation of the failure mode and effects analysis management method> In this way, the processing performed by the failure mode and effects analysis management system 210 is realized by the cooperation of software and hardware resources. That is, the processor 211 provided in the failure mode and effects analysis management system 210 loads a program that realizes each function of the failure mode and effects analysis management system 210 into a storage device 212 such as a RAM and executes it to realize these functions. Therefore, the processes performed by the failure mode effects analysis management system 210 described above are such that, by the processor executing the program recorded in the memory, from the results of past failure mode effects analysis and the list of failure causes, which is the history when a defect occurs, the failure mode representing the content of the defect is discriminated, from the results of past failure mode effects analysis and the list of failure causes, the occurrence degree of the defect for the failure mode is calculated, from the results of past failure mode effects analysis and the work performance data representing the work performance, the impact degree when a defect occurs for the failure mode is calculated, from the results of past failure mode effects analysis and the list of failure causes, the detection degree of the defect for the failure mode is calculated, and the results of the failure mode effects analysis for the failure mode are updated. It can be regarded as a failure mode effects analysis management method.

[0064] Also, the program operating in the failure mode effects analysis management system 210 is a program for causing a computer to realize a function of discriminating the failure mode representing the content of the defect from the results of past failure mode effects analysis and the list of failure causes, which is the history when a defect occurs, a function of calculating the occurrence degree of the defect for the failure mode from the results of past failure mode effects analysis and the list of failure causes, a function of calculating the impact degree when a defect occurs for the failure mode from the results of past failure mode effects analysis and the work performance data representing the work performance, a function of calculating the detection degree of the defect for the failure mode from the results of past failure mode effects analysis and the list of failure causes, and a function of updating the results of the failure mode effects analysis for the failure mode.

[0065] Note that the program for realizing this embodiment can of course be provided by means of communication, and can also be provided by being stored in a recording medium such as a CD-ROM.

[0066] As described above, this embodiment has been explained, but the technical scope of the present invention is not limited to the scope described in the above embodiment. It is clear from the description of the claims that those obtained by making various changes or improvements to the above embodiment are also included in the technical scope of the present invention.

Explanation of Reference Numerals

[0067] 1... Information processing system, 120... FMEA data, 130... List of failure causes, 131... Date and time of failure occurrence, 132... Method of failure discovery, 140... Work performance data, 210... Failure mode and effects analysis management system, 230... Work performance data generation device, 250... FMEA generation device, 270... List of failure causes generation device, 310... Failure mode analysis section, 320... Severity calculation section, 330... Occurrence calculation section, 340... Detection calculation section, 350... FMEA update section, 500... Occurrence table, 600... Severity table, 700... Detection table

Claims

1. A failure mode analysis unit that discriminates a failure mode representing the content of a defect from the results of past failure mode effect analyses and a defect cause list that is the history when a defect occurred; An occurrence degree calculation unit that calculates the occurrence degree of a defect with respect to the failure mode from the results of the past failure mode effect analysis and the defect cause list; An impact degree calculation unit that calculates the impact degree when a defect occurs with respect to the failure mode from the results of the past failure mode effect analysis and work performance data representing work performance; A detection degree calculation unit that calculates the detection degree of a defect with respect to the failure mode from the results of the past failure mode effect analysis and the defect cause list; An update unit that updates the results of the failure mode effect analysis for the failure mode; A failure mode effect analysis management system comprising the above.

2. The failure mode analysis unit discriminates the failure mode by the distance between vectors after converting the defect content described in the defect cause list for the failure mode into a vector. The failure mode effect analysis management system according to Claim 1.

3. The occurrence degree calculation unit calculates the occurrence degree from the defect occurrence date and time information described in the defect cause list for the failure mode. The failure mode effect analysis management system according to Claim 1.

4. The occurrence degree calculation unit calculates the occurrence degree by either a method of updating based on a predefined occurrence degree table based on the defect occurrence date and time information or a method of updating relatively from the defect occurrence frequencies of all the failure modes. The failure mode effect analysis management system according to Claim 3.

5. The impact degree calculation unit calculates the contribution rate of each process related to the target variable using machine learning from the work performance data, and sets the calculated contribution rate as the impact degree of each process. The failure mode effect analysis management system according to Claim 1.

6. The impact degree calculation unit calculates the impact degree based on at least one of the target variables determined by the user. The failure mode effect analysis management system according to Claim 5.

7. The detection degree calculation unit calculates the detection degree based on the defect discovery method information described in the defect cause list for the failure mode. The failure mode effect analysis management system according to Claim 1.

8. The detection degree calculation unit calculates the detection degree by either one of the following methods: referring to and updating the most recent failure cause list based on a predefined detection degree table based on the defect discovery method information, and referring to all the failure cause lists and updating with an average value. The failure mode effect analysis management system according to claim 7.

9. The update unit notifies the user of the failure mode discriminated by the failure mode analysis unit, the occurrence degree calculated by the occurrence degree calculation unit, the impact degree calculated by the impact degree calculation unit, and the detection degree calculated by the detection degree calculation unit. When the user selects to update, the update unit updates the result of the failure mode effect analysis for the failure mode. The failure mode effect analysis management system according to claim 1.

10. The failure mode effect analysis is updated each time the failure cause list is created. The failure mode effect analysis management system according to claim 1.

11. By a processor executing a program recorded in a memory, discriminate a failure mode representing the content of a defect from the results of past failure mode effect analyses and a failure cause list that is the history when a defect occurred, calculate the occurrence degree of the defect for the failure mode from the results of the past failure mode effect analyses and the failure cause list, calculate the impact degree when a defect occurred for the failure mode from the results of the past failure mode effect analyses and work performance data representing the work performance, calculate the detection degree of the defect for the failure mode from the results of the past failure mode effect analyses and the failure cause list, update the result of the failure mode effect analysis for the failure mode, Failure mode effect analysis management method.

12. A failure mode effect analysis management system that performs failure mode effect analysis, a failure mode effect analysis occurrence device that sends the results of past failure mode effect analyses to the failure mode effect analysis management system, a failure cause list occurrence device that sends a failure cause list that is the history when a defect occurred, and a work performance data occurrence device that sends work performance data representing work performance, and the failure mode effect analysis management system includes a failure mode analysis unit that discriminates a failure mode representing the content of a defect from the results of past failure mode effect analyses and the failure cause list, an occurrence degree calculation unit that calculates the occurrence degree of the defect for the failure mode from the results of the past failure mode effect analyses and the failure cause list, An impact calculation unit that calculates the impact level when a defect occurs for the failure mode from the results of the past failure mode effect analysis and the work performance data; A detection degree calculation unit that calculates the detection degree of the defect for the failure mode from the results of the past failure mode effect analysis and the defect cause list; An update unit that updates the results of the failure mode effect analysis for the failure mode; An information processing system comprising the above.

Images