Inspection support system and inspection support method

Abstract

Narrow down the scope of testing appropriately. [Solution] The system manages graph data as process management data representing the flow of multiple processes. Each of the multiple nodes includes a process node containing information about that process and an entity node containing information about the entity associated with that process. This includes: If there is a defective product, the system identifies candidate cause entities as potential causes of the defect in a traceback starting from the defective product, narrows down the cause entities based on the relative degree of the defect of each candidate entity, identifies the product ID of a product with the same manufacturing conditions as the defective product in a traceforward starting from the narrowed-down cause entity, and designates the product with the identified product ID as the product to be inspected.

Description

Technical Field

[0001] The present invention generally relates to data processing techniques for inspection support.

Background Art

[0002] Inspections of products, equipment, etc. are generally carried out. Appropriately narrowing down the inspection target contributes to efficient inspection and thus to inspection support. Regarding narrowing down the inspection target, for example, it is conceivable to use the techniques disclosed in Patent Document 1 or 2. That is, it is conceivable to use the techniques disclosed in Patent Document 1 or 2 to identify the scope of influence of a defect and narrow down the inspection target based on the identified scope of influence. [[ID=***]]

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] When an abnormal product is found in the sampling inspection of products, if all products belonging to the range related to the abnormal product are set as inspection targets, the inspection burden is large and it is not realistic.

Means for Solving the Problems

[0005] The system manages graph data as process management data representing the flow of multiple processes. Each node includes a process node containing information about that process and an entity node containing information about entities associated with that process. If there is an abnormal product, the system identifies candidate cause entities as potential causes of the abnormality in a traceback starting from the abnormal product, narrows down the cause entities based on the relative degree of abnormality of each candidate entity, and in a traceforward starting from the narrowed-down cause entities, identifies the product ID of a product with the same manufacturing conditions as the abnormal product, and designates the product with the identified product ID as the product to be inspected. [Effects of the Invention]

[0006] According to the present invention, the target of inspection can be appropriately narrowed down. [Brief explanation of the drawing]

[0007] [Figure 1] This shows an example of the overall system configuration, including the inspection support system according to the embodiment. [Figure 2] An example of the logical structure of a testing support system is shown. [Figure 3] An example of a part of the structure of the process control model according to the embodiment is shown. [Figure 4] A schematic example of assigning a pseudo-lot ID is shown below. [Figure 5] This document shows the flow of the inspection target narrowing process and an example of the structure of quality control data. [Figure 6] A detailed example of S500 to S502 in Figure 5 is shown. [Figure 7] Detailed examples of S503 and S504 in Figure 5 are shown. [Figure 8] Detailed examples of S505 and S506 in Figure 5 are shown. [Figure 9] Examples of when retesting is necessary are shown. [Figure 10] Examples of when retesting is necessary are shown. [Figure 11] An example of S505 is shown in Figure 5. [Figure 12] A modified example of S503 in Figure 5 is shown. [Modes for carrying out the invention]

[0008] In the following explanation, "interface device" may refer to one or more interface devices. These one or more interface devices may be at least one of the following: An I / O interface device is one or more I / O (Input / Output) interface devices. An I / O (Input / Output) interface device is an interface device to at least one of the following: an I / O device and a remote display computer. The I / O interface device to the display computer may be a communication interface device. The at least one I / O device may be either a user interface device, such as an input device like a keyboard and a pointing device, or an output device like a display device. A communication interface device which is one or more communication interface devices. One or more communication interface devices may be one or more identical communication interface devices (for example, one or more NICs (Network Interface Cards)) or two or more different communication interface devices (for example, a NIC and an HBA (Host Bus Adapter)).

[0009] Furthermore, in the following explanation, "memory" refers to one or more memory devices, which are examples of one or more storage devices, and may typically be main memory devices. At least one memory device in memory may be a volatile memory device or a non-volatile memory device.

[0010] Furthermore, in the following explanation, "persistent storage device" may refer to one or more persistent storage devices, which are examples of one or more storage devices. Persistent storage devices are typically non-volatile storage devices (e.g., auxiliary storage devices), and specifically may be, for example, HDDs (Hard Disk Drives), SSDs (Solid State Drives), NVME (Non-Volatile Memory Express) drives, or SCMs (Storage Class Memory).

[0011] Furthermore, in the following explanation, "storage device" may refer to at least memory, including both memory and persistent storage.

[0012] Furthermore, in the following explanation, "processor" may refer to one or more processor devices. At least one processor device may typically be a microprocessor device such as a CPU (Central Processing Unit), but may also be other types of processor devices such as a GPU (Graphics Processing Unit). At least one processor device may be single-core or multi-core. At least one processor device may be a processor core. At least one processor device may be a broad-sense processor device such as a circuit that is a collection of gate arrays according to a hardware description language that performs some or all of the processing (e.g., FPGA (Field-Programmable Gate Array), CPLD (Complex Programmable Logic Device), or ASIC (Application Specific Integrated Circuit)).

[0013] In the following description, the function may be described in terms of a "yyy unit", but the function may be realized by one or more computer programs being executed by a processor, or by one or more hardware circuits (such as an FPGA or an ASIC), or by a combination thereof. When the function is realized by a program being executed by a processor, since the defined processing is performed while appropriately using a storage device and / or an interface device, etc., the function may be regarded as at least a part of the processor. The processing described with the function as the subject may also be the processing performed by the processor or a device having the processor. The program may be installed from a program source. The program source may be, for example, a program distribution computer or a computer-readable storage medium (such as a non-temporary storage medium). The description of each function is an example, and a plurality of functions may be combined into one function, or one function may be divided into a plurality of functions.

[0014] In the following description, when describing without distinguishing between elements of the same kind, common reference numerals among the reference numerals are used, and when distinguishing between elements of the same kind, reference numerals may be used.

[0015] In the following description, "4M" means the first letters of the four elements of Man, Machine, Method, and Material. In the following description, the four elements constituting "4M" are denoted as "Man", "Machine", "Method", and "Material", and these are collectively referred to as "4M elements".

[0016] FIG. 1 shows a configuration example of an entire system including an inspection support system according to an embodiment.

[0017] In this embodiment, the inspection support system 1 is a physical computer system (one or more physical computers) and has an interface device 51, a storage device 52, and a processor 53 connected thereto. The inspection support system 1 may also be a logical computer system based on a physical computer system (for example, a virtual machine or a cloud computing system).

[0018] The interface device 51 is connected to network 2 (for example, the Internet or a WAN (Wide Area Network)). One or more data generators 5 (for example, data generators 5a to 5c) and the user device 35 are connected to network 2. The data generators 5 and the user device 35 are connected to the inspection support system 1 via network 2.

[0019] The data generation device 5 is a device that collects or generates actual data. For example, data generation device 5a may be a barcode reader that acquires worker work logs, or a PC or server that collects work logs. Data generation device 5b may be a machine that processes parts or assembles finished products. Data generation device 5c may be a sensor that collects inspection information of RFID (Radio Frequency IDentifier) ​​attached to parts or finished products. The actual data collected or generated by these data generation devices 5 may be transmitted via the network 2 to the inspection support system 1 (and / or a storage device connected to the network 2 and located outside the inspection support system 1). "Actual data" may be data generated or collected on site (e.g., a manufacturing site), and in particular may be at least a portion of data relating to entities belonging to the 4M elements, such as data relating to equipment belonging to Machine, data relating to workers belonging to Man, data relating to parts and products belonging to Material, and data relating to procedures belonging to Method. Hereinafter, entities belonging to the 4M elements may be referred to as "4M entities".

[0020] The user device 35 is an information processing device (computer) such as a personal computer or smartphone. The inspection support system 1 may be the server and the user device 35 may be the client. The inspection support system 1 transmits filtering result information representing the result of narrowing down the inspection target to the user device 35, and the user device 35 displays the inspection target represented by the filtering result information. In this embodiment, the "user" may be the person in charge of inspection.

[0021] Figure 2 shows an example of the logical configuration of the inspection support system 1.

[0022] The storage device 52 of the inspection support system 1 stores, for example, management data 15. The management data 15 contains data for each 4M entity (for example, data representing the process in which the 4M entity is involved, the time period of work or operation of the 4M entity, and the hourly status of the 4M entity). In this embodiment, the management data 15 includes data representing the process control model described later, and quality control data described later. The management data 15 may also include statistical data.

[0023] The inspection support system 1 includes middleware 291 and application 8, and the processor 53 executes this software. Application 8 may reside outside of the inspection support system 1.

[0024] Middleware 291 includes an API (Application Programming Interface) section 11 and a processing section 12. Middleware 291 may communicate with the user device 35 via or without application 8. In the latter case, middleware 291 has a UI (User Interface) section (not shown), and the UI section may communicate with the user device 35.

[0025] The API unit 11 communicates with the application 8. For example, the API unit 11 receives a first request from the application 8 and sends a second request (for example, a request to perform narrowing down the target of inspection) based on the first request to the processing unit 12. The API unit 11 receives a response to the second request from the processing unit 12 and, based on that response, returns a response to the first request from the application 8 to the application 8.

[0026] The processing unit 12 receives a request (e.g., a second request) via the API unit 11 (or UI unit), and based on the request, it refers to or updates the management data 15 and returns a response based on the result to the requester. The processing unit 12 includes an abnormality content identification unit 200, a traceback unit 201, a cause identification unit 202, a cause narrowing unit 203, an impact scope identification unit 204, a priority setting unit 205, a re-inspection target determination unit 206, a result output unit 207, a data management unit 208, and a pseudo-lot ID assignment unit 209. The data management unit 208 creates or updates the management data 15, for example, by including data from the data generator 5 in the management data 15 either as is or after processing. Other functions 200 to 207 and 209 will be described later.

[0027] Figure 3 shows an example of a data structure representing part of the process control model that the control data 15 represents.

[0028] A process control model represents the sequence (flow) of multiple processes and, for each of the one or more 4M elements associated with each process, information about the 4M entities belonging to that 4M element.

[0029] Specifically, for example, a process control model is a graph containing 150 nodes and edges. The edges in the graph may include undirected edges, but typically directed edges are sufficient. A process control model can be a Directed Acyclic Graph (DAG).

[0030] Node 150 holds information about the process, information about 4M entities, and relevance information for linking the information about 4M entities to the information about the process. Hereinafter, Node 150 holding information about the process may be referred to as "Process Node 150A," Node 150 holding information about 4M entities may be referred to as "4M Node 150M," and Node 150 holding relevance information may be referred to as "Relevance Node 150R." Relevance Node 150R, which holds relevance information, may exist between 4M Node 150M, but it is included in one of those 4M Node 150M and does not necessarily have to exist as a node.

[0031] The information held by process node 150A includes A-ID (the ID of the process corresponding to process node 150A) and O-Info. (information regarding the operation of the process corresponding to process node 150A). O-Info. may include, for example, information indicating when and what kind of operation occurred (for example, information including the start time and operation time of product processing for each product).

[0032] The information held by 4M node 150M includes the ID of the 4M entity corresponding to 4M node 150M and O-Info., which is information regarding the operation of the 4M entity corresponding to 4M node 150M. O-Info. may include, for example, information indicating when and what kind of operation occurred (for example, information including information indicating the start and end times of operation for each operation or for each product). In the example shown in Figure 3, 4M node 150M includes Man node 150MW, Machine node 150MM, Material node 150MP, and Method node 150MS. Also, in the example shown in Figure 3, the IDs of the 4M entities include Man-ID (worker ID), Mac.-ID (equipment ID), Mat.-ID (part or product ID), and Met.-ID (procedure ID). For each 4M element, there may be one 4M node 150M associated with a single process node 150A for that 4M element, and each 4M entity belonging to that 4M element (e.g., Man) may include an entity ID (e.g., for each of worker A, worker B, ...), or there may be one for each 4M entity belonging to that 4M element.

[0033] The information held by the relationship node 150R includes at least a portion of the process information (e.g., at least the A-ID) and at least a portion of the information of the 4M entity associated with the process information (e.g., at least the 4M entity ID). Alternatively, a 4M node 150M may be linked to a process node 150A via a single edge, without using the relationship node 150R, and the information held by the relationship node 150R may be associated with that edge. A single process node 150A can be associated with at least one 4M entity's 4M node 150M via at least one relationship node 150R.

[0034] In the example shown in Figure 3, the thick arrows represent edges. Also, in the example shown in Figure 3, the solid or dashed curved arrows represent examples of the direction of information traversal. For example, one process node 150A1 is associated with 4M nodes 150MP1, 150MP2, 150MW, 150MM, and 150MS through relationship nodes 150R1 to 150R5. The processing unit 12 identifies relationship nodes 150R1 to 150R5 using the A-ID of process node 150A1 as a key, and identifies 4M nodes 150MP1, 150MP2, 150MW, 150MM, and 150MS using the IDs in the information of relationship nodes 150R1 to 150R5 as a key.

[0035] The user device 35 may issue a request to the application 8 to narrow down the inspection target, and the API unit 11 of the middleware 291 may receive a request from the application 8 to execute the inspection target narrowing process. The processing unit 12 may, in response to the request received by the API unit 11, refer to the management data 15 and perform the inspection target narrowing process as data processing that includes traversing nodes in the process control model represented by the management data 15, according to a specified unit or a specified target (e.g., a process, 4M element, or 4M entity). The processing unit 12 may return narrowing result information representing the result of narrowing down the inspection target to the application 8 through the API unit 11. The application 8 may transmit the narrowing result information received through the API unit 11 to the user device 35.

[0036] This embodiment will be described in more detail below. In the following description, the relationship between the first node and the second node via a single edge can be expressed as "the first node belongs to the second node" or "the first node is directly related to the second node."

[0037] Figure 4 schematically shows an example of assigning a pseudo-lot ID.

[0038] It is possible to use the ID of a product flowing through multiple processes as a key to trace the process nodes 150A of those processes, or to trace the 4M nodes 150M associated with each of the one or more process nodes 150A.

[0039] However, when an anomaly occurs, even using the ID of the object (e.g., product) where the anomaly occurred as a key, it is not easy to identify the scope of the anomaly's impact.

[0040] On the other hand, while the product moves between processes, the start and end of each process are repeated. For products that have completed a process, the next process begins. By focusing on both the process units and the interactions between processes, a pair of start and end times can be considered a pseudo-lot, and these pseudo-lots can be seen as flowing between processes.

[0041] Therefore, in this embodiment, a pseudo-lot ID assignment unit 209 is provided, and the pseudo-lot ID assignment unit 209 assigns an ID to a pseudo-lot, which is a pair of the start time and end time of a process, for each process. Since the end time of one process is associated with the start time of the next process, the same pseudo-lot ID is assigned to the pseudo-lot containing the end time of one process and to the pseudo-lot containing the start time of the next process (the start time of the next process which is associated with the end time of the previous process). As a result, the same pseudo-lot ID is assigned between processes. This makes it possible to traverse the process management model using the pseudo-lot ID as a key.

[0042] For each process, the start time and end time may be the actual start and end times of the process, or the start time of the process may be the time when the process node 150A is generated, and the end time of the process and the start time of the next process may be the time when the process node 150A of the next process is generated. The process node 150A may be generated by the data management unit 208 and included in the management data 15. For example, the data management unit 208 may identify the start of a process based on data from the data generator 5, generate the process node 150A for the identified process, and include the process node 150A in the management data 15. Alternatively, the data management unit 208 may identify one or more 4M entities associated with the identified process based on data from the data generator 5, generate a 4M node 150M for each identified 4M entity of the 4M element to which the 4M entity belongs, and associate the 4M node 150M with the process node 150A.

[0043] Figure 5 shows the flow of the inspection target narrowing process and an example of the quality control data structure.

[0044] Quality control data 550 is included in control data 15. Quality control data 550 includes quality standard data 551 and anomaly definition data 552.

[0045] Quality standard data 551 represents one or more quality standard values ​​and multiple quality value ranges. The "quality standard value" is the standard value of the quality value, and may be, for example, the median (or average) or acceptance value of the quality value of products judged to be normal in inspection. The "quality value" is a value that represents the quality of the product, and may be, for example, the quantity of a certain element that the product possesses. The "median" is an example of the first quality standard value. The "acceptance value" is an example of the second quality standard value, and is the lower limit of the quality value of products judged to be normal.

[0046] In quality standard data 551, multiple quality value ranges are defined within the range below the median. A attention level is associated with each quality value range. In abnormality definition data 552, the need for reinspection is associated with each attention level. "Attention level" represents the degree of need for attention. The further the quality value is from the pass value, the lower the attention level, and the further the quality value is from the pass value, the higher the attention level. "Need for reinspection" represents the degree of need for reinspection. In this embodiment, the degree of need for reinspection is categorized as "not required," "priority A," "priority B," "priority C," and "stop," in order of increasing need. "Not required" means the need for reinspection is the lowest (no reinspection is needed), and "stop" means the need for reinspection is the highest (for example, manufacturing equipment is stopped and all products are required to be reinspected). "Priority A," "priority B," and "priority C" each mean that the products assigned that priority require reinspection. As the attention level increases, the number of items requiring reinspection increases.

[0047] The inspection target narrowing process is initiated by the processing unit 12 when an abnormality is found during the inspection. The object being "inspected" in this paragraph may be any of products or machines, but in this embodiment, it is assumed to be a product, particularly a finished product.

[0048] The process of narrowing down the inspection target includes identifying the type of anomaly (S500), tracing back (S501), identifying the cause (S502), narrowing down the cause (S503), identifying the scope of impact (S504), setting the priority (S505), deciding on the target for re-inspection (S506), and outputting the results (S507). Detailed examples of S500 to S507 are described below.

[0049] Figure 6 shows a detailed example of steps S500 to S502 in Figure 5.

[0050] In inspection Z, which is a sampling inspection, an abnormality is detected in product "200" (product with ID "200") from finished product H (lot). Specifically, it is detected that the quality value of product "200" that was the subject of inspection is below the median value (an example of the first quality standard value). The abnormality identification unit 200 refers to the quality standard data 451 to identify the quality value range to which the quality value of product "200" belongs, and identifies the attention level associated with the identified quality value range. Here, it is assumed that attention level "3" is identified. As shown in the figure, the Material node 150MPh of finished product H may belong to the process node 150Az of inspection Z. The Material node 150MPh (e.g., O-Info.) may contain the ID of each product (individual) in finished product H (lot).

[0051] In S500, the abnormality identification unit 200 generates abnormality data 601 which includes data indicating a warning level of "3". The abnormality data 601 includes data indicating the abnormality in addition to the warning level of "3". The abnormality may be detected manually or automatically in inspection Z, and the data indicating the abnormality may be entered manually or by an inspection device including a computer. The abnormality data 601 is referenced in subsequent processing to narrow down the items to be reinspected.

[0052] In S501, the traceback unit 201 performs a traceback that includes traversing the graph representing the process control model (DAG in this embodiment) in reverse direction, using the product ID "200" (an example of the ID of the object in which an anomaly was detected) as the key. "Tracing a directed graph in reverse direction" means traversing edges and nodes in the opposite direction to the direction of the edges in a directed graph. As a result, the traceback is performed by traversing upstream from the 4M node 150MPh of the finished product H. In the traceback, all process nodes 150A and all 4M nodes 150M that include product ID "200" are identified. That is, all processes and all 4M entities involved in the manufacturing process in which product "200" in which an anomaly was detected are identified. According to the example shown in Figure 6, using product ID "200" in the Material node 150MPh of the finished product H as the key, the following node 150 is identified. In other words, each of the following nodes (for example, the O-Info. of each node) contains product ID "200", and therefore, node 150 is identified in the traceback. • Process node 150Az (process node representing inspection Z) • Material node 150MPg (a 4M node representing the finished product G) is directly connected to process node 150Az. • Process node 150Ay (a process node representing process Y) is directly connected to Material node 150MPg. • Directly connected to process node 150Ay are Man node 150MWe (a 4M node representing worker E), Method node 150MSf (a 4M node representing method F), and Material node 150MPd (a 4M node representing intermediate material D). • Process node 150Ax (a process node representing process x) is directly connected to Material node 150MPd. The Machine node 150MMb (a 4M node representing Machine B), the Method node 150MSc (a 4M node representing Method C), and the Material node 150MPa (a 4M node representing Raw Material A) are directly connected to the process node 150Ax.

[0053] In S502, the cause identification unit 202 identifies 4M node 150M from node 150 identified in the traceback. Here, Material node 150MPa representing raw material A, Machine node 150MMb representing machine B, Method node 150MSc representing method C, Man node 150MWe representing worker E, and Method node 150MSf representing method F are identified. In other words, the cause identification unit 202 identifies the 4M entities associated with the ID of the product in which the anomaly was detected as the cause of the anomaly detected for product "200," specifically raw material A, machine B, method C, worker E, and method F. The 4M entities identified as the cause of the anomaly in S502 can be called "candidate cause 4M entities."

[0054] Figure 7 shows detailed examples of S503 and S504 in Figure 5.

[0055] In S503, the cause narrowing unit 203 narrows down the cause 4M entities from the 4M candidate cause entities identified in S502. Specifically, for example, the cause narrowing unit 203 performs the following for every 150M 4M nodes of the 4M candidate cause entities identified in S502. Let's take the 150M 4M nodes of one 4M candidate cause entity as an example. (S503-1) The cause narrowing unit 203 refers to the O-Info. of the 4M node 150M and identifies the degree of relevance of the candidate cause 4M entity to the product for each product ID, based on one or more perspectives. Any perspective may be used, such as time (for example, from start time to end time) or speed. Here, it is assumed that "time" is used. The degree of relevance for each product ID is identified based on the values ​​for each of the one or more perspectives recorded for each product ID in the O-Info. (S503-2) The cause narrowing unit 203 identifies a statistical value (e.g., average value) of the degree of relevance (e.g., time) for each of the candidate cause 4M entities, based on the degree of relevance for each product. (S503-3) The cause narrowing unit 203 determines whether the correlation of product “200” in which the abnormality was detected satisfies predetermined conditions with respect to the statistical values ​​identified in S503-2. For example, the cause narrowing unit 203 determines whether the time (an example of correlation) that the candidate cause 4M entity was involved with product “200” in which the abnormality was detected is longer than the average time (an example of a statistical value for correlation) identified in S503-2 for the candidate cause 4M entity. (S503-4) If the result of the determination in S503-3 is true, the cause narrowing unit 203 selects the candidate cause 4M entity as the cause 4M entity. If the result of the determination in S503-3 is false, the cause narrowing unit 203 does not select the candidate cause 4M entity as the cause 4M entity.

[0056] As shown in the example in Figure 7, in S503, through the above process, the cause narrowing unit 203 narrows down the five candidate cause 4M entities (raw material A, machine B, method C, worker E, and method F) to two cause 4M entities (machine B and method F).

[0057] In the above example, in S503-3, it is determined whether the relevance of product "200" satisfies predetermined conditions with respect to the statistical value. However, in this example, it may also be determined whether the relevance of product "200" falls within a specific range based on the statistical value. If the relevance of product "200" falls within the so-called normal range (for example, the range defined by the cause narrowing unit 203 based on the relevance statistical value), the candidate cause 4M entity does not need to be considered a cause 4M entity. If the relevance of product "200" does not fall within the so-called normal range, the candidate cause 4M entity may be considered a cause 4M entity.

[0058] In S504, the impact scope identification unit 204 performs a trace forward, which includes traversing the graph representing the process control model in the forward direction, starting from the 4M node 150M of the cause 4M entity narrowed down in S503. "Tracing a directed graph in the forward direction" means traversing edges and nodes in the same direction as the edges in the directed graph. As a result, the trace forward is performed by traversing downstream from the 4M node 150M of the cause 4M entity. In the trace forward, the product ID of the product manufactured under the same manufacturing conditions as the cause 4M entity is identified. As shown in the example in Figure 7, the trace forward is performed starting from the Machine node 150MMb of Machine B, one of the two cause 4M entities (Machine B and Method F).

[0059] Here, "manufacturing conditions" may be those of a simulated lot, or other manufacturing conditions (e.g., temperature, mode, etc.). Specifically, for example, the following processes may be performed. (S504-1) The impact scope identification unit 204 identifies the time period in which product "200" was involved from the O-Info. of Machine node 150MMb of machine B, and identifies a pseudo-lot (time period) and its pseudo-lot ID that overlap all or part of the identified time period from the O-Info. of process node 150Ax to which Machine node 150MMb belongs. (S504-2) The impact scope identification unit 204 identifies the product ID in the impact scope by tracing downstream (towards the Material node 150MPh of the finished product H) using the pseudo-lot ID identified in S504-1 as a key. For example, the impact scope identification unit 204 performs at least one of the following for each of the one or more process nodes 150A, which consist of process node 150Ax and process nodes 150A downstream of process node 150A. The product ID identified in S504-2 is the product ID in the impact scope. (2A) The impact scope identification unit 204 identifies a pseudo-lot (time) corresponding to the pseudo-lot ID identified in S504-1, and identifies product IDs whose time overlaps in whole or in part with that pseudo-lot (time) from the O-Info. of each 4M node 150M belonging to the process node 150A. That is, a pseudo-lot is identified in which all or part of the time in which the product "200" in which the abnormality was detected was involved overlaps with the time in which the causative 4M entity was involved, and the product IDs of products whose time overlaps in whole or in part with the identified pseudo-lot are identified from all 4M nodes 150M belonging to either the process node 150A to which the 4M node 150M of the originating causative 4M entity belongs, or all process nodes 150A downstream thereof. (2B) The impact scope identification unit 204 identifies manufacturing conditions that overlap in whole or in part with the pseudo-lot (time) identified in (2A) from the O-Info. of each 4M node 150M belonging to the process node 150A, and identifies product IDs corresponding to the same manufacturing conditions. In other words, the product IDs of all products manufactured under the same manufacturing conditions that overlap in whole or in part with the pseudo-lot identified in (2A) are identified from all 4M nodes 150M belonging to either the process node 150A to which the 4M node 150M of the originating cause 4M entity belongs, or all process nodes 150A downstream thereof.

[0060] Figure 8 shows detailed examples of S505 and S506 in Figure 5.

[0061] In S505, the priority setting unit 205 acquires raw data related to product "200" in which an anomaly was detected, from the raw data store associated with the 4M node 150M of the originating cause 4M entity. The "raw data store" may be a collection of raw data (for example, data measured by sensors, etc.). The raw data store may be included in the management data 15 for each 4M entity, or it may be stored in the storage device 52 as data not included in the management data 15. The raw data collected and stored for each 4M entity from the data generator 5, etc., may be predetermined. If the originating cause 4M entity is machine B, then the time series of temperature data inside machine B is acquired as raw data related to product "200". An example of the temperature change represented by that time series of temperature data is shown in Figure 9. Note that acquiring raw data is a necessary process for setting the need for reinspection, and is not necessarily required depending on the originating cause 4M entity.

[0062] In S505, the priority setting unit 205 sets multiple different re-inspection requirements based on the acquired raw data. For example, as shown in the example in Figure 9, in the temperature changes identified from the raw data, "×" is set for the range 903 with the highest temperature, "△" is set for ranges close to the range with "×" (for example, ranges 902 and 904 adjacent to range 903), and "〇" is set for ranges far from the range with "×" (for example, range 901 adjacent to range 902 and range 905 adjacent to range 904). The "range with the highest temperature" is an example of a range considered to have had the greatest impact on the abnormality of product "200," and may depend on the manufacturing conditions considered in the identification of the affected range (S504). The "×," "△," and "〇" here are base values ​​for priority and may correspond to values ​​representing the problem level. Each of "×," "△," and "〇" can be referred to as a "rank value" from now on. Furthermore, in Figure 8, the table of machine B may have columns for ID and temperature, with the ID of the product recorded in the ID column and the temperature recorded in the temperature column. For each product, the priority setting unit 205 may convert the product temperature into a rank value of "×", "△", or "〇" based on the temperature, time, and the relationship exemplified in Figure 9 (the relationship between time and range 901-905).

[0063] Furthermore, the determination of which reinspection requirements are set for which manufacturing conditions, or in other words, what rank values ​​are set, varies depending on the 4M element to which the 4M entity belongs, as illustrated in Figure 10. For example, if the 4M element is Machine, the rank value may differ depending on the range of instance values ​​of manufacturing conditions such as temperature and vibration frequency. If the 4M element is Man, the rank value may differ depending on the range of instance values ​​of manufacturing conditions such as years of experience, or on the manufacturing condition (instance value) of whether or not the person in charge of the product in which the anomaly was detected is the same person. If the 4M element is Method, the rank value may differ depending on the type of manufacturing condition (instance value) such as mode, or on the determination result of the manufacturing condition (instance value) such as the captured image. If the 4M element is Material, the rank value may differ depending on the range of manufacturing conditions (instance value) such as numerical values ​​measured from the material, or on the manufacturing condition (instance value) of whether or not the product in which the anomaly was detected is the same as the purchased part number or purchasing company. In this way, the rules for converting the instance values ​​of manufacturing conditions into rank values ​​may differ depending on the 4M element to which the 4M entity belongs.

[0064] In S505, the priority setting unit 205 performs the following for each of the multiple rank values ​​set with respect to the manufacturing conditions of machine B (an example of a cause 4M entity). This makes it possible to associate the rank value set in terms of the manufacturing conditions of machine B with the product identified from the finished product G or H in terms of the time related to those manufacturing conditions. (S505-1) Determine the time for the temperature range (an example of manufacturing conditions) corresponding to the rank value from the raw data or Machine node 150MMb. (S505-2) The pseudo-lot ID of a pseudo-lot that overlaps with all or part of the temperatures identified in S505-1 is identified from process node 150Ax to which Machine node 150MMb belongs. (S505-3) By tracing the process model downstream using the pseudo-lot ID identified in S505-2 as a key, a pseudo-lot with the same pseudo-lot ID is identified from process node 150Ay or 150Az to which the finished product G or H (an example of a 4M entity that may include items to be reinspected) belongs. (S505-4) Identify product IDs whose timeline overlaps in whole or in part with the pseudo-lot identified in S505-3 from the Material node 150MPg or 150MPh of finished product G or H. (S505-5) Associate the rank value with the product ID identified in S505-4.

[0065] In S505, the priority setting unit 205 sets a priority for each rank value. According to the example shown in Figure 8, the relationship between rank value and priority is as follows: "Priority A" is set for "×" (i.e., an abnormal temperature that is the same as or near the abnormal sample (product)). "Priority B" is set for "△" (i.e., a temperature that is far from the abnormal temperature). "Priority C" is set for "〇" (i.e., a temperature that is sufficiently far from the abnormal temperature (e.g., a normal temperature)).

[0066] In S506, the re-inspection target determination unit 206 identifies all products requiring re-inspection corresponding to the attention level (e.g., "3") represented by the abnormality content data 601 from the abnormality definition data 452, and designates products with product IDs associated with the identified re-inspection needs (e.g., priority) as products requiring re-inspection. If the number of products designated for re-inspection does not meet the predetermined conditions, the re-inspection target determination unit 206 further narrows down the products to be re-inspected based on priority.

[0067] Subsequently, in S507, the result output unit 207 displays the refined result information, including the product ID of the product to be reinspected, on the user device 35.

[0068] The above explanation focuses only on machine B as a cause 4M entity for the sake of simplicity. However, if there are multiple cause 4M entities, the same processing as for machine B may be performed for each cause 4M entity. For each cause 4M entity, the rank values ​​determined based on the manufacturing conditions are arranged along an axis, with the time periods overlapping all or part of the pseudo-lots being considered. A priority may then be set by comprehensively considering the multiple rank values ​​arranged along that axis. A specific example is shown in Figure 11. Specifically, for method F, which is another cause 4M entity of machine B, multiple rank values ​​are set based on the manufacturing conditions for product "200". For each rank value, a pseudo-lot in which all or part of the time period corresponding to the manufacturing conditions for which that rank value is set overlaps is identified from process node 150Ay to which method F belongs, and the product IDs in which all or part of the time period overlaps with that pseudo-lot are identified from the Material node 150MPg or 150MPh of finished product G or H. The rank values ​​for the manufacturing conditions of machine B and the rank values ​​for the manufacturing conditions of method F are arranged along a common pseudo-lot ID axis (i.e., the time axis), and the identified product IDs are also arranged along this axis. As a result, multiple rank values ​​of different 4M entities are arranged along the axis for each identified product ID. Priorities are set based on these multiple rank values. For example, if there is even one "×" on the axis, a high priority may be set.

[0069] Furthermore, the displayed filtering results information may include the product ID for each process identified in the traceback performed using product ID "200" as the key, or for each process identified in the traceforward performed starting from the cause 4M entity. Also, if a product consists of multiple parts, it is possible that the product ID of the finished product may not be assigned to the O-Info. of process node 150A (or the 4M node 150M belonging to it) due to two or more parts being combined into one in a certain process. In such cases, the process ID may be included in the filtering results information for each process identified in the above traceback or traceforward.

[0070] Furthermore, in S503 of Figure 5, instead of or in addition to the process described above, the causative 4M entity may be identified based on the statistical data exemplified in Figure 12. The statistical data may include, for example, data representing the number of abnormal cases for each 4M entity. The "number of abnormal cases" is the number of products in which abnormalities were detected, or in other words, the number of times a candidate causative 4M entity was identified. The statistical data may be the data included in the management data 15. A 4M entity with a relatively large number of abnormal cases is selected as the causative 4M entity. For example, after a causative 4M entity is identified by the method described with reference to Figure 7, if that causative 4M entity has a relatively large number of abnormal cases in the statistical data, it may be confirmed as the causative 4M entity. Also, if two or more causative 4M entities are identified from the statistical data among the causative 4M entities identified by the method described with reference to Figure 7, one or more causative 4M entities with a relatively large number of abnormal cases may be confirmed as causative 4M entities from among those two or more causative 4M entities.

[0071] Although one embodiment has been described above, this is merely an example for the purpose of explaining the present invention, and is not intended to limit the scope of the present invention to this embodiment alone. The present invention can be carried out in various other forms.

[0072] For example, the inspection support according to the embodiment may be applied to maintenance. For example, the statistical data illustrated in Figure 12 may include, for each 4M entity, the number of abnormalities and a history of the time when the number of abnormalities was added, for each attention level. The processing unit 12 can refer to the statistical data and identify the trend (e.g., rate of increase) of the number of abnormalities for each 4M entity, for each attention level. For example, the processing unit 12 may receive a specification from the user via the user device 35 regarding the attention level and conditions related to the trend of the number of abnormalities, and for that attention level, identify a 4M entity (e.g., machine B) from the statistical data whose trend of the number of abnormalities satisfies the user-specified conditions, and present the identified 4M entity as a 4M entity recommended for maintenance. In this way, inspections in the user's maintenance work can be supported. [Explanation of Symbols]

[0073] 50: Inspection support system

Claims

1. A storage device that stores process control data representing the flow of multiple processes related to product manufacturing. A processor connected to the aforementioned storage device and Equipped with, The aforementioned process management data is data representing a graph having multiple nodes and multiple edges, The plurality of nodes, for each of the plurality of processes, A process node containing information about the process, The entity node belonging to the process node of the process includes information about the entity associated with the process, Includes, Edges between process nodes are directed edges that represent the flow of the process. If there is a defective product in which an abnormality has been detected, the processor will (A) In a traceback performed by referring to the process control data, starting from the entity node of the entity to which the abnormal product belongs, the entity corresponding to the entity node that has information about the abnormal product is designated as a candidate cause entity, which is an entity that is a candidate for the cause of the abnormality. (B) For each of one or more potential causal entities, the degree of the abnormality of the potential causal entity is identified based on the process control data, and if the degree is relatively high, the potential causal entity is designated as the causal entity. (C) For at least one of the cause entities narrowed down from the one or more candidate cause entities, in a trace forward performed starting from the said cause entity and referring to the process control data, the product ID of a product with the same manufacturing conditions as the defective product is identified. (D) Present the user result information that includes the identified product ID as the product ID of the product being inspected. Inspection support system.

2. (C) The processor is For each identified product ID, the instance value of the manufacturing conditions in at least one causal entity for the product of that product ID is converted to one of several rank values, which are multi-level values, according to the conversion rule associated with that causal entity, and a priority is set for that product ID based on the obtained rank value. Based on the priority assigned to each product ID, the product ID is identified. The inspection support system according to claim 1.

3. The memory device stores abnormality definition data, which is data representing one or more priorities as the need for re-examination for each of the multiple attention levels. (C) The processor is One or more priorities corresponding to the attention level for the detected anomaly are identified from the anomaly definition data, From the product IDs to which priority has been set, identify the product IDs to which the priority has been set to the same as one or more of the priorities identified above. The inspection support system according to claim 2.

4. The aforementioned at least one causal entity is two or more causal entities, (C) The processor is For each identified product ID, for the product of that product ID, the instance value of the manufacturing condition in each of the two or more causal entities is converted into a rank value according to the conversion rule associated with that causal entity, and a priority is set for that product ID based on the two or more rank values ​​obtained for the two or more causal entities. The inspection support system according to claim 2.

5. The aforementioned processor, Based on the aforementioned process control data, for each of the multiple processes, one or more pseudo-lots are identified, each consisting of a start time and an end time for that process. For each of the multiple pseudo-lots, identify the pseudo-lot having an end time associated with the start time of the pseudo-lot from the process immediately preceding the process containing the pseudo-lot, and / or identify the pseudo-lot having a start time associated with the end time of the pseudo-lot from the process immediately following the process containing the pseudo-lot. For each series of pseudo-lots in the aforementioned multiple processes, the same pseudo-lot is assigned to each pseudo-lot in that series of pseudo-lots. (C) The processor For at least one causative entity, identify the pseudo-lot ID of the pseudo-lot in which all or part of the time period in which the defective product was involved overlaps, Using the aforementioned pseudo-lot ID as the key, the trace forward is performed. The inspection support system according to claim 1.

6. A product having the same manufacturing conditions as the defective product is a product in which, for each downstream process of the process to which the at least one causative entity belongs, the time of the entity belonging to that process overlaps in whole or in part with the pseudo-lot corresponding to the identified pseudo-lot ID. The inspection support system according to claim 5.

7. A product manufactured under the same manufacturing conditions as the defective product is a product manufactured under the same manufacturing conditions as the pseudo-lot corresponding to the identified pseudo-lot ID, for all or part of the time, for each downstream process of the process to which the at least one causative entity belongs, with respect to the entities belonging to that process. The inspection support system according to claim 5.

8. In (B), for each of the one or more candidate causal entities, the processor, based on the process management data, For each product ID of the product involved with the suspected causal entity, the degree of relevance of the suspected causal entity to that product ID is identified based on one or more perspectives. Based on the relevance for each product ID, we identify statistical values ​​for relevance. As for the degree of the abnormality, the degree of relevance of the abnormal product is identified, If the identified correlation of the abnormal product is higher than the identified statistical value, the candidate cause entity is designated as the cause entity. The inspection support system according to claim 1.

9. In (B), For each of the one or more candidate causative entities, the degree of the anomaly is the number of past anomalies relating to that candidate causative entity. The aforementioned cause entity is an entity that has a relatively large number of abnormal cases among the one or more candidate cause entities. The inspection support system according to claim 1.

10. Each entity is a 4M entity belonging to one of the following categories: Man, Machine, Method, or Material. The inspection support system according to claim 1.

11. If there is a defective product in which an abnormality has been detected, (A) In a traceback performed by referencing process control data starting from the entity node of the entity to which the abnormal product belongs, the entity corresponding to the entity node that has information about the abnormal product is designated as a candidate cause entity, which is an entity that is a candidate for the cause of the abnormality. (B) For each of one or more potential causal entities, the degree of the abnormality of the potential causal entity is identified based on the process control data, and if the degree is relatively high, the potential causal entity is designated as the causal entity. (C) For at least one of the cause entities narrowed down from the one or more candidate cause entities, in a trace forward performed starting from the said cause entity and referring to the process control data, the product ID of a product with the same manufacturing conditions as the defective product is identified. (D) Present the user result information that includes the identified product ID as the product ID of the product being inspected. This is done by computer. The aforementioned process control data is data representing the flow of multiple processes related to the manufacturing of a product, and is data representing a graph having multiple nodes and multiple edges. The plurality of nodes, for each of the plurality of processes, A process node containing information about the process, The entity node belonging to the process node of the process includes information about the entity associated with the process, Includes, Edges between process nodes are directed edges that represent the flow of the process. Testing support methods.

Images