Data management system and data management method

Abstract

A system according to the present invention identifies one or more change points in business performance from performance data, and calculates, for each of one or more calculation points including the identified one or more change points, a change point score which is a numerical value regarding a management element instance related to the calculation point. The system generates numerical value data including the calculated one or more change point scores, and outputs the numerical value data to an external system.

Description

Data management system and data management method

[0001] The present invention relates generally to data management, for example to assist in analysis or visualization.

[0002] In business (e.g., manufacturing), factors that affect results (e.g., productivity and quality in manufacturing) such as KPIs (Key Performance Indicators) include changes in the 4Ms (Machine, Material, Method, and Human) or 5M1E (4M + Measurement and Environment). The "results" here are, for example, objective variables, and the "factors" are, for example, explanatory variables.

[0003] For example, US Pat. No. 6,299,649 discloses a process for extracting information indicative of changes or modifications to manufacturing-related information.

[0004] Japanese Patent Application Laid-Open No. 2021-33729

[0005] In the following explanation, a "change point" means a time when a change occurs. For example, in the manufacturing industry, the definition of a "change point" is common, meaning a time when some kind of change occurs in production. Specific examples of change points include when there is a worker reassignment, a change in equipment, or a change in lot. A "change point" can be any time when a change occurs, regardless of whether or not human intent is involved. Note that a change that involves human intent can also be called a "change."

[0006] Business processes are analyzed or visualized to contribute to improving results (for example, productivity and quality in the manufacturing industry). For business process analysis or visualization, it is desirable for change points, which are factors in the results, to be numerical data (quantitative data). However, change points are generally not explicitly recorded as numerical data, and sometimes there is no explicit record of change points. As a result, there is a risk that information that could affect the results may be overlooked.

[0007] The system identifies one or more change points in business performance from performance data, calculates a change point score, which is a numerical value, for each of one or more calculation points including the identified one or more change points, for the management element instance related to the calculation point, generates numerical data including the calculated one or more change point scores, and outputs the numerical data to an external system.

[0008] According to the present invention, it is possible to support analysis or visualization that contributes to improving business results.

[0009] 1 shows an example of the configuration of a data management system according to an embodiment; 2 shows an example of the configuration of a work performance table; 3 shows an example of the configuration of a maintenance record table; 4 shows an example of the configuration of a change history table; 5 shows an example of the configuration of a measurement performance table; 6 shows an example of the configuration of a condition management table; 7 shows an example of the configuration of a change point process management table; 8 shows an example of the configuration of a measurement value process management table; 9 shows an example of the configuration of a change point table; 10 shows an example of the configuration of an item management table; 11 shows the flow of a change point data generation process; 12 shows the flow of an item selection process; 13 schematically shows an example of a change point score; 14 shows an example of output data generation for analysis or visualization; 15 shows an example of visualization; 16 shows a schematic example of data management according to a modified example;

[0010] In the following description, an "interface device" may be one or more interface devices. The one or more interface devices may be at least one of the following: - An I / O interface device that is one or more I / O (Input / Output) interface devices. The I / O (Input / Output) interface device is an interface device for at least one of an I / O device and a remote display computer. The I / O interface device for the display computer may be a communication interface device. The at least one I / O device may be a user interface device, for example, either an input device such as a keyboard and a pointing device, or an output device such as a display device. - A communication interface device that is one or more communication interface devices. The one or more communication interface devices may be one or more homogeneous communication interface devices (e.g., one or more NICs (Network Interface Cards)) or two or more heterogeneous communication interface devices (e.g., a NIC and an HBA (Host Bus Adapter)).

[0011] In the following description, "memory" refers to one or more memory devices, which are an example of one or more storage devices, and may typically be a primary storage device. At least one memory device in the memory may be a volatile memory device or a non-volatile memory device.

[0012] In the following description, a "persistent storage device" may refer to one or more persistent storage devices, which are an example of one or more storage devices. A persistent storage device may typically be a non-volatile storage device (e.g., an auxiliary storage device), and more specifically, may be, for example, a hard disk drive (HDD), a solid state drive (SSD), a non-volatile memory express (NVME) drive, or a storage class memory (SCM).

[0013] In the following description, the term "storage device" may refer to at least one of memory and persistent storage device.

[0014] In the following description, a "processor" may refer to one or more processor devices. The at least one processor device may typically be a microprocessor device such as a CPU (Central Processing Unit), but may also be another type of processor device such as a GPU (Graphics Processing Unit). The at least one processor device may be a single-core or multi-core. The at least one processor device may also be a processor core. At least one processor device may be a processor device in a broad sense, such as a circuit that is a collection of gate arrays written in a hardware description language that performs some or all of the processing (for example, an FPGA (Field-Programmable Gate Array), a CPLD (Complex Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit)).

[0015] In the following description, data that produces an output in response to an input may be described using expressions such as "xxx table." However, this data may have any structure (for example, structured data or unstructured data), or may be a learning model such as a neural network, genetic algorithm, or random forest that generates an output in response to an input. Therefore, "xxx table" can be referred to as "xxx information." In the following description, the structure of each table is an example, and one table may be divided into two or more tables, or all or part of two or more tables may be one table.

[0016] In the following description, functions may be described using the expression "yyy unit." However, the functions may be realized by one or more computer programs executed by a processor, by one or more hardware circuits (e.g., FPGAs or ASICs), or by a combination thereof. When a function is realized by a program executed by a processor, the specified processing is performed using a storage device and / or an interface device, etc., as appropriate, and therefore the function may be considered to be at least a part of the processor. Processing described using a function as the subject may be processing performed by a processor or a device having the processor. A 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 (e.g., a non-transitory storage medium). The description of each function is an example; multiple functions may be combined into one function, or one function may be divided into multiple functions.

[0017] In the following description, when elements of the same type are described without distinction, common reference symbols are used, and when elements of the same type are described with distinction, reference symbols are used.

[0018] The unit of "date and time" may be year, month, day, hour, minute, and second, or may be a coarser or finer unit than that.

[0019] In the following description, with respect to business operations such as manufacturing, each element in a management element set such as 4M or 5M1E, for example, human being, machine, method, material, measurement, and environment, will be collectively referred to as a "business management element."

[0020] FIG. 1 shows an example of the configuration of a data management system according to an embodiment.

[0021] In this embodiment, the data management system 100 is a physical computer system (one or more computers). However, instead, it may be a logical computer system based on a physical computer system (for example, a system as a cloud computer service based on a cloud infrastructure). The data management system 100 communicates with a manufacturing site system 110, an analysis system 120, and a visualization system 130. Communication with at least one of these systems 110, 120, and 130 may be performed via a communication network. The communication network may be the Internet, a wide area network (WAN), or a local area network (LAN).

[0022] The manufacturing site system 110 is a system that includes an input system for data related to the manufacturing of products. This "input system" may include one or more sensors provided at the manufacturing site that output measurement value data, or may include an information processing terminal (e.g., a personal computer or smartphone) that accepts manual data input by a worker, manager, or other person. In this embodiment, manufacturing is line manufacturing, but the present invention can also be applied to other manufacturing methods such as job shop manufacturing and cell manufacturing. Data such as measurement value data is input from the manufacturing site system 110 to the data management system 100.

[0023] The analysis system 120 is a physical or logical computer system that analyzes the data output from the data management system 100. Examples of the analysis will be described later.

[0024] The visualization system 130 is a system including a display device, and may be, for example, an input / output device including an input device and a display device, or may be a remote information processing terminal (for example, a client computer). The visualization system 130 visualizes the data output from the data management system 100, and typically displays a screen based on the data.

[0025] The data management system 100 includes an interface device 51, a storage device 52, and a processor 53 connected thereto.

[0026] The interface device 51 communicates with the shop floor system 110, the analysis system 120 and the visualization system 130.

[0027] The storage device 52 stores computer programs executed by the processor 53 and data input and output by the processor 53. Examples of the data include performance data 151, condition management data 152, change point data 153, process management data 154, and item management data 155. The performance data 151 is data representing production performance at the manufacturing site and includes data input from the manufacturing site system 110. The condition management data 152 is data representing the conditions for extracting change points. The change point data 153 is numerical data related to change points, specifically, data on change point scores. The process management data 154 is data representing the relationship between change points and measurement values, processes, and business management elements. The item management data 155 represents the items of data included in the output data for analysis or visualization.

[0028] By executing the computer program, the processor 53 realizes functions such as a data acquisition unit 171, a change point data generation unit 172, an item selection unit 173, and an output data generation unit 174. The data acquisition unit 171 receives data from the manufacturing site system 110 and stores the data in the performance data 151. The change point data generation unit 172 generates change point data 153 based on the performance data 151 in accordance with the extraction conditions represented by the condition management data 152. The change point data generation unit 172 stores the generated change point data 153 in the storage device 52. The item selection unit 173 selects an item of data to be output based on the change point data 153 and the process management data 154, and stores data representing the selected item in the item management data 155. The output data generation unit 174 acquires data corresponding to the data item represented by the item management data 155 from the performance data 151 and the change point data 153, and generates output data based on the acquired data. The output data generation unit 174 outputs the generated output data to the analysis system 120 and / or the visualization system 130 .

[0029] FIG. 2 shows an example of the structure of the work performance table 200 .

[0030] The work performance table 200 is included in the performance data 151. The work performance table 200 represents the relationship between processes, products, and change point categories. A "change point category" is a category that belongs to a change point. For example, any of the business management elements (human, machine, method, material, measurement, or environment) may be classified as a change point category, and an instance that belongs to any of the business management elements may be a change point category. For example, the change point category for an instance that belongs to human may be worker A, B, etc. The change point category for an instance that belongs to machine may be equipment X, equipment Y, etc. The change point category for an instance that belongs to material may be material lot A, material lot B, etc. In addition, any category such as the content of maintenance (e.g., inspection, paint nozzle replacement) may be adopted as a change point category.

[0031] The work performance table 200 has a record for each work performance. The record has information such as a process ID 201, an individual ID 202, a date and time 203, a worker name 204, and a material lot ID 205. In the example shown in Fig. 2, examples of instances belonging to a business management element are a worker and a material lot, but the work performance table 200 may have columns for instances belonging to other types of business management elements instead of or in addition to one or both of the worker and the material lot. Furthermore, there may be a work performance table 200 for each business management element.

[0032] The process ID 201 represents the ID of the process. The individual ID 202 represents the ID of the product (individual). The date and time 203 represents the date and time. The date and time may be divided into the start date and time and the end date and time of the process. The worker name 204 represents the ID of the worker who worked on the product in the process at the date and time. The material lot ID 205 represents the ID of the material lot of the product on which the worker worked in the process at the date and time.

[0033] FIG. 3 shows an example of the configuration of the maintenance record table 300 .

[0034] The maintenance record table 300 is included in the performance data 151. The maintenance record table 300 represents the maintenance performance. The maintenance record table 300 has a record for each maintenance performance. The record has information such as a facility ID 301, a date and time 302, and a content 303.

[0035] The equipment ID 301 indicates the ID of the equipment to be maintained. The date and time 302 indicates the date and time when the maintenance was performed. The details 303 indicate the details of the maintenance.

[0036] FIG. 4 shows an example of the configuration of the change history table 400 .

[0037] The change history table 400 is included in the performance data 151. The change history table 400 represents the change performance. The change history table 400 has a record for each change performance. The record has information such as a date and time 401, a process ID 402, an individual ID 403, an object 404, and a phenomenon 405.

[0038] The date and time 401 indicates the date and time when the change was made. The process ID 402 indicates the ID of the process where the change occurred. The individual ID 403 indicates the ID of the product related to the change. The object 404 indicates the business management element where the change occurred or the instance belonging to it. The phenomenon 405 indicates the content of the change.

[0039] FIG. 5 shows an example of the configuration of the measurement value result table 500 .

[0040] The measurement value result table 500 is included in the result data 151. The measurement value result table 500 represents measurement results. The measurement value result table 500 has a record for each measurement result. The record has information such as an individual ID 501, a date and time 502, and a measurement value 503. A measurement value 503 may exist for each result factor such as a KPI. Therefore, for example, the measurement values ​​503 include a measurement value 1 (503A), a measurement value 2 (503B), and so on. For example, the measurement value 1 may be a temperature based on a measurement value from one or more temperature sensors. The measurement value 2 may be a humidity based on a measurement value from one or more humidity sensors.

[0041] FIG. 6 shows an example of the configuration of a condition management table 600 .

[0042] The condition management table 600 is included in the condition management data 152. The condition management table 600 represents various conditions for each change point, including the extraction condition for that change point. The condition management table 600 has a record for each change point. The record has information such as a change point management ID 601, a table name 602, a change point target column name 603, a category condition column name 604, a category condition 605, a linking flag 606, and an extraction classification 607.

[0043] The change point management ID 601 represents the ID of the change point management. The table name 602 represents the table name of the table from which the change point was identified. The change point target column name 603 represents the column name from which the change point was identified. The category condition column name 604 represents the column name of the column that contains information that corresponds to the change point category of the change point. The category condition 605 represents information that corresponds to the change point category of the change point. The linking flag 606 represents whether or not linking to an individual ID is required. The extraction classification 607 represents the conditions that correspond to the change point.

[0044] According to the condition management table 600 illustrated in FIG. 6, for example, change points are extracted as follows:

[0045] According to the change point management ID 601 "001," the table name 602 is "work history," the change point target column name 603 is "worker name," the category condition column name 604 is "process ID," and the extracted classification 607 is "monitor immediately after change." Therefore, in the work history table 200 illustrated in FIG. 2 , if the worker name 204 and the worker name 204 immediately following it are different for the same process, a set of the immediately following worker name 204 and the date / time 203 and individual ID 202 corresponding to the immediately following worker name 204 for that process is extracted as a change point. Note that, because it is possible to identify the individual ID 202 from the work history table 20 in this way, the linking flag 606 for the change point management ID 601 "001" is "unnecessary." Furthermore, in the work record table 200, the process ID 201 column and the worker name 204 column do not contain multiple values ​​(multiple process IDs or multiple worker names), so there is no need to specify the category condition 605 ("-").

[0046] According to the change point management ID 601 "002," the table name 602 is "maintenance record," the change point target column name 603 is "content," the category condition column name 604 is "equipment ID," the category condition 605 is "inspection," and the extracted classification 607 is "close monitoring of long-term progress." Therefore, in the maintenance record table 300 illustrated in FIG. 3 , if the date and time indicated by the date and time 302 corresponding to the content 303 "inspection" has passed a predetermined period of time since the date and time corresponding to the immediately preceding "inspection," the combination of the date and time 302 corresponding to the content 303 "inspection," the equipment ID 301 corresponding to the content 303 "inspection," and the content 303 "inspection" is extracted as a change point. Note that, because the linking flag 606 for the change point management ID 601 "002" is "needed," an individual ID is linked to this extracted change point. For example, the maintenance record corresponding to the content 303 "inspection" performed most recently in the date and time 203 in the work history table 200 is linked to the individual ID of each record in the work history table 200. Here, "most recent" refers to the maintenance record corresponding to the content "inspection" whose date and time 302 is earlier and closest to the date and time 203.

[0047] The contents of the records set in the condition management table 600 may be input, for example, through the visualization system 130. Furthermore, the extracted classification 607 may affect the calculation of the change point score (for example, a parameter (for example, a coefficient) in the formula for the change point score). For example, the extracted classification 607 may be "focusing on immediately after the change" or "focusing on long-term progress." The tendency of the change point score differs depending on whether the extracted classification 607 is "focusing on immediately after the change" or "focusing on long-term progress." The calculation of the change point score will be described later.

[0048] FIG. 7 shows an example of the configuration of a change point process management table 700 .

[0049] The change point process management table 700 is included in the process management data 154. The change point process management table 700 has a record for each change point management ID. The record has information such as a change point management ID 701, a process ID 702, and a classification 703. The change point management ID 701 represents the ID of the change point management. The process ID 702 represents the process ID. The classification 703 represents the business management element to which the change point belongs.

[0050] FIG. 8 shows an example of the configuration of a measurement value process management table 800 .

[0051] The measurement value process management table 800 is included in the process management data 154. The measurement value process management table 800 has a record for each measurement value as a factor. The record has information such as a table name 801, a column name 802, a process ID 803, and a classification 804. The table name 801 indicates the table name. The column name 802 indicates the column name of the measurement value. The process ID 803 indicates the process ID. The classification 804 indicates the business management element to which the measurement value relates.

[0052] FIG. 9 shows an example of the configuration of a change point table 900 .

[0053] The change point table 900 is included in the change point data 153. The change point table 900 has a record for each product. The record has information such as an individual ID 901 and a change point score 902 for each change point. In this embodiment, for each instance (value) of the category condition column name 604 in the condition management table 600 shown in FIG. 6 , a change point that matches the change point target column name 603 and the category condition 605 corresponds to one column of the change point score 902. For example, the change point management ID 601 "001" corresponds to a worker change for each process ID (i.e., the column for change point score 902A, the column for change point score 902B, ...). The conversion store management ID "002" corresponds to the content for each equipment ID (e.g., the column for equipment X inspection 902C).

[0054] The individual ID 901 represents the ID of the product. The change point score 902 represents a numerical value as the change point score. For example, a process A worker change 902A represents the change point score as a worker change in process A. An equipment X inspection 902C represents the change point score as an inspection of equipment X. A material C lot 902E represents the change point score as a change in the material C lot.

[0055] FIG. 10 shows an example of the configuration of the item management table 1000 .

[0056] The item management table 1000 is included in the item management data 155. The item management table 1000 has a record for each item corresponding to an element of output data. The record has information such as a table name 1001, a column name 1002, and a category 1003.

[0057] The table name 1001 indicates the table name. The column name 1002 indicates the column name. The category 1003 indicates whether the information in the column name belongs to a change point or to a change point association.

[0058] An example of the processing performed in this embodiment will be described below.

[0059] FIG. 11 shows the flow of the change point data generation process.

[0060] The change-point data generation unit 172 sets a change-point extraction period (S1101). This period may be a period designated in advance, or may be a period designated by the user via the visualization system 130.

[0061] The change point data generation unit 172 performs steps S1102 to S1104 for each data item for which change points are to be extracted. The "data item for which change points are to be extracted" is an instance of a column corresponding to the category condition column name 604 for each change point ID in the condition management table 600.

[0062] The change point data generation unit 172 identifies one or more change points that meet the extraction conditions from the table represented by the table name 602 in the condition management table 600 for each change point management ID 601 (S1102). The "extraction conditions" include a change point target column name 603, a category condition column name 604, and a category condition 605. Specifically, the change point data generation unit 172 identifies one or more change points for the change point category classification from the reference range corresponding to the change point target column name 603 and the category condition 605 for each instance of the category condition column name 604 in the table represented by the table name 602. Note that records related to the reference range in the table represented by the table name 602 are records having dates and times 203, 302, 401, or 502 that represent dates and times that belong to the change point extraction period. Furthermore, if there are no change points that meet the extraction conditions, S1103 and S1104 are skipped for the data item.

[0063] The change point data generation unit 172 links an individual ID to each change point (S1103). One or more individual IDs can be linked to each change point. For example, for one change point category classification, if work on multiple products occurs between a certain change point and the next change point in a certain process, the individual IDs of each of the multiple products are linked to that change point. Specifically, individual IDs are linked as follows, for example: If the table represented by the table name 602 is the work performance table 200, the individual ID 202 is linked to an element that can be a change point category (e.g., the worker name 204 or the material lot ID 205). If the table represented by the table name 602 is the change history table 400, the individual ID 403 may be linked to an element that can be a change point category (e.g., an element identified from the object 404 and the phenomenon 405). The individual ID 202 corresponding to the change point category is the linked individual ID. Note that linking of individual IDs may be performed only when the linking flag 606 is "required." Alternatively, the value of the linking flag 606 may be "not required" only when an individual ID is linked to the table represented by the table name 602 with respect to the change point management ID, so that linking of individual IDs is essentially always required. When the table represented by the table name 602 is the maintenance record table 300, an individual ID is not linked to an element (e.g., content 303) that can be a change point category. When the table represented by the table name 602 is the change history table 400, an individual ID 403 may not be linked to an element that can be a change point category. The individual ID 202 or 501 corresponding to the date and time 203 or 502 that matches the date and time 302 or 401 corresponding to the change point category is the linked individual ID.

[0064] For each change point, there is a calculation point set as one or more calculation points from the change point to the next change point. For each change point, the calculation point set includes the change point but does not include the next change point. For each change point, the first calculation point in the calculation point set is the change point. For each change point, each calculation point in the calculation point set may be a time or an associated individual ID, and is associated with the change point category of the change point. The change point data generation unit 172 calculates a change point score for each calculation point, and records the individual ID 901 associated with the calculation point and the calculated change point score 902 in the change point table 900 (S1104). The calculation of the change point score follows the extracted classification 607 corresponding to the change point identified in S1102. The method of calculating the change point score will be described later.

[0065] When steps S1102 to S1104 are completed for all data items targeted for change point extraction, the change point table 900 is complete. For example, a change point score for a change in worker for each process ID is generated for a change point with a change point management ID of "001," so that a process A worker change 902A, a process B worker change 902B, etc. are generated in the change point table 900. For a change point with a change point management ID of "002," a change point score for the content "inspection" is generated for each equipment ID, so that an equipment X inspection 902C, an equipment Y inspection (not shown), etc. are generated in the change point table 900. The change point data generation unit 172 stores the change point table 900 in the storage device 52 (S1105). In S1105, for example, the change point table 900 in the work area of ​​memory is stored in a persistent storage device.

[0066] FIG. 12 shows the flow of the item selection process.

[0067] The item selection unit 173 sets a target period (S1201). The "target period" is a period to be analyzed or visualized. This period may be a period specified in advance, or may be a period specified by the user through the visualization system 130.

[0068] The item selection unit 173 performs steps S1202 to S1204 for each change point data item. The "change point data item" is the column name represented by the change point score 902 in the change point table 900.

[0069] The item selecting unit 173 determines whether any of the change point scores 902 is equal to or greater than a predetermined threshold (S1202). Note that the range of values ​​that the change point scores can take may be the same regardless of the change point score 902 (for example, a range of 0 to 1), or may be different. The "threshold" may be common to all change point scores 902, or may be prepared for each change point score 902.

[0070] If the determination result in S1202 is true (S1202: Yes), the item selection unit 173 adds a change point record corresponding to the change point score 902 to the item management table 1000 (S1203). A "change point record" is a record whose category 1003 is "change point". The column name 1002 in the change point record is the column name of the change point score 902 that has a change point score equal to or greater than the threshold. If there is a change point association for the change point corresponding to the change point score 902, the item selection unit 173 adds a change point association record corresponding to the change point association to the item management table 1000 (S1204). A "change point association record" is a record whose category 1003 is "change point association". Specifically, if there is a process ID 803 and a classification 804 in the measurement value process management table 800 that match the process ID 702 and classification 703 corresponding to the change point score 902, the item selection unit 173 records the table name 801 and column name 802 corresponding to the process ID 803 and classification 804 in the item management table 1000 as table name 1001 and column name 1002.

[0071] When steps S1202 to S1204 have been completed for all change point data items, the item management table 1000 is completed. The item selection unit 173 stores the item management table 1000 in the storage device 52 (S1205). In S1205, for example, the item management table 1000 in the work area of ​​memory is stored in a persistent storage device.

[0072] The calculation of the change point score will be described in detail below.

[0073] For each calculation point, a change point score is calculated for the change point category of the calculation point. The change point category is determined according to the category condition column name 604 and the category condition 605. That is, for example, for huMan, the change point category classification is worker, and the change point categories are workers A, B, C, .... For Machine, the change point category classification is equipment, and the change point categories are equipment X, Y, Z, ....

[0074] The change point score corresponds to the likelihood of a negative impact on results (e.g., productivity or quality), typically the likelihood of an anomaly occurring. In this embodiment, the higher the change point score, the greater the likelihood of a negative impact on results.

[0075] The change score for each calculated point reflects at least one of the following three score elements: (a) First score element: the change score of the previous time; (b) Second score element: the degree of risk reduction; and (c) Third score element: the degree of risk increase.

[0076] The "previous change point score" is the change point score of the calculation point immediately preceding the calculation point in question, and / or the change point score of the change point immediately preceding the current change point to which the calculation point in question belongs, for the change point category associated with the change point score of the calculation point in question.

[0077] The higher the risk reduction, the less likely it is to adversely affect the outcome, and therefore the lower the change point score tends to be. The shorter the time interval from the previous calculation point (and / or the elapsed time from the previous change point to the current change point or current calculation point), the higher the risk reduction tends to be. For example, the shorter the time elapsed since the previous maintenance was performed, the higher the risk reduction tends to be (e.g., a reduction in the risk of failure).

[0078] The higher the risk increase, the greater the likelihood of a negative impact on the results, and therefore the higher the change point score tends to be. The risk increase tends to be higher the longer the time interval from the previous calculation point (and / or the elapsed time from the previous change point to the current change point or current calculation point). For example, the longer the time that has elapsed since the previous maintenance was performed, the higher the risk increase may tend to be (e.g., an increase in the risk of failure, an increase in the degree of contamination of the paint nozzle, etc.).

[0079] The calculation of the change point score (for example, the trend of the change point score) is affected by the extracted classification 607. Below, an example will be taken where the process is Process A, the change point category classification is worker, and the extracted classification 607 is “gazing immediately after the change.”

[0080] FIG. 13 shows a schematic example of the change-point score.

[0081] The calculation method (calculation method of change point score) for process A when the change point category is any of workers A, B, C, ... and the extracted classification 607 is "watch immediately after change" follows the calculation formula for the change point score below. Note that the calculation method of the change point score may differ depending on the classification 804 (change point category classification) and the extracted classification 607. For example, depending on the classification 804 and the extracted classification 607, some of the first to third score elements described above may not be reflected in the change point score. R ｃ (n) = α(Δt ｃ，ｎ ) R ｃ (n-1)-β(Δt ｃ，ｎ ) + γ(Δt ｃ，ｎ )

[0082] R ｃ (n) is the change-point score of the nth calculation point of the change-point category c. ｃ，ｎ is the time interval from the (n-1)th calculation point. The coefficients α, β, and γ may be constants, or may be variable values ​​according to the time interval (Δt), as shown in graph 1350 illustrated in FIG. 13 (the horizontal axis is Δt, and the vertical axis is the coefficient value). For example, α and β may be smaller as Δt increases, and γ may be a constant until Δt exceeds a certain value, and once Δt exceeds the certain value, γ may be larger as Δt increases.

[0083] α (Δtｃ，ｎ ) corresponds to the forgetting degree and may take a value between 0 and 1, for example. ｃ，ｎ ) is Δt ｃ，ｎ (That is, a long time interval means forgetting, so the influence of the change point score at the (n-1)th calculation point becomes smaller.)

[0084] β (Δt ｃ，ｎ ) corresponds to the proficiency level, and Δt ｃ，ｎ The smaller the change point score, the greater the probability of failure (i.e., a shorter time interval means that the user will become accustomed to the change point). Familiarity is an example of risk reduction. The higher the familiarity, the lower the risk of abnormality occurring, and therefore the lower the change point score tends to be.

[0085] γ (Δt ｃ，ｎ ) is the initial value according to the forgetting degree. ｃ，ｎ ) is Δt ｃ，ｎ The larger the initial value according to the degree of forgetting, the greater the risk of abnormality occurrence, and therefore the change point score tends to be high.

[0086] 13, the horizontal axis represents calculation points (e.g., time or individuals) for process A, and the vertical axis represents the change point score "change in process A worker." Along this horizontal axis, the change point category (worker A, B, or C) for process A and the change point score are shown for each calculation point. The display mode of the change point category (color, pattern, and line type of the block), the display mode of the change point score corresponding to the change point category (color, pattern, and line type of the circle), and the display mode of the time interval Δ corresponding to the change point category (line type of the arrow) are all consistent.

[0087] As shown in graph 1300, for each change point category, the change point score is calculated not only for each of the one or more change points, but also for each of one or more calculation points that include one or more change points. In other words, for each calculation point, the change point data generation unit 172 calculates, as a change point score, the risk of adversely affecting the results for the change point category related to that calculation point in step A.

[0088] According to the example shown in FIG. 13, there are change points 1311, 1312 and 1313.

[0089] The change point 1311 means that the operator of process A has changed to operator B for the first time. Therefore, the change point score 1301 (change point category is operator B) corresponding to the change point 1311 is γ(Δt ｃ，ｎ ) is highly influenced by

[0090] After that, as shown by the arrow 1302, worker B continues to work on process A. ｃ，ｎ ) becomes larger, but β(Δt ｃ，ｎ ) becomes larger, resulting in a gradually lower change point score.

[0091] The change point 1312 means that the worker in process A has been changed to worker C for the first time in a long time. ｃ，ｎ Therefore, the change point score 1303 corresponding to the change point 1312 (the change point category is worker C) is γ(Δt ｃ，ｎ ), it is higher than the change point score 1323 of worker C in the previous session.

[0092] The change point 1313 indicates that the worker in process A has returned to worker B. Since the worker has returned to worker B after a certain period of time, albeit in a short period of time, α(Δt ｃ，ｎ ) and β(Δt ｃ，ｎ ), as a result, the change point score 1304 (change point category is worker B) corresponding to the change point 1313 is not very different from the immediately preceding change point score 1324 of worker B.

[0093] FIG. 14 shows an example of generating output data for analysis or visualization.

[0094] The output data generation unit 174 acquires, for each record in the item management table 1000, columns corresponding to the table names 1001 and 1002 in the record from the performance data 151 and / or change point data 153, and generates an output table 1400. The output table 1400 is included in the output data to the analysis system 120 or the visualization system 130.

[0095] The output table 1400 is composed of a column for individual ID 1401, a column for date and time 1402, acquired columns (a column for change point association 1403 and a column for change point 1404), and a column for KPI 1405. KPI 1405 may be, for example, a binary value of "0" indicating good (non-defective product) or "1" indicating bad (defective product). Either the column for individual ID 1401 or the column for date and time 1402 may be omitted. The column for change point association 1403 corresponds to the table names 1001 and 1002 in the item management table 1000 and is acquired from the performance data 151 (e.g., the measurement value performance table 500). The column for change point 1404 corresponds to the table names 1001 and 1002 in the item management table 1000 and is acquired from the change point data 153 (e.g., the change point table 900).

[0096] The change-point related 1403 is typically a measured value, i.e., a numerical value. The change point 1404 is also a numerical value. The KPI 1405 is an example of a target variable and may be calculated by a regression equation using one or more explanatory variables. In addition to the measured value, a change-point score can be used as an explanatory variable. That is, with a simple change such as adding an explanatory variable (without substantially changing the calculation method of the KPI 1405), the output data generation unit 174 can accurately calculate the KPI 1405. Note that the change-point related 1403, such as a measured value, may not be used as an explanatory variable, and only the change point 1404 may be used.

[0097] Furthermore, all change point scores and all measurement values ​​linked to an individual ID may be used as explanatory variables corresponding to that individual ID (i.e., the output table 1400 may be configured with a column for all change point scores and a column for all measurement values), but as illustrated in FIG. 14 , the change point scores and measurement values ​​used as explanatory variables may be narrowed down to the change point scores of interest and their associated measurement values. According to FIG. 12 , the change point scores and measurement values ​​recorded in the item management table 1000 are narrowed down to the change point scores corresponding to S1202: Yes and their associated measurement values. In this way, analysis or visualization is narrowed down to the change points of interest and their associated measurement values, thereby enabling low-load, high-precision analysis or efficient visualization.

[0098] There may be multiple KPIs such as quality and lead time, and an output table may be generated for each KPI.

[0099] FIG. 15 shows an example of visualization.

[0100] The visualization system 130 displays a management screen 1500 based on the output data from the data management system 100. The management screen 1500 is, for example, a GUI (Graphical User Interface).

[0101] The management screen 1500 includes a visualization specification UI (User Interface) 1510 and a visualization UI 1520 .

[0102] The visualization specification UI 1510 displays GUI components, such as check boxes, for selecting item options for each item type. Item types include process, change point score, measurement value, and KPI. Via the visualization specification UI 1510, the user specifies, for each item type, the process, change point (change point score), measurement value, and KPI to be visualized. From the output data, options belonging to the process, the change point, the measurement value, and the KPI are identified, and the identified options are displayed in a selectable manner. For example, if the output data includes only the output table 1400 illustrated in FIG. 14 , the process option is only "Process A," the change point option is "Process A Operator Change," the measurement value options are only "Measurement Value 1" and "Measurement Value 2," and the KPI option is only "Quality."

[0103] The visualization UI 1520 visualizes change points (change point scores). Specifically, the visualization UI 1520 visualizes information about options selected through the visualization specification UI 1510. According to the example shown in FIG. 15 , options such as "Process A," "Process A Operator Change," "Equipment X Maintenance," "Measurement Value 1," and "Quality" are selected through the visualization specification UI 1510. Therefore, the visualization UI 1520 displays the change point scores, measurement value 1, and KPI for "Process A Operator Change" and "Equipment X Maintenance" on a graph with the horizontal axis representing time or individuals and the vertical axis representing KPI, measurement value, or change point score.

[0104] Note that "Process A Operator Change" is displayed above "Equipment X Maintenance," but this is to prevent the display range of "Process A Operator Change" from overlapping with the display range of "Equipment X Maintenance." However, since the ranges of the change point scores for "Process A Operator Change" and "Equipment X Maintenance" are the same (for example, a range from 0 to 1), their display ranges may overlap. In other words, multiple change point scores may be displayed in the same coordinate system.

[0105] Although one embodiment has been described above, this is merely an example for explaining the present invention, and the scope of the present invention is not limited to this embodiment. The present invention can be implemented in various other forms.

[0106] For example, in the embodiment, the business is a manufacturing industry and the element corresponding to the individual ID is a product, but the present invention is applicable to businesses other than manufacturing, and the element corresponding to the individual ID is not limited to a product. For example, if the business is a general company order-taking business, the individual ID may be an order number or a sales order number, etc. Furthermore, if the business is a logistics business, the individual ID may be a package number, a container (e.g., a pallet) number, a truck number, etc.

[0107] Furthermore, the data configuration to be managed need not be limited to the data configuration exemplified in the embodiment. For example, the change point data 153 and the performance data 151 may be associated with the graph data exemplified in FIG. 16 . In the graph represented by the graph data, each node is a business element (e.g., a process) in a business, or a business management element in 4M or 5M1E. Each edge (e.g., a directed edge) represents a relationship between a business element and a management element. A column of change points (change point scores) in the change point table 900 or a column of measurement values ​​in the measurement value performance table 500 may be associated with the node.

[0108] The above description can be summarized, for example, as follows: The following summary may include supplementary explanations and explanations of modifications of the above description.

[0109] A data management system (for example, data management system 100) includes an interface device (for example, interface device 51), a storage device (for example, storage device 52), and a processor (for example, processor 53) connected thereto.

[0110] The interface device is communicatively coupled to a data source (e.g., the shop floor system 110) and an external system (e.g., the visualization system 130 and / or the analysis system 120).

[0111] A data source is a source of performance data (e.g., performance data 151) that represents the performance of a business (e.g., manufacturing industry). The business involves management element instances (e.g., workers A, B, ..., equipment X, Y, ...) that are instances of one or more types of business management elements in a management element set (e.g., 4M or 5M1E) that includes at least one type of business management element from 4M (Machine, Material, Method, human Man).

[0112] The processor acquires performance data from a data source via an interface device and stores the performance data in a storage device. The processor identifies one or more change points in business performance for each of one or more change point items from the performance data. A "change point item" may also be called a type or classification of a change point, and may be, for example, "change in worker in process A," "change in worker in process B," "inspection of equipment X," ... identified from a record in the condition management table 600.

[0113] The processor calculates a change point score, which is a numerical value, for each of one or more calculation points including one or more identified change points of one or more change point items, for the management element instance (e.g., change point category) related to the calculation point. All of the one or more calculation points may be change points, or the one or more calculation points may include calculation points other than change points.

[0114] The processor generates numerical data (e.g., output data including output table 1400) including one or more calculated change point scores for each of one or more change point items among the one or more change point items, and outputs the numerical data to an external system.

[0115] An external system performs visualization or analysis based on the numerical data. The numerical data contributes to the visualization or analysis. Therefore, it can support analysis or visualization that contributes to improving business results. Specifically, for example, by extracting change points from performance data, it is expected that all factors that may affect the results will be extracted without omission, thereby supporting analysis or visualization. Furthermore, for example, it is easy to analyze the impact of change points on results (e.g., productivity and quality) using systematic methods such as machine learning.

[0116] The change point score for each calculation point may be based on at least one of the following (a) to (c): (a) for the management element instance associated with the change point score of the calculation point, the change point score of the calculation point immediately preceding the calculation point and / or the change point score of the change point immediately preceding the current change point to which the calculation point belongs, (b) for the management element instance associated with the change point score of the calculation point, the risk reduction degree that tends to decrease as the time interval from the previous calculation point to the calculation point and / or the time interval from the previous change point to the calculation point or the current change point becomes shorter, and (c) for the management element instance associated with the change point score of the calculation point, the risk increase degree that tends to increase as the time interval from the previous calculation point to the calculation point and / or the time interval from the previous change point to the calculation point or the current change point becomes longer.

[0117] As a result, an appropriate numerical value can be expected as the change point score. Specifically, the importance of a change point depends on the time interval (elapsed time) from the occurrence of a previous change such as "for the first time in a long time," "for the first time," or "aging deterioration," and since this dependency is reflected in the change point score, an appropriate numerical value can be expected as the change point score. Note that the "previous change point" may be the change point immediately before the current change point or an earlier change point. Similarly, the "previous calculation point" may be the calculation point immediately before the calculation point or an earlier calculation point. Note that an example of (a) is the above-mentioned R ｃ (n-1). An example of (b) is the above-mentioned β(Δt ｃ，ｎ ) An example of (c) is the above-mentioned α(Δt ｃ，ｎ ) and γ(Δt ｃ，ｎ In the embodiment, for each calculation point, in the calculation of the change point score, Δt ｎ is the time interval from the (n-1)th calculation point to the nth calculation point, while Δt ｎ may be the time interval from the (nm)th calculation point (m is any natural number) to the nth calculation point.

[0118] At least one of one or more change point items may be associated with a focused element (e.g., extracted classification 607). For each change point item, if a focused element is associated with the change point item, the processor may determine a trend of the change point score according to the associated focused element and calculate a change point score according to the determined trend. This may result in an appropriate value being obtained as the change point score. The determination of the focused element associated with the change point item may be performed, for example, in S1104, and the change point data generation unit 172 may determine a trend of the change point score according to the determined focused element and calculate a change point score according to the determined trend. "Determining the trend of the change point score" may mean determining a value to be entered into one or more parameters in a change point score calculation formula and entering the value, or may mean selecting a calculation formula corresponding to the focused element from among multiple change point score calculation formulas. The calculation formula may be an example of a calculation method. The calculation method may involve using a machine learning model that outputs the change point score and inputs values ​​(e.g., explanatory variables) as factors of the change point score, or the change point score may be obtained by inputting values ​​as factors into such a machine learning model.

[0119] When there is a change point item associated with "immediately after a change" as an element to be watched (for example, a change point item associated with "watch immediately after a change"), the determined tendency of the change point score for each calculation point for the change point item may be a tendency that satisfies the following (x) and (y) for the management element instance associated with the change point score of the calculation point. In the case of watching immediately after a change, the shorter the time interval, the higher the degree of risk reduction such as proficiency, and therefore the smaller the change point score tends to be. Meanwhile, the longer the time interval, the higher the degree of risk increase such as forgetfulness, and therefore the larger the change point score tends to be. As a result, an appropriate numerical value can be expected for the change point score. (x) The shorter the time interval from the calculation point previous to the calculation point and / or the time interval from the change point previous to the current change point to which the calculation point belongs to and the calculation point or the current change point, the smaller the change point score. (y) The longer the time interval from the previous calculation point to the calculation point and / or the time interval from the previous change point to the calculation point or the current change point, the larger the change point score.

[0120] When there is a change point item associated with long-term progress as an element to be monitored, the determined tendency of the change point score for each calculation point for that change point item may be larger the longer the time interval from the calculation point previous to that calculation point to that calculation point and / or the time interval from the change point previous to the current change point to which that calculation point belongs to to that calculation point or the current change point, for the management element instance associated with the change point score of that calculation point. In the case of monitoring long-term progress, the longer the time interval between executions (e.g., maintenance executions) related to the management element instance, the higher the degree of risk increase, and therefore the change point score tends to be larger, and as a result, an appropriate numerical value can be expected as the change point score.

[0121] The processor may select, from one or more change point items, a change point item for which a change point score satisfying a predetermined condition has been calculated. One or more of the one or more change point items may be selected change point items. This allows the change point items with change point scores included in the numerical data to be narrowed down to appropriate change point items. Therefore, even if the actual data is massive, narrowing down to the change point items of interest for analysis or visualization is expected to improve the accuracy of the user's judgment when viewing the visualized information (e.g., making it easier to understand the data) and / or improve the accuracy of the analysis and reduce the amount of calculation required for analysis. Note that the "predetermined condition" for the change point score may be equal to or greater than a threshold, as in the embodiment, or may be another condition. Furthermore, the "threshold" may be a fixed value, or the maximum or average value of one or more change point scores belonging to the change point items.

[0122] In the performance data, an individual ID, which is the ID of a management element entity belonging to Material, may be associated with a management element entity related to business performance or a date and time for the management element entity. For each calculation point, the processor may identify from the performance data an individual ID associated with the management element entity associated with the change point score of the calculation point or an individual ID associated with the date and time for the management element entity, and associate the identified individual ID with the calculation point. For each change point score, the numerical data may include an individual ID associated with the calculation point corresponding to the change point score. In this way, the numerical data includes, in addition to the change point score, an individual ID associated with the calculation point corresponding to the change point score, which can contribute to visualization and / or analysis.

[0123] The performance data may include a measurement value for each individual ID for one or more measurement items. The processor may acquire measurement values ​​for one or more measurement items among one or more measurement items (e.g., measurement value 1, measurement value 2, ...). The numerical data may include, for each individual ID, one or more change point scores for one or more change point items as well as one or more acquired measurement values ​​for one or more measurement items. This can further contribute to visualization and / or analysis.

[0124] The processor may calculate a response variable (e.g., a KPI) for each individual ID using one or more change point scores for one or more change point items and one or more measurement values ​​for one or more measurement items as multiple explanatory variables. The numerical data may further include the calculated response variable for each individual ID. This can further contribute to visualization and / or analysis.

[0125] The external system may be a visualization system (e.g., visualization system 130) that performs visualization based on numerical data. Visualization based on numerical data may display, on a coordinate system with the time series of calculation points as the horizontal axis, a list of change point scores for at least one of one or more change point items (in the example of FIG. 15 , “Process A Operator Change” and “Equipment X Maintenance”), a list of measurement value scores for at least one of one or more measurement items (in the example of FIG. 15 , “Measurement Value 1”), and a list of objective variables (in the example of FIG. 15 , “KPI (Quality)”). This is expected to improve the accuracy of user judgment. Note that “at least one of one or more change point items” may include a predetermined change point item or may include a change point item selected by the user from among the one or more change point items. Similarly, “at least one of one or more measurement items” may include a predetermined measurement item or may include a measurement item selected by the user from among the one or more measurement items.

[0126] The external system may be an analysis system that performs analysis based on numerical data having explanatory variables and objective variables for each individual ID. Because the generated and output numerical data has a configuration that can be interpreted by the analysis system, it is expected that the accuracy of analysis by the analysis system will improve with little (e.g., virtually no) design change to the analysis system.

[0127] 100: Data management system

Claims

1. An apparatus comprising: an interface device communicatively connected to a data source and an external system; a storage device; and a processor connected to the interface device and the storage device, wherein the data source is a source of performance data that represents the performance of a business operation, wherein the business operation involves management element instances that are instances of one or more types of business management elements in a management element set that includes at least one type of business management element out of 4M (Machine, Material, Method, human Man), and wherein the processor: acquires the performance data from the data source through the interface device; stores the acquired performance data in the storage device; identifies one or more change points in the performance of the business operation for each of one or more change point items from the performance data; and calculates a change point score, which is a numerical value, for each of one or more calculation points that include the identified one or more change points of the one or more change point items, for the management element instance related to the calculation point, generating numerical data including one or more calculated change point scores for each of the one or more change point items among the one or more change point items, and outputting the numerical data to the external system.

2. The data management system of claim 1, wherein the change point score for each calculation point is based on at least one of the following (a) to (c): (a) for the management element instance associated with the change point score of the calculation point, the change point score of the calculation point immediately preceding the calculation point and / or the change point score of the change point immediately preceding the current change point to which the calculation point belongs; (b) for the management element instance associated with the change point score of the calculation point, the risk reduction degree tends to become smaller as the time interval from the previous calculation point to the calculation point and / or the time interval from the previous change point to the calculation point or the current change point becomes shorter; and (c) for the management element instance associated with the change point score of the calculation point, the risk increase degree tends to become larger as the time interval from the previous calculation point to the calculation point and / or the time interval from the previous change point to the calculation point or the current change point becomes longer.

3. The data management system of claim 1, wherein at least one of the one or more change point items is associated with a focused element, and the processor, for each change point item, when a focused element is associated with the change point item, determines a trend of the change point score according to the focused element associated with the change point item, and calculates a change point score according to the determined trend.

4. The data management system of claim 3, wherein when there is a change point item associated with an immediately after change as an element to be watched, the determined tendency of the change point score for each calculation point for the change point item satisfies the following (x) and (y) for the management element instance associated with the change point score of the calculation point: (x) the smaller the shorter the time interval from the calculation point previous to the calculation point and / or the time interval from the change point previous to the current change point to which the calculation point belongs to and the calculation point or the current change point, and (y) the larger the longer the time interval from the previous calculation point to the calculation point and / or the time interval from the previous change point to the calculation point or the current change point.

5. The data management system of claim 3, wherein when there is a change point item associated with a long-term course as an element to be monitored, the determined tendency of the change point score for each calculation point for that change point item is larger for the management element instance associated with the change point score of that calculation point, the longer the time interval from the calculation point previous to that calculation point to that calculation point, and / or the time interval from the change point previous to the current change point to which that calculation point belongs to that calculation point to that calculation point or the current change point.

6. The data management system of claim 1, wherein the processor selects, from the one or more change point items, a change point item for which a change point score that satisfies a predetermined condition has been calculated, and the one or more change point items from the one or more change point items are each a selected change point item.

7. The data management system of claim 1, wherein the performance data associates an individual ID, which is the ID of a management element entity belonging to a Material, with a management element entity related to the business performance or a date and time for the management element entity, and the processor, for each calculation point, identifies from the performance data an individual ID associated with the management element entity associated with the change point score of the calculation point or a date and time for the management element entity, and associates the identified individual ID with the calculation point, and the numerical data includes, for each change point score, an individual ID associated with the calculation point corresponding to the change point score.

8. The data management system of claim 7, wherein the performance data includes a measurement value for each individual ID for one or more measurement items, the processor acquires measurement values ​​for one or more of the one or more measurement items, and the numerical data includes, for each individual ID, one or more change point scores for the one or more change point items as well as one or more acquired measurement values ​​for the one or more measurement items.

9. The data management system of claim 8, wherein the processor calculates a dependent variable for each individual ID using the one or more change point scores for the one or more change point items and the one or more measurement values ​​for the one or more measurement items as multiple explanatory variables, and the numerical data further includes the calculated dependent variable for each individual ID.

10. The data management system of claim 9, wherein the external system is a visualization system that performs visualization based on the numerical data, and the visualization based on the numerical data displays a sequence of change point scores for at least one of the one or more change point items, a sequence of measurement value scores for at least one of the one or more measurement items, and a sequence of target variables in a coordinate system with the time series of calculation points as the horizontal axis.

11. The data management system according to claim 9, wherein the external system is an analysis system that performs analysis based on numerical data having explanatory variables and objective variables for each individual ID.

12. A data management method comprising the steps of: acquiring performance data from a data source; identifying one or more change points in business performance for each of one or more change point items from the performance data; calculating a numerical change point score for a management element instance related to one or more calculation points including the identified one or more change points of the one or more change point items; generating numerical data including the calculated one or more change point scores for each of the one or more change point items of the one or more change point items; and outputting the numerical data to an external system, wherein the data source is a source of performance data that is data representing the performance of the business; and the business involves management element instances that are instances of one or more types of business management elements in a management element set that includes at least one type of business management element out of 4M (Machine, Material, Method, human Man).

13. A computer program that causes a computer to perform the following actions: acquire performance data from a data source; identify one or more change points in business performance for each of one or more change point items from the performance data; calculate a numerical change point score for a management element instance related to one or more calculation points including the identified one or more change points of the one or more change point items; generate numerical data including the calculated one or more change point scores for each of the one or more change point items of the one or more change point items; and output the numerical data to an external system; wherein the data source is a source of performance data that is data representing the performance of the business; and the business involves management element instances that are instances of one or more types of business management elements in a management element set that includes at least one type of business management element out of 4M (Machine, Material, Method, human Man).

Images