Data processing device, method and program

Abstract

To provide a data processing device, a method and a program capable of reducing a burden on a user monitoring manufacturing data.SOLUTION: The data processing device comprises an acquisition unit, a first calculation unit and an output unit. The acquisition unit acquires state data about a state of each product and order data about an order of manufacturing each product. The first calculation unit, using the state data and the order data acquired by the acquisition unit, calculates a first score related to change of the state data acquired by the acquisition unit based on a first change model expressing change of the state data using the order data. The output unit outputs the first score.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a data processing device, a method, and a program. [Background technology]

[0002] In the manufacturing industry, it is important to quickly identify the cause of a product entering a specific state. For example, if a product enters an abnormal state that differs from its normal state, quickly identifying the cause will lead to maintaining and improving yield. Many manufacturing industries use monitoring of various data acquired during the manufacturing process to help detect anomalies and identify their causes.

[0003] The content of the data varies. For example, data about the manufacturing conditions of a product includes the names of materials and equipment used in the manufacturing process. Data about the product's condition includes the size, physical characteristics, and appearance quality of the manufactured product. Usually, this data is associated with information such as an ID or serial number that can identify each individual product.

[0004] By monitoring each item of individual product data, it may be possible to detect abnormalities in products or equipment. For example, if the individual product data values ​​of some products in a group of products manufactured during a certain period are different from normal values, it is possible that an abnormality has occurred in those products. In this case, the manufacturing data, including data on manufacturing conditions, is scrutinized to identify the cause of the product abnormality. For example, if the manufacturing data identifies that the abnormal product was manufactured only on a specific piece of equipment, it is possible that that equipment is the cause of the abnormality.

[0005] The more processes and devices required to complete a product, the more data needs to be monitored. Furthermore, with the recent development of IoT (Internet of Things) technology, various data related to manufacturing can be easily acquired. As a result, the number of manufacturing data items has increased dramatically. This makes manual monitoring of manufacturing data difficult. Given these circumstances, there is a demand for a device that can assist users in monitoring manufacturing data. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Patent Publication No. 2021-071896 Summary of the Invention [Problem to be solved by the invention]

[0007] The present disclosure provides a data processing device, method, and program that can reduce the burden on users of monitoring manufacturing data. [Means for solving the problem]

[0008] A data processing device according to an embodiment includes an acquisition unit, a first calculation unit, and an output unit. The acquisition unit acquires status data relating to the status of each product and sequence data relating to the manufacturing sequence of each product. The first calculation unit uses the status data and sequence data acquired by the acquisition unit to calculate a first score relating to a change in the status data acquired by the acquisition unit based on a first change model that expresses a change in the status data using the sequence data. The output unit outputs the first score. The first calculation unit calculates a first score based on a goodness of fit of the first change model to the state data and the order data and a penalty for the complexity of the first change model. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a block diagram showing a data processing device according to the first embodiment. [Figure 2] FIG. 2 is a diagram illustrating an example of manufacturing data. [Figure 3] FIG. 3 is a flowchart showing a process of calculating the first score in the first calculation unit. [Figure 4] Figure 4 shows an image of the application of the regression model. [Figure 5] FIG. 5 is a diagram illustrating an example of a result of sorting data by the output unit. [Figure 6] FIG. 6 is a diagram showing an example of a diagram generated by visualization data based on the result of applying the first change model to the state data. [Figure 7] FIG. 7 is an example of a display screen of the analysis result displayed on the display unit when both the first score and the application result of the first change model are received. [Figure 8] FIG. 8 is a block diagram illustrating a data processing device according to the second embodiment. [Figure 9] FIG. 9 is a flowchart showing the process of calculating the second score in the second calculation unit. [Figure 10A] FIG. 10A is a diagram showing an example of data generated by the output unit when the manufacturing condition data that caused the change in the status data can be acquired. [Figure 10B] FIG. 10B is a diagram showing an example of data generated by the output unit when the manufacturing condition data that caused the change in the status data cannot be acquired. [Figure 11] This is an example of visualization using a scatter plot that combines sequence data, manufacturing condition data, and status data. [Figure 12] FIG. 12 is an example of a display screen of the analysis results displayed on the display unit when the first score, the application result of the first change model, and the second score are received. [Figure 13] FIG. 13 shows an example of a display screen of the analysis results when a scatter diagram combining sequence data, manufacturing condition data, and status data is displayed. [Figure 14] FIG. 14 is an example of a display screen in the case where the degree of freedom when the first score is calculated matches the number of elements of the manufacturing condition data when the second score is calculated. [Figure 15]FIG. 15 is a block diagram illustrating a data processing device according to the third embodiment. [Figure 16] FIG. 16 is a diagram showing an example of a display screen displayed on the display unit in the third embodiment. [Figure 17] FIG. 17 is a schematic diagram showing the relationship between the information amount change button and the switching button and the amount of information in the second display area according to the third embodiment. [Figure 18] FIG. 18 is a diagram showing an example of a display screen including an interface for accepting an operation for changing the priority of the first display area and an operation for changing the priority of the second display area. [Figure 19] FIG. 19 shows a modified example in which details of the analysis results are displayed. [Figure 20] FIG. 20 is a block diagram showing a data processing device according to the fourth embodiment. [Figure 21] FIG. 21 is a diagram showing the relationship between manufacturing data and manufacturing process data. [Figure 22] FIG. 22 is a diagram showing an example of a display screen displayed on the display unit in the fourth embodiment. [Figure 23] FIG. 23 is a diagram showing the relationship between manufacturing data and user data. [Figure 24] FIG. 24 is a block diagram illustrating a data processing device according to the fifth embodiment. [Figure 25] FIG. 25 is a block diagram showing the hardware configuration of a data processing device. DETAILED DESCRIPTION OF THE INVENTION

[0010] (First embodiment) A first embodiment will be described. A data processing device according to the first embodiment uses order information related to products to infer the cause of a product entering a certain state, and visualizes the cause based on the inferred result. In this way, the data processing device supports a user in discovering the cause of a product entering a certain state.

[0011] 1 is a block diagram showing a data processing device according to a first embodiment. The data processing device 1 includes an acquisition unit 101, a first calculation unit 102, and an output unit 103. The data processing device 1 is connected to a manufacturing database 104 and a display unit 105. The manufacturing database 104 and the display unit 105 may be provided separately from the data processing device 1, or may be provided in the data processing device 1, for example.

[0012] The acquiring unit 101 acquires manufacturing data from the manufacturing database 104. The manufacturing data includes key data for identifying each product, status data relating to the status of each product, and sequence data relating to the manufacturing sequence of each product. The acquiring unit 101 inputs the manufacturing data to the first calculating unit 102.

[0013] The first calculation unit 102 calculates a first score related to a change in the state data using the state data and the order data. The first score is a score that indicates how well the first change model fits the actual state data and order data. The first change model is a model that indicates a change in the state data using the order data.

[0014] The output unit 103 generates visualization data from the first score or information explaining the first score. Then, the output unit 103 outputs the visualization data to the display unit 105. The visualization data is data for visually presenting to the user candidate causes of changes in the status data.

[0015] The manufacturing database 104 stores manufacturing data. The manufacturing database 104 may be configured as a general relational database management system (RDBMS). The manufacturing database 104 may be, for example, a NoSQL (Not only SQL) database. Furthermore, the manufacturing data stored in the manufacturing database 104 may be configured as a file in a predetermined format such as CSV (Comma Separated Value).

[0016] The display unit 105 displays various types of information based on the visualized data and may be configured with various display devices such as a liquid crystal display, an organic EL display, or the like.

[0017] Next, the operation of the data processing device 1 of the first embodiment will be described. First, the operation of the acquisition unit 101 will be described. The acquisition unit 101 acquires manufacturing data used to calculate the first score from the manufacturing database 104. The manufacturing data will be described in detail below. FIG. 2 is a diagram showing an example of manufacturing data. Here, the vertical direction in the manufacturing data of FIG. 2 is called columns, and the horizontal direction is called rows. In this case, the column direction of the manufacturing data represents items of product state and order. Furthermore, the row direction of the manufacturing data represents data for each individual product.

[0018] In the example of FIG. 2, the manufacturing data has an ID as key data. The key data is useful for identifying a specific state, for example, an abnormal state. If product identification is not required, the manufacturing data may not include key data. Furthermore, if the manufacturing data is recorded in a CSV format or the like and does not include any particular key data, the acquisition unit 101 may assign a serial number or the like to the acquired manufacturing data or the like as key data.

[0019] In the embodiment, the acquisition unit 101 acquires at least two or more pieces of manufacturing data that satisfy a condition, i.e., two or more rows. Then, the acquisition unit 101 inputs the manufacturing data to the first calculation unit 102. An acquisition condition for the manufacturing data is, for example, whether or not the ID of the manufacturing data is included in an ID list of products that are the subject of cause identification. In the case of such an acquisition condition, the acquisition unit 101 may acquire manufacturing data having an ID included in the list. Here, the number of manufacturing data acquired by the acquisition unit 101, i.e., the number of rows, is represented as D.

[0020] The conditions for acquiring manufacturing data may be specified by the user. For example, the conditions may be specified to acquire manufacturing data for a group of products manufactured in a period of time such as an arbitrary hour or an arbitrary day. The following further describes the status data and order data, which are elements of manufacturing data.

[0021] First, we will explain the order data. For the purposes of the following explanation, the order data is defined as {O k :k=1...L}, where L represents the number of items in the acquired order data. The products to be analyzed in this embodiment are manufactured according to a specific L order. The order data is data relating to the order in which such products are manufactured. The order data may include, for example, data on the order in which each process in the manufacture of the product begins, i.e., the order in which work on the product is started. In this way, when the order data is data indicating the order in which each process begins, the value of L in the order data is equal to the number of processes.

[0022] Here, the sequence data does not have to be data per process. For example, the sequence data may be data per production line, per manufacturing facility, per manufacturing device, per product lot, per batch, or per production unit. The sequence data may also be data per manufacturing base, etc. For example, if a product travels from factory A to factory B and factory C before arriving at an inspection site, the sequence data may be recorded by the time period during which the product is at each factory and inspection site.

[0023] The order in the order data may be represented by a sequential number starting from 1, or may be represented by time. For example, the order data may be data on the start time of each process in the manufacture of a product, or the time when work on the product begins in each process. O1 and O2 in FIG. 2 are examples in which order data is recorded by sequential numbers, and O in FIG. L-1 , O L is an example in which the order data is recorded by time. As shown in Figure 2, the order data does not necessarily have to be recorded in a uniform format.

[0024] In addition, the sequence data may be, for example, instead of the start time of construction, the time of passage, the time of completion, or a number representing a certain period or a certain number of product units, a character string, a serial number, a manufacturing number, an ID, the number of times the device has been processed, or any other type of data that can determine the sequence of the products.

[0025] Furthermore, in the embodiment, for example, a situation is assumed in which a product is automatically or manually moved by a belt conveyor, a cart, etc., and processed or inspected in predetermined equipment, etc. In addition, for example, even in a situation in which a product is fixed and processed or inspected in different equipment, etc. for each process, the order can be specified, so the technology of the first embodiment can be applied.

[0026] Furthermore, the product associated with the sequence data does not have to be a product that goes through multiple processes. For example, even if the process is not explicitly defined, if the product goes through multiple devices or if multiple operations are performed on the product, the sequence data may be the order in which the product arrived at each device or each work location.

[0027] Furthermore, when the order data is recorded by time, the order data may include data of products that are in the same order due to the influence of the accuracy of the time recording, etc. However, it is desirable that the order data be recorded so that the order is unique.

[0028] Next, the state data will be described. For the purposes of the following discussion, the state data will be defined as {Y i :i=1...N}, where N represents the number of items of acquired state data.

[0029] Status Data Y l represents the status of an individual product, for example. The status of a product is represented, for example, as a measurement value of an inspection performed on the individual product. l includes data on measurements related to inspection items, such as product dimensions and weight. Depending on the type of product, status data Yl may include data on measurements of electrical and physical properties of the product. l may be, for example, a measurement value relating to an inspection item at the time of product shipment.

[0030] Also, the status data Y l The status data Y may be any data other than the measurement values ​​related to the inspection items, which can grasp the status of the product. For example, the status data Y l may be data of the result of some judgment. For example, the status data Y l The acquisition unit 101 acquires any state data Y that the user has determined to be useful for analysis. l You may obtain.

[0031] Here, the state data and the order data may be combined into one piece of data as shown in FIG. 2, or may be separate pieces of data associated with each other by a common ID.

[0032] Next, the operation of the first calculation unit 102 will be described. i and ordered data O k When receiving the first score S1(Y i ,O k ) is calculated. FIG. 3 is a flowchart showing the process of calculating the first score in the first calculation unit 102.

[0033] First, the first calculation unit 102 initializes k and i to initial values, for example, 1. Then, in step S1, the first calculation unit 102 calculates the state data Y i and ordered data O k The first change model f(Y i )=aO k The first change model applies to the ordered data O k Using the state data Y i In the embodiment, the first change model is a model that represents a change in the order data Ok is a regression model with a step basis.

[0034] Figure 4 shows an image of the application of a regression model. Figure 4 (A) shows the application of a regression model to ordered data O k The state data Y is sorted in order of time. i The vertical axis of (A) in Fig. 4 is the distribution of the state data Y i The horizontal axis is the measured value of the order data O k 4B shows the values ​​of the order data O k represents the step basis generated from the time-series state data Y i and the state data Y at time 0 i When the distribution of O is obtained, k The basis function x changes stepwise at time tn (n=0…D) tn Considering this, the state data Y i The change of these basis functions x tn The weights are the weights of the basis functions x tn The first calculation unit 102 calculates the state data Y i and ordered data O k Based on each basis function x tn weights, i.e., the function f(Y i ) to calculate the regression coefficients for the state data Y i and ordered data O k For example, the first calculation unit 102 applies the first change model to the order data O k The ordered state data Y is represented by i The weights are calculated by a method such as the least squares method using the state data Y i 4(c) shows the results when the first change model shown in (B) of Figure 4 is applied to the scatter plot of (A) of Figure 4. In other words, (c) of Figure 4 shows the results when the first change model that best fits the scatter plot of (A) of Figure 4 is y = ax tn This indicates that

[0035] In step S2, the first calculation unit 102 calculates the state data Y i The first score S1(Y i ,O k ) is calculated. i ,O k ) is the state data Y i and ordered data O k The goodness of fit of the first change model to the state data Y i and ordered data O k The first score is a value representing the degree of fit of the first change model to the model. For example, an existing evaluation index such as the residual sum of squares (RSS) or log-likelihood may be used as the first score. Other examples of the first score include a correlation coefficient, a regression coefficient, a mean square error (MSE), a mean root mean square error (RMSE), a mean absolute error (MAE), a coefficient of determination (R2), and indices derived from these evaluation indices. Furthermore, the first score may be an evaluation index calculated from multiple indices among these indices.

[0036] In the embodiment, as a more preferable example, a score indicating the sum of the degree of fit to the data and a penalty for the complexity of the model may be used as the first score. i The changes in include not only the changes due to the original cause of the product reaching a specific state, but also various noises due to events other than the original cause, such as measurement errors and process changes. When the influence of noise is large, if the score is simply based on the high degree of fit, the score of the model that is overly fit to the changes including the noise will be high. In this case, the state data Y due to the cause i It becomes difficult to capture changes in the data. In an embodiment, the Bayesian Information Criterion (BIC), which is known as a score that indicates the sum of the goodness of fit to the data and a penalty for the complexity of the model, is used as the first score. The BIC is calculated using Equation (1). Here, in Equation (1), ln(L) is the maximum logarithmic likelihood, n is the number of samples, and k is the degree of freedom of the data.

number

[0037] Furthermore, under the Gaussian error model, equation (1) can be rewritten as equation (2), where n in equation (2) is the number of samples, Yi true is the state data Y i The measured value of Yi pred is the state data Y calculated from the first change model. i is the predicted value.

number

[0038] In equation (1) or (2), a small BIC value means that the model fits the data well. That is, a small BIC value indicates that the ordered data O k is the state data Y i In this way, the order data O k and state data Y i The first change model f(Y i )=aO k To what extent is the state data Y i The extent to which the change in the

[0039] Returning now to the description of FIG. 3, in step S3, the first calculation unit 102 determines whether k is greater than L. If it is determined in step S3 that k is not greater than L, the first calculation unit 102 adds 1 to k and returns the process to step S1. In step S3, if k is greater than L, that is, if the current state data Y i All ordered data in O k If it is determined that the processing for has been completed, the first calculation unit 102 moves the processing to step S4.

[0040] In step S4, the first calculation unit 102 determines whether i is greater than N. If it is determined in step S4 that i is not greater than N, the first calculation unit 102 adds 1 to i and returns the process to step S1. If it is determined in step S4 that i is greater than N, that is, if all the state data Y i All ordered data in O k 3 is completed, the first calculation unit 102 ends the process of Fig. 3. In this case, the first calculation unit 102 outputs the first score or information explaining the first score obtained in the process of Fig. 3 to the output unit 103. Here, the information explaining the first score may be, for example, the state data Y i The result of applying the first score to the state data Y i The result of applying it to, for example, the state data Y i , ordered data O k and status data Y i and ordered data O k The data may be a function representing a first variation model that best fits the

[0041] In this way, the first calculation unit 102 calculates the state data Y i The first change model is applied to the first score S1(Y i ,O k ) is calculated. This allows the state data Y i The change in the state data Y i The causes of changes can be identified.

[0042] In this embodiment, it is assumed that a causal change occurs once or several times in a certain period, and therefore an index with a penalty for complexity, such as BIC, is preferable. However, the first score is not necessarily limited to BIC. For example, the Akaike Information Criterion (AIC), Deviance Information Criterion (DIC), Generalized Information Criterion (GIC), Widely Available Akaike Information Criterion (WAIC), Widely Available Bayesian Information Criterion (WBIC), or a derivative index thereof may be used as the first score.

[0043] In the embodiment, the step basis is generated from the order data. i Any basis other than the step basis may be used as long as it can capture the change in . For example, a sigmoid basis or a ramp basis may be used, or a Fourier basis or a wavelet basis may be used.

[0044] Furthermore, in the above example, the first change model is a regression model. i For example, if a moving average model is used as the first change model, the state data Y i Noise that is presumed to have little relation to the cause of the change can be removed. Such a first change model can respond to changes that are more likely to be caused. When a moving average model is used as the first change model, a threshold for detecting changes and the number of changes, etc., can be used as the first score instead of BIC, etc. Alternatively, the first score may be calculated directly using a machine learning method such as a support vector machine, deep learning, etc.

[0045] Next, a description will be given of the operation of the output unit 103. After the first calculation unit 102 calculates the first score, the output unit 103 generates visualized data from the first score or information explaining the first score.

[0046] First, the operation of the output unit 103 when receiving the first score S1(Y i ,O k ) is the state data Y i and ordered data O k The output unit 103 outputs the first score S1(Y i ,O k ) for example, order data O kThe output unit 103 may then generate visualized data based on the sorted data and output the generated visualized data to the display unit 105 as an analysis result. In addition, the output unit 103 may k The first score S1(Y i ,O k ) order. The output unit 103 may then generate visualized data based on the sorted data and output the generated visualized data to the display unit 105 as the analysis result. In this case, the first score S1(Y i ,O k ) and ordinal data O k The visualized data can be generated so that the rankings are displayed in a ranked format. Therefore, the user can sort the order data O k You can check the results in order of importance by checking

[0047] FIG. 5 is a diagram showing an example of the result of sorting data by the output unit 103. FIG. 5 shows the first score S1(Y i ,O k ) in ascending order, that is, the state data Y i Order data by likelihood of being related to the cause of change k This shows an example where the first score S1(Y i ,O k ) and ordinal data O k is displayed on the display unit 105, the user can determine the status data Y from the displayed numerical value. i The causes of the changes can be considered.

[0048] Next, the operation of the output unit 103 when receiving the application result of the first change model as information explaining the first change model from the first calculation unit 102 will be described. When receiving the application result of the first change model, the output unit 103 generates visualized data in a format that can be visually recognized by the user, such as a diagram, from the application result of the first change model. By using the application result of the first change model, the user can iThe change point in the distribution of the first score can be confirmed as visual information rather than as a numerical value like the first score. This allows the user to understand the results more intuitively.

[0049] Figure 6 shows the status data Y i 6 shows an example of a diagram generated from visualization data based on the result of applying the first change model to the order data O5 having the smallest first score in FIG. 5. The vertical axis of FIG. 6 represents the state data Y i The horizontal axis represents the time of the order data O5. The scatter diagram 201 in FIG. 6 shows the state data Y i That is, the output unit 103 outputs the state data Y i are sorted in the order of time of the order data O5, and each sorted state data Y i 6 is plotted at the corresponding time points. i 6 is a line graph showing the first change model applied to the state data Y i In this way, it becomes easier for the user to intuitively understand how well the first change model fits the distribution of the state data Y i The degree of change in may be visually indicated.

[0050] Furthermore, when receiving both the first score and the application result of the first change model, the output unit 103 may combine both to generate visualized data that can be displayed. This allows the user to determine the state data Y from the numerical values ​​and the scatter diagram. i The causes of the changes can be investigated.

[0051] The format of the visualization data will now be described. The visualization data can be generated as, for example, image data or drawing data. Alternatively, the visualization data may be in a format that can be displayed on the display unit 105, such as html (Hypertext Markup Language), xml (eXtensible Markup Language), or JSON (JavaScript (registered trademark) Object Notation) format.

[0052] FIG. 7 is an example of a display screen of the analysis results displayed on the display unit 105 when the output unit 103 receives both the first score and the application result of the first change model. The display screen of FIG. 7 is displayed based on visualized data. As shown in FIG. 7, the display screen of the analysis results includes a first display area 300. The first display area 300 includes a creation date display field 301, a condition data name display field 302, and second display areas 303, 304, and 305. Here, FIG. 7 shows the display screen of the analysis results for the condition data Y i This is an example of a display screen when there is one status data Y i If there are two or more, all the state data Y i The first display areas 300 for the respective status data Y i If any score has been calculated for each state data Y, the state data Y is ranked based on the score. i A first display area 300 for may be displayed.

[0053] The creation date display field 301 is a display field for displaying the creation date of the manufacturing data used in the analysis, for example, as information to be analyzed.

[0054] The status data name display field 302 displays the status data Y to be analyzed. i This is a display field for displaying the name of the item.

[0055] The second display areas 303, 304, and 305 correspond to the status data Y displayed in the corresponding first display area 300. i and ordered data O kA maximum of L second display areas can be provided for one first display area.

[0056] Second display areas 303, 304, and 305 include ordinal data name display fields 3031, 3041, and 3051. Second display areas 303, 304, and 305 also include at least one of a display area for the first score and a display area for a diagram showing the application results of the first change model. For example, second display area 303 includes both first score display area 3032 and a display area 3033 for a diagram showing the application results of the first change model. On the other hand, second display areas 304 and 305 only include first score display areas 3042 and 3052.

[0057] Here, the amount of information displayed in each of second display areas 303, 304, and 305 may be determined, for example, by the value of the first score. For example, when the value of the first score is smaller than a predetermined threshold Th1, both the first score and a diagram representing the application result of the first change model are displayed. When the value of the first score is larger than the predetermined threshold Th1 and smaller than a threshold Th2 that is larger than the threshold Th1, either the first score or the diagram representing the application result of the first change model, for example, the value of the first score, is displayed. When the value of the first score is larger than the predetermined threshold Th2, for example, a message indicating that the score value is larger than the threshold may be displayed, as shown in first score display area 3052 in FIG. 8. In this case, the corresponding second display area itself may be hidden.

[0058] The display priority of second display areas 303, 304, and 305 may also be determined according to the value of the first score. For example, the priority may be determined so that second display areas are arranged from the top of the display screen in ascending order of first score.

[0059] Furthermore, if the value of the first score is smaller than a threshold value Th1, the corresponding second display area may be emphasized. The emphasis may be achieved, for example, by coloring the corresponding second display area, by bolding the name of the ordinal data displayed in the corresponding second display area, or by attaching a warning mark or the like to the corresponding second display area. Conversely, if the value of the first score is larger than a threshold value Th2, the corresponding second display area may be made less noticeable. The less noticeable the second display area may be made, for example, by lightening the color of the second display area.

[0060] As described above, according to the first embodiment, the status data Y i and sequence data regarding the manufacturing sequence of the product. k The first score calculated using i Ordinal data O related to changes in k That is, in the first embodiment, the state data Y i Without using information that directly indicates the cause of the change in the state data Y i Therefore, since the user does not need to monitor the status data one by one, it is expected that the burden on the user will be reduced.

[0061] In the first embodiment, the priority and amount of information when the analysis result is displayed are determined based on the value of the first score. i It is possible to prioritize checking the manufacturing sequence that is expected to be highly related to changes in the manufacturing process, for example, specific passage points during manufacturing. In addition, since the number of points that need to be checked is reduced, it is expected that the burden on the user to check will be reduced. Furthermore, it is expected that the number of times the user overlooks something will be reduced.

[0062] In addition, the first score in the first embodiment is the state data Y i Ordered data O k Status data Y based on whether or not there is a change and the number of changes iFor example, if the product goes into a specific state such as an abnormal state due to an event such as changing the conditions before and after maintenance, changing the lot of parts, or a malfunctioning device that was operating normally, the status data Y i often changes when certain events occur in a specific order. In other words, in such cases, the state data Y i can change significantly at a certain point in time. i If the cause of the change is the state data Y in the process where the specific event occurred, i On the other hand, the number of changes in the state data Y i If measurement errors are included in the status data Y or if the order of products is changed between processes, i can change in a complex manner. i The number of changes in the first score is considered to be large. i Ordered data O k Status data Y based on whether or not there is a change and the number of changes i The score indicates the likelihood of a change in the state data Y being in a specific state such as an abnormal state. i The first variation model that matches the sequence data O can be determined. k In this case, the state data Y i It can be assumed that there was some cause that caused the product to be in a specific state, such as an abnormal state, in the order or time when the changes in the status data Y i Ordinal data O related to changes in k can be estimated without using information on the variations in manufacturing conditions.

[0063] (Second embodiment) Next, a second embodiment will be described. As in the first embodiment, the data processing device according to the second embodiment also uses order information to estimate the cause of a product entering a certain state, and visualizes the cause based on the estimation result. In this way, the data processing device supports the user in discovering the cause of a product entering a certain state. In the first embodiment, the first calculation unit 102 calculates the order data O k Based on the state data Y i In this case, a first score relating to the change in the status data is calculated. On the other hand, if manufacturing condition data, which is information indicating the manufacturing conditions of the product, can be acquired from the manufacturing database 104, a second score relating to the change in the status data can be calculated based on the manufacturing condition data in the same manner as the first score. If the first score and the second score are available, the relationship between the order that caused the change in the status data and the manufacturing conditions can be presented to the user by comparing the first score and the second score. In this way, the second embodiment is an example that is applied when manufacturing condition data can be acquired. Hereinafter, the description of the same parts as in the first embodiment will be omitted or simplified, and only the parts that are different from the first embodiment will be described.

[0064] 8 is a block diagram showing a data processing device according to the second embodiment. The data processing device 1 includes an acquisition unit 101, a first calculation unit 102, an output unit 103, and a second calculation unit 106. As in the first embodiment, the data processing device 1 is connected to a manufacturing database 104 and a display unit 105.

[0065] In the second embodiment, the acquisition unit 101 acquires manufacturing data. In the second embodiment, the manufacturing data is data used to calculate the first score and the second score, and includes key data for identifying each product, status data related to the status of each product, sequence data related to the manufacturing sequence of each product, and manufacturing condition data related to the manufacturing conditions of each product. The acquisition unit 101 inputs the key data, status data, and sequence data of the manufacturing data to the first calculation unit 102. The acquisition unit 101 also inputs the key data, status data, and manufacturing condition data of the manufacturing data to the second calculation unit 106.

[0066] The second calculation unit 106 calculates a second score related to the change in the condition data based on the second change model using the condition data and the manufacturing condition data. The second score is a score that indicates how well the second change model fits the actual condition data and the manufacturing condition data. The second change model is a model that indicates the change in the condition data using the manufacturing condition data.

[0067] The output unit 103 generates visualization data from the first score or information explaining the first score, and the second score or information explaining the second score. Then, the output unit 103 outputs the visualization data to the display unit 105. As in the first embodiment, the visualization data is data for visually presenting to the user potential causes of a large change in the manufacturing data.

[0068] Next, the operation of the data processing device 1 of the second embodiment will be described. First, the operation of the acquisition unit 101 will be described. In the second embodiment, the acquisition unit 101 acquires manufacturing data including manufacturing condition data from the manufacturing database 104. The manufacturing condition data will be described in detail below. For the purpose of the following description, the manufacturing condition data is assumed to be {C j :j=1...M}, where M represents the number of items of acquired manufacturing condition data.

[0069] In addition, information such as the names of materials used in the product and the names of equipment used to process or assemble the product may be used as manufacturing conditions recorded as manufacturing condition data. More generally, information related to 5M1E may be used as manufacturing conditions. 5M1E is an acronym for Man, Machine, Material, Method, Measurement, and Environment, and is widely known as six factors for managing manufacturing processes. Man information includes information such as the name of the person who processes the product. Machine information includes information such as the name of the equipment used to manufacture the product, the name of the production line, and the state of the equipment during processing, such as temperature and pressure. Material information includes information such as the ID or name of the material used to manufacture the product and the ID or name of the parts that make up the product. Method information includes information such as the product processing method and the type of processing program. Measurement information includes information such as the name of the equipment where the measurement was performed and the measurement location on the product where the measurement was performed. Environment information includes information such as the name of the factory building where the measurement was performed and the temperature and humidity at the time of the measurement. The acquisition unit 101 acquires manufacturing condition data from the manufacturing database 104 so that it includes the manufacturing conditions necessary for analysis. At this time, the acquiring unit 101 may acquire the manufacturing condition data so as to include manufacturing conditions that the user has determined to be useful for analysis and visualization.

[0070] Next, the operation of the first calculation unit 102 will be described. i and ordered data O k When receiving the first score S1(Y i ,O k The first score is calculated in the same manner as in the first embodiment.

[0071] Next, the operation of the second calculation unit 106 will be described. i and manufacturing condition data C j When receiving the second score S2(Y i ,C j) is calculated. FIG. 9 is a flowchart showing the process of calculating the second score in the second calculation unit 106. Here, the second score may be calculated basically in the same manner as the first score. Therefore, the explanation of FIG. 9 will be simplified.

[0072] First, the second calculation unit 106 initializes i and j to initial values, for example, 1. Then, in step S101, the second calculation unit 106 calculates the state data Y i and manufacturing condition data C j The second change model f(Y i )=aC j The second change model is applied to the manufacturing condition data C j Using the state data Y i The second variation model may be a regression model similar to the first variation model. That is, the second variation model represents, for example, a variation of the manufacturing condition data C j may be a regression model with step basis.

[0073] Here, the manufacturing condition data C j If the data is nominal scale data such as material or part ID, material or part name, manufacturing condition data C j After being converted into dummy variables using a method such as One Hot Encoding, the second change model is applied. j The method of converting the data into dummy variables is not limited to One Hot Encoding. For example, j To convert the manufacturing condition data C into a dummy variable, methods such as Label Encoding, Count Encoding, Target Encoding, and Leave one out Encoding may be used. j If the data can be used to apply a regression model without converting it into a dummy variable such as a ratio scale, these conversion methods do not need to be used.

[0074] In step S102, the second calculation unit 106 calculates the state data Y i The second score S2(Yi ,C j ) is calculated. i ,C j ) is the state data Y i and manufacturing condition data C j The goodness of fit of the second change model to the state data Y i and manufacturing condition data C j The second score is a value that represents the degree of conformance of the second change model to the condition data Y. Here, the second score is preferably a score that can be compared with the first score. For example, the BIC, which is the same as the first score, may be used as the second score. Of course, each score that can be used as the first score may be used as the second score. The second score indicates how well the second change model conforms to the condition data Y. i For example, by displaying the ranking of the first score and the ranking of the second score in parallel, the user can easily understand the change in the order data O. k and manufacturing condition data C j Which item is in status data Y? i It is possible to evaluate whether a change in

[0075] In step S103, the second calculation unit 106 determines whether j is greater than M. If it is determined in step S103 that j is not greater than M, the second calculation unit 106 adds 1 to j and returns the process to step S101. In step S103, if j is greater than M, that is, if the current state data Y i All manufacturing condition data in C j If it is determined that the processing for has been completed, the second calculation unit 106 moves the processing to step S104.

[0076] In step S104, the second calculation unit 106 determines whether i is greater than N. If it is determined in step S104 that i is not greater than N, the second calculation unit 106 adds 1 to i and returns the process to step S101. If it is determined in step S104 that i is greater than N, that is, if all the state data Y iAll manufacturing condition data in C j 9 is completed, the second calculation unit 106 ends the processing of FIG. 9. In this case, the second calculation unit 106 outputs the second score or information explaining the second score obtained in the processing of FIG. 9 to the output unit 103. Here, the information explaining the second score may be, for example, the state data Y i The result of applying the second score to the state data Y i The result of applying it to, for example, the state data Y i , manufacturing condition data C j and status data Y i and manufacturing condition data C j The data may be a function representing a second variation model that best fits the

[0077] Next, the operation of the output unit 103 will be described. After the first calculation unit 102 calculates the first score and the second calculation unit 106 calculates the second score, the output unit 103 generates visualization data from the first score or information explaining the first score and the second score or information explaining the second score. The output unit 103 may further perform processing to compare the first score and the second score. Status data Y i The cause of the change occurs at a certain timing, and when the order of the products is changed, the first score is considered to be the same as the score of the true cause. Therefore, by comparing the first score and the second score, the corresponding manufacturing condition data C j It can be determined how likely it is that the first score and the second score are the cause of the change. For example, if the first score and the second score are about the same, the corresponding manufacturing condition data C j The change in state data Y i On the other hand, if there is a difference between the first score and the second score, the corresponding manufacturing condition data C j The change in state data Y iIn this case, it is suggested that there is another cause. Therefore, in such a case, the output unit 103 may generate visualization data to notify the user that there may be another cause of the change.

[0078] The first score to be compared with the second score will be explained. First, the degree of freedom in calculating the second score is determined by the manufacturing condition data C j The manufacturing condition data C is determined by the condition number minus 1. j The number of conditions is the manufacturing condition data C j For example, if one of two processing conditions can be selected as the processing condition for a certain process, the manufacturing condition data C j The condition number of is 2, and the degree of freedom when calculating the second score is 1. On the other hand, the degree of freedom when calculating the first score can take any range from 0 to L-1. Therefore, when comparing the first score with the second score, the degree of freedom of the first score may be the degree of freedom when the maximum likelihood score is obtained, for example, when BIC is used, the degree of freedom when BIC is smallest. Alternatively, the ordered data O k For example, the order data O k The degree of freedom of manufacturing condition data C j By aligning the number of elements to the order data O k Comparison can be made while suppressing overfitting. When specifying degrees of freedom for calculation as described above, greedy methods, orthogonal matching pursuit (OMP), Lasso regression, or their derivative methods may be used. These methods make it possible to perform regression calculations and obtain scores by specifying not only the maximum likelihood degrees of freedom but also any degrees of freedom. Therefore, these methods are preferable.

[0079] Furthermore, when the first score and the second score are compared, a Δ score, which is the difference between the two scores, may be used as a reference value for comparison. For example, if the Δ score is 0, it indicates that the first score and the second score are exactly the same value. Also, the larger the Δ score, the greater the gap between the first score and the second score. In the present embodiment, when the first score and the second score are compared on the assumption that the first score is equivalent to the true cause, the smaller the Δ score, the closer the manufacturing condition data C corresponding to the second score at that time is. j is highly likely to be the true cause, and the larger the Δ score, the more likely the manufacturing condition data C j There is a high possibility that there is a cause other than the above. In other words, by using the Δ score, j The confidence level for the true cause of the error can be presented. When the first and second scores are BIC, the Δ score is calculated by the Bayes factor BF shown in Jeffreys, H. 1961. Theory of probability. Oxford University Press. 01 and have the relationship shown in formula (3). Δ score ~2lnBF 01 (3)

[0080] Based on Equation (3), the logarithm of the reference value may be doubled and used as the reference value for comparing the delta scores. Of course, the user may set an appropriate reference value depending on the situation. This allows the comparison results to be evaluated according to the user's needs.

[0081] FIG. 10A shows the status data Y i Manufacturing condition data C that caused the change j 10A is a diagram showing an example of data generated by the output unit 103 when the status data Y i 10A shows an example in which the first score and the second score for the order data O k and manufacturing condition data C jare sorted in ascending order of scores. The data may be arranged in any order that is easy for the user to see, such as the order of processing, rather than in the order of scores. According to FIG. 10A, the order data O5 shows the smallest first score when the degree of freedom is 1. The smallest first score value for the order data O5 matches the value of the second score for the manufacturing condition data C8. This means that the manufacturing condition data C8 is i In this way, the user can easily determine the state data Y i The causes of the changes can be considered.

[0082] FIG. 10B shows the status data Y i Manufacturing condition data C that caused the change j 10B is a diagram showing an example of data generated by the output unit 103 when the manufacturing condition data C8 cannot be acquired. In FIG. 10B, since the manufacturing condition data C8 cannot be acquired, the delta score between the smallest first score value for the order data O5 and the second score value for the manufacturing condition data C4 and the delta score between the smallest first score value for the order data O5 and the second score for the manufacturing condition data C9 are both large values. This means that the manufacturing conditions other than the manufacturing condition data C4 and the manufacturing condition data C9 are not included in the status data Y i In this case, the comparison of the first score suggests that the state data Y i In the example of FIG. 10B, the first score is smallest when the degree of freedom of the order data O5 is 1. Therefore, the order data O5 changes the status data Y i It can be assumed that a change of

[0083] Furthermore, manufacturing condition data C j The diagrams related to this example correspond to FIG. 6, for example, and the vertical axis represents the status data Y i The horizontal axis represents the value of manufacturing condition data C j Alternatively, the manufacturing condition data C j For example, the vertical axis represents the status data Y i The horizontal axis represents the value of manufacturing condition data C jManufacturing condition data C such as scatter plots and histograms j and state data Y i On the other hand, as in this embodiment, the order data O k If the information on the order data O is also available, the order data O shown in Figure 11 can be used. k , manufacturing condition data C j , state data Y i In Fig. 11, the vertical axis represents the state data Y i The horizontal axis is the value of the ordinal data O k In FIG. 11, a line 202 of a first change model applied based on the order data O5 and a line 204 of a second change model applied based on the manufacturing condition data C8 that is highly related to the order data O5 are displayed on a scatter diagram 201 at each time of the order data O5, and the display mode of a part 203 of the scatter diagram 201 is adjusted according to the content of the manufacturing condition data C8. The adjustment of the display mode is, for example, an adjustment of color. In FIG. 11, the scatter diagram before the change in the manufacturing condition data C8 is plotted with black dots, and the scatter diagram after the change is plotted with white dots. With the display as shown in FIG. 11, the state data Y i 11, the display mode is adjusted by adjusting the order data O5 and the manufacturing condition data C8. j The plots are colored differently depending on the change in the manufacturing condition data C. On the other hand, the adjustment of the display mode is not limited to changing the color of the plots. For example, the adjustment of the display mode can be performed by changing the shape of the plot points, etc. j This may be done in any manner that allows the change in the

[0084] The operation of the output unit 103 will be further described. Fig. 12 is an example of a display screen of the analysis results displayed on the display unit 105 when the first score, the application result of the first change model, and the second score are received. The display screen of Fig. 12 is displayed based on visualized data. As shown in Fig. 12, the display screen of the analysis results includes a first display area 300. The first display area 300 includes a creation date display field 301, a condition data name display field 302, and second display areas 303, 304, 305, 306, and 307. Here, Fig. 12 shows the display screen of the analysis results for the condition data Y to be analyzed. i This is an example of a display screen when there is one status data Y i If there are two or more, all the state data Y i The first display areas 300 for the respective status data Y i If any score has been calculated for each state data Y, the state data Y is ranked based on the score. i A first display area 300 for may be displayed.

[0085] The creation date display field 301 and the status data name display field 302 in the first display area 300 are the same as those shown in Fig. 7. Therefore, a description of these will be omitted.

[0086] The second display areas 303, 304, and 305 correspond to the status data Y displayed in the corresponding first display area 300. i and ordered data O k On the other hand, the second display areas 306 and 307 are display areas for displaying information about the combination of the status data Y displayed in the corresponding first display area 300. i and manufacturing condition data C j A maximum of L×M second display areas can be provided for one first display area.

[0087] Second display areas 303, 304, and 305 include ordinal data name display fields 3031, 3041, and 3051. Second display areas 303, 304, and 305 also include at least one of a display area for the first score and a display area for a diagram representing the application result of the first change model. For example, second display area 303 includes both first score display area 3032 and a display area 3033 for a diagram representing the application result of the first change model. On the other hand, second display areas 304 and 305 do not include a display area for a diagram representing the application result of the first change model, and instead include first score display areas 3042 and 3052.

[0088] The second display areas 306 and 307 include manufacturing condition data name display fields 3061 and 3071. The second display areas 306 and 307 also include at least one of a display area for the second score and a display area for a scatter diagram showing the relationship between the manufacturing condition data and the status data. For example, the second display area 306 includes both a second score display area 3062 and a display area 3063 for a scatter diagram showing the relationship between the manufacturing condition data and the status data. On the other hand, the second display area 307 does not include a display area for a scatter diagram showing the relationship between the manufacturing condition data and the status data, but includes a second score display area 3072. The scatter diagram showing the relationship between the manufacturing condition data and the status data may be, for example, a display area for the respective manufacturing condition data C j The state data Y corresponding to i Here, C in Figure 11 81 is the manufacturing condition data C8 before the change, and C 82 is the changed manufacturing condition data C8.

[0089] The amount of information displayed in each second display area may be determined by, for example, the value of the first score and the value of the second score. For example, when the value of the first score is smaller than a predetermined threshold Th1, both the first score and a diagram representing the application result of the first change model are displayed. When the value of the first score is larger than a predetermined threshold Th1 and smaller than a threshold Th2 larger than the threshold Th1, either the first score or the diagram representing the application result of the first change model, for example, the value of the first score, is displayed. Similarly, when the value of the second score is smaller than a predetermined threshold Th3, both the second score and the diagram representing the application result of the second change model are displayed. When the value of the second score is larger than a predetermined threshold Th3 and smaller than a threshold Th4 larger than the threshold Th3, either the second score or the diagram representing the application result of the second change model, for example, the value of the second score, is displayed. When the value of the second score is larger than a predetermined threshold Th4, for example, a message indicating that the score value is larger than the threshold may be displayed, similar to the first score. In this case, the corresponding second display area itself may be hidden. The threshold value Th1 and the threshold value Th3 may be the same value. Similarly, the threshold value Th2 and the threshold value Th4 may be the same value.

[0090] In addition, manufacturing condition data C, which shows a second score similar to the first score, j If there is manufacturing condition data C j The second display area may be highlighted by, for example, changing the color of the frame, making the name of the manufacturing condition bold, or displaying a specific mark.

[0091] The display priority of the second display areas 303, 304, 305, 306, and 307 may also be determined according to the values ​​of the first and second scores. In this case, if the first and second scores can be compared, the first and second scores may be compared together, and the priority may be determined such that the second display areas are arranged from the top of the display screen in ascending order of score value. On the other hand, if the first and second scores cannot be compared, the first and second scores may be compared separately, and the priority may be determined such that the second display areas based on the ordering data are arranged in ascending order of first score value, and the second display areas based on the manufacturing condition data are arranged in descending order of second score value.

[0092] 11. In addition, instead of the scatter diagram showing the relationship between the manufacturing condition data and the state data shown in FIG. 12, the order data O shown in FIG. k , manufacturing condition data C j , state data Y i A combined scatter plot may be displayed.

[0093] Figure 13 shows the order data O k , manufacturing condition data C j , state data Y i 13 is an example of a display screen of the analysis result when a scatter diagram is displayed in which the ordered data O with the smallest first score is displayed. k In this example, second display areas 3034, 3035, and 3036 for the related manufacturing condition data C8, C4, and C9 are displayed in the second display area 303 for the order data O5. The second display areas 3034, 3035, and 3036 include manufacturing condition data name display fields 3034a, 3035a, and 3036a. The second display areas 3034, 3035, and 3036 also include a display area for the second score and a display area for the order data O5. k , manufacturing condition data C j , state data Y iand at least one of a display area for a scatter plot combined with the second score. For example, second display area 3034 includes display area 3034b for the second score and display area 3034c for the scatter plot. On the other hand, second display areas 3035 and 3036 include only display areas 3035b and 3036b for the second score. Here, the amount of information displayed in each second display area 3034, 3035, 3036 may be determined, for example, by the value of the second score.

[0094] Such ordered data O k , manufacturing condition data C j , state data Y i The combined scatter plot allows for ordinal data k Manufacturing condition data C based on j The state data Y changes with i Therefore, the user can visually distinguish the change in the order data O k Manufacturing condition data C based on j It is easy to evaluate.

[0095] Here, the degree of freedom when the first score is calculated and the manufacturing condition data C when the second score is calculated are j If the number of elements in the first display area 303 matches the number of elements in the second display area 305, a display screen including second display areas 303, 304, and 305 as shown in Fig. 14 can also be displayed. The following description will be given with reference to Fig. 14.

[0096] The second display areas 304, 305, and 306 include manufacturing condition data name display fields 3037, 3043, and 3053. The second display areas 304, 305, and 306 also include a display area for the Δ score and a display area for the order data O. k , manufacturing condition data C j , state data Y i For example, the second display area 304 includes a display area 3038 for the Δ score and a display area for the order data O. k , manufacturing condition data C j , state data Y iOn the other hand, the second display areas 304 and 305 include only Δ score display areas 3044 and 3054. Here, the second display areas may include order data name display fields instead of the manufacturing condition data name display fields 3037, 3043, and 3053.

[0097] Here, the amount of information displayed in each of the second display areas 303, 304, and 305 may be determined, for example, by the value of the Δ score. For example, if the value of the Δ score is within a threshold range R1, which is a range close to 0, the amount of information displayed in each of the second display areas 303, 304, and 305 may be determined by the Δ score and the order data O. k , manufacturing condition data C j , state data Y i In addition, if the value of the Δ score is wider than the threshold range R1 and within a threshold range R2 wider than the threshold range R1, the Δ score and the order data O are displayed. k , manufacturing condition data C j , state data Y i Either one of the scatter plots combining the above two scatter plots, for example, the value of the Δ score, is displayed. Similarly, if the value of the Δ score is outside the threshold range R2, the fact that the value of the Δ score is outside the threshold range may be displayed. In this case, the corresponding second display area itself may be hidden.

[0098] The display priority of the second display areas 303, 304, and 305 may also be determined according to the value of the Δ score. For example, the priority may be determined so that the second display areas are arranged from the top of the display screen in ascending order of the Δ score.

[0099] Furthermore, if the Δ score value is within the threshold range R1, the corresponding second display area may be emphasized. The emphasis may be achieved, for example, by coloring the corresponding second display area, bolding the name of the manufacturing condition data or sequence data displayed in the corresponding second display area, or adding a warning mark or the like to the corresponding second display area. Conversely, if the Δ score value is outside the threshold range R2, the corresponding second display area may be made less noticeable. Making the second display area less noticeable may be achieved, for example, by lightening the color of the second display area.

[0100] As described above, according to the second embodiment, in addition to the first embodiment, the status data Y i and manufacturing condition data C regarding the manufacturing conditions of the product j The second score calculated using i Manufacturing condition data C related to changes in j The user can estimate the corresponding manufacturing condition data C by comparing the first score with the second score. j The change in state data Y i It is possible to assess whether the change is a plausible cause of the change.

[0101] In the second embodiment, the priority and the amount of information when the analysis result is displayed are determined based on both the first score and the second score. i It is possible to check, with priority given to, the manufacturing sequence and manufacturing conditions that are expected to be highly related to the change in the amount of the product.

[0102] (Third embodiment) Next, a third embodiment will be described. As in the first and second embodiments, the data processing device according to the third embodiment also uses order information to infer the cause of a product entering a certain state, and visualizes the cause based on the inferred result. In this way, the data processing device supports the user in discovering the cause of the product entering a certain state.

[0103] FIG. 15 is a block diagram illustrating a data processing device according to a third embodiment. The data processing device 1 includes an acquisition unit 101, a first calculation unit 102, a second calculation unit 106, and an output unit 103. In the third embodiment, the output unit 103 is connected to an operation unit 107. The operation unit 107 may be provided separately from the data processing device 1, or may be provided in the data processing device 1. As described above, the third embodiment differs from the first and second embodiments in that the content of the visualization data generated by the output unit 103 is changed in response to an operation from the operation unit 107. More specifically, the amount of information displayed and the display priority are changed in response to the operation. Below, descriptions of the same parts as in the second embodiment will be omitted or simplified, and only the parts different from the second embodiment will be described. Note that the operation unit 107 may be included in the data processing device 1 of the first embodiment, rather than in the data processing device 1 of the second embodiment. The operation in this case is also similar to that of the third embodiment described below.

[0104] 16 is a diagram showing an example of a display screen displayed on display unit 105 in the third embodiment. The display screen shown in FIG. 16 includes a first display area 300 similar to the display screen shown in FIG. 12. In FIG. 16, first display area 300 includes second display areas 303, 304, 306, and 307. First display area 300 may further include second display area 305 shown in FIG. 12. Furthermore, first display area 300 may be the first display area 300 shown in FIG. 7, the first display area 300 shown in FIG. 13, or the first display area 300 shown in FIG. 14.

[0105] 16, each of second display areas 303, 304, 306, and 307 further includes an amount of information change button 401, 402, 403, and 404. When any of the amount of information change buttons 401, 402, 403, and 404 is selected using operation unit 107, the amount of information in the corresponding second display area increases or decreases.

[0106] For example, the second display area 303 in FIG. 16 is in a state where the amount of information is increasing. That is, the second display area 303 displays all of the name of the ordered data, the first score, and a diagram showing the application result of the first change model. When the information amount change button 401 is selected in this state, the amount of information in the second display area 303 decreases. Specifically, only the name of the ordered data and the first score are displayed in the second display area 303. When the amount of information decreases, the second display area is displayed, for example, collapsed on the display unit 105. When the amount of information is changed from an increased state to a decreased state, the display may be a reduced display instead of a collapsed display.

[0107] On the other hand, the second display area 304 in FIG. 16 is in a state where the amount of information is reduced. That is, only the name of the ordered data and the first score are displayed in the second display area 304. When the change amount of information button 404 is selected in this state, the amount of information in the second display area 304 increases. Specifically, in addition to the name of the ordered data and the first score, a diagram showing the application result of the first change model is displayed in the second display area 304. When the amount of information increases, the second display area is displayed, for example, as expanded on the display unit 105. When the amount of information is changed from a state where the amount of information is reduced to a state where the amount of information is increased, the display may be an enlarged display instead of an expanded display.

[0108] Here, it is desirable that information amount change buttons 401, 402, 403, and 404 include information relating to a change in the amount of information in the second display area. The information relating to a change in the amount of information in the second display area is displayed as, for example, "+" or "-". Specifically, when the amount of information in the second display area is increasing, "+" may be displayed on information amount change buttons 401, 402, 403, and 404, and when the amount of information in the corresponding second display area is decreasing, "-" may be displayed on information amount change buttons 401, 402, 403, and 404.

[0109] In the first and second embodiments, the second display areas whose first scores or second scores are greater than the threshold are hidden. In the third embodiment, the first display area 300 includes a switch button 405 for switching between displaying and hiding the second display areas whose first scores or second scores are greater than the threshold. For example, while the switch button 405 displays "show more," the second display areas whose first scores or second scores are greater than the threshold are collapsed and hidden. When the switch button 405 is selected in this state, the collapsed second display areas are displayed. Then, the switch button 405 displays "hide." When the switch button 405 is selected in this state, the second display areas whose first scores or second scores are greater than the threshold are collapsed and hidden again. Here, all the collapsed second display areas may be displayed, or only a predetermined number of them may be displayed. The display order of the second display areas may be determined by the first score and / or the second score. When a predetermined number of second display areas are displayed, only the display areas with the highest first scores and / or second scores may be displayed.

[0110] 17 is a schematic diagram showing the relationship between the amount of information change button and the switching button and the amount of information in the second display area according to the third embodiment. As shown in FIG. 17, the amount of information in the second display area, which displays the first score or the second score and a diagram showing the application result of the model, is the largest. The amount of information in the second display area, which is not displayed, is the smallest. The amount of information in the second display area can be increased or decreased using the amount of information change button and the switching button.

[0111] As described above, the first display area displays the state data Y i There can be as many state data as there are to be analyzed. i When there are a plurality of first display areas, an information amount change button and a switching button may be provided for each first display area.

[0112] 18 is a diagram showing an example of a display screen having an interface for accepting an operation for changing the priority of the first display area and an operation for changing the priority of the second display area. The display screen shown in FIG. 18 includes a first display area 300 similar to the display screen shown in FIG. 16. In FIG. 18, the first display area 300 includes second display areas 303, 304, 306, and 307. The first display area 300 may further include the second display area 305 shown in FIG. 12. Furthermore, the first display area 300 may be the first display area 300 shown in FIG. 7, the first display area 300 shown in FIG. 13, or the first display area 300 shown in FIG. 14.

[0113] In FIG. 18 , the first display area 300 further includes priority change buttons 501 and 502. The priority change button 501 includes a radio button for accepting an operation to change the display priority of the multiple first display areas in order of score or in an order specified by the user, and a radio button for accepting an operation to change the priority in ascending or descending order of the score or the specified order. The priority change button 502 includes a radio button for accepting an operation to change the display priority of the multiple second display areas in order of score or in an order specified by the user, and a radio button for accepting an operation to change the priority in ascending or descending order of the score or the specified order. Here, ascending order of score refers to, for example, ascending order of the first score and the second score. Similarly, descending order of score refers to, for example, descending order of the first score and the second score. The user's specified order can be arbitrarily specified by the user, for example, in order of the first score, the second score, or the like.

[0114] In FIG. 18, the first display area or the second display area is rearranged according to the priority designated by the priority change button 501 or the priority change button 502.

[0115] 18 shows radio buttons as an interface for changing the priority level. The interface for changing the priority level may be a check box, a pull-down menu, or the like. When a status data name display field 302 is selected, priority level change buttons 501 and 502 may be displayed in the corresponding first display area 300.

[0116] 19, when a cursor 601 is placed on a scatter diagram in a display area 3033 of a diagram showing the application result of the first change model, a detailed analysis result 602 at the position of the cursor 601 may be displayed. The detailed analysis result may include, for example, an ID, corresponding state data Y i The measurements of and the corresponding ordinal data O k The value of the order in the first change model may be included. Note that if there are multiple display areas for a diagram representing the application result of the first change model or if multiple analysis results have been generated, a button for displaying details of the analysis results may be provided separately. In this case, when the button is selected, details of the multiple analysis results may be displayed.

[0117] Additionally, the first display area 300 may include an interface for changing the thresholds for the first score and the second score.

[0118] As described above, in the third embodiment, the status data Y i , ordered data O k , manufacturing condition data C j Regarding the state data Y iThe amount and priority of information to be displayed in the second display area are determined based on the first score or the second score regarding the change in the information. The user can also change the amount of information and the display priority of the second display area through the operation unit 107. The user can also change the display priority of the first display area through the operation unit 107. This allows the user to not only prioritize monitoring information that is expected to be highly related to an abnormality, but also to check the analysis results from more diverse perspectives. For example, in daily monitoring work, the user can check items that are expected to be highly related to an abnormality, which are displayed first, and in situations where detailed monitoring is required, can also monitor items that are relatively less related to an abnormality.

[0119] (Fourth embodiment) Next, a fourth embodiment will be described. As in the first to third embodiments, the data processing device according to the fourth embodiment also uses order information to infer the cause of a product entering a certain state, and visualizes the cause based on the inference result. In this way, the data processing device supports the user in discovering the cause of the product entering a certain state.

[0120] 20 is a block diagram showing a data processing device according to a fourth embodiment. The data processing device 1 includes an acquisition unit 101, a first calculation unit 102, a second calculation unit 106, an output unit 103, a manufacturing process database 108, and a user database 109. In the fourth embodiment, the output unit 103 generates visualization data according to manufacturing process data recorded in the manufacturing process database 108 and user data recorded in the user database 109.

[0121] The manufacturing process database 108 records manufacturing process data that indicates the manufacturing process of a product. Usually, a manufacturing process is a subdivision of the work that is carried out until the final product is completed. Then, manufacturing data is usually recorded for each process. In this embodiment, the manufacturing process data is data that indicates to which manufacturing process each item of the manufacturing data acquired by the acquisition unit 101 belongs. FIG. 21 is a diagram showing the relationship between manufacturing data and manufacturing process data. For example, in FIG. 21, order data O1 and manufacturing condition data C1 belong to process A of the manufacturing process data. Such status data Y i , ordered data O k , manufacturing condition data C j The manufacturing process data is data indicating to which process each of the above belongs. In the fourth embodiment, the output unit 103 generates visualization data using not only the first score or the second score but also the manufacturing process data.

[0122] 22 is a diagram showing an example of a display screen displayed on display unit 105 in the fourth embodiment. The display screen shown in FIG. 22 includes a first display area 300 similar to the display screen shown in FIG. 16. In FIG. 22, first display area 300 includes second display areas 303, 304, 306, and 307. First display area 300 may further include second display area 305 shown in FIG. 12. Furthermore, first display area 300 may be the first display area 300 shown in FIG. 7, the first display area 300 shown in FIG. 13, or the first display area 300 shown in FIG. 14.

[0123] 22, the first display area 300 includes a manufacturing process name display field 701 and a priority change button 702. The manufacturing process name display field 701 displays the current analysis target status data Y i The priority change button 702 includes a radio button for accepting an operation to change the display priority of the plurality of first display areas in order of score, order of process, or order specified by the user, and a radio button for accepting an operation to change the priority in ascending or descending order of score, order of process, or specified order.

[0124] The first display area is rearranged according to the priority specified by the priority change button 702. For example, in Fig. 22, when ascending order of process sequence is specified, the first display area is rearranged in descending order of the manufacturing process to which the status data Y7 to be measured belongs.

[0125] 22, the second display area also includes a manufacturing process name display field 703. The manufacturing process name display field 703 displays the order data O of the corresponding second display area. k Or manufacturing condition data C j This is a display field for displaying the name of the manufacturing process to which the sequence data O5 belongs. For example, FIG. 22 shows that the sequence data O5 with the smallest first score belongs to process D. In other words, FIG. 22 shows that status data Y7 in process K is the data that indicates an abnormality, and is therefore the subject of analysis, but it suggests that the cause of the change in status data Y7 was process D.

[0126] Here, the first display area 300 may have an interface for changing the threshold values ​​for the first score and the second score for each process. Such an interface is effective when it is known which processes are more likely to cause abnormalities and which are less likely to cause abnormalities.

[0127] The user database 109 records user data of the user who performs the analysis. In the embodiment, users may be, for example, workers in each manufacturing process, managers of each manufacturing process, or managers responsible for overall manufacturing. The user data is data that indicates the priority of display for each of these user attributes. FIG. 23 is a diagram showing the relationship between manufacturing data and user data. The user data may be data indicating, for example, whether or not to display each measurement value. In FIG. 23, analysis results for manufacturing data marked with a circle are displayed. On the other hand, analysis results for manufacturing data not marked with a circle are not displayed. The user data may be, for example, data indicating whether or not to display each manufacturing process. In this case, for example, only analysis results for the process that the user is responsible for are displayed.

[0128] Here, the user data may include numerical data representing the level of the user. For example, a user with a low numerical value may be shown analysis results for a portion of the manufacturing data, while a user with a high numerical value may be shown analysis results for a large amount of the manufacturing data. For example, the numerical value for the level of an inexperienced user may be set low, while the numerical value for the level of an experienced or managerial user may be set high. In FIG. 23, the numerical values ​​are set in three levels, but the numerical values ​​do not necessarily have to be set in three levels.

[0129] As described above, according to the fourth embodiment, visualization data is generated taking into account not only the scores but also the manufacturing process data. This allows for a display that is easy for the user to monitor, such as displaying the analysis results in the order of the manufacturing processes. Furthermore, by using not only the scores but also user data, a display that is tailored to the monitoring target and / or proficiency level of the user can be realized. This reduces the burden on the user.

[0130] In the fourth embodiment, the manufacturing process database 108 and the user database 109 are included in the data processing device 1. However, the manufacturing process database 108 and the user database 109 may be provided separately from the data processing device 1.

[0131] (Fifth embodiment) Next, a fifth embodiment will be described. As in the first to fourth embodiments, the data processing device according to the fifth embodiment also uses order information to infer the cause of a product entering a certain state, and visualizes the cause based on the inference result. In this way, the data processing device supports the user in discovering the cause of the product entering a certain state.

[0132] 24 is a block diagram showing a data processing device according to a fifth embodiment. The data processing device 1 includes an acquisition unit 101, a first calculation unit 102, a second calculation unit 106, an output unit 103, and an analysis database 110. The analysis database 110 may be included in the data processing device 1 of the first embodiment, rather than in the data processing device 1 of the second embodiment. The operation in this case is also similar to that of the fifth embodiment described below.

[0133] In the fifth embodiment, the first calculation unit 102 records the first score or information explaining the first score in the analysis database 110. Similarly, the second calculation unit 106 records the second score or information explaining the second score in the analysis database 110.

[0134] Furthermore, in the fifth embodiment, the acquisition unit 101 records data related to the acquired manufacturing data in the analysis database 110. For example, the acquisition unit 101 records data on acquisition conditions for manufacturing conditions such as the acquisition period of the manufacturing data.

[0135] Furthermore, in the fifth embodiment, the output unit 103 generates visualization data using data recorded in the analysis database 110. For example, the output unit 103 may generate graphs showing the progress of the first score and the second score in addition to the diagram showing the application results of the first change model and the diagram showing the application results of the second change model described above. Furthermore, the output unit 103 may generate visualization data that allows a comparison between the previous analysis result and the current analysis result.

[0136] As described above, in the fifth embodiment, past first scores or information explaining the first scores, and past second scores or information explaining the second scores are recorded in the analysis database 110. This allows visualization data to be provided without re-applying a change model, etc. This is effective for daily monitoring of manufacturing data, etc.

[0137] Next, the hardware configuration of the data processing device according to each of the above-described embodiments will be described. Fig. 25 is a block diagram showing the hardware configuration of the data processing device 1. The data processing device 1 includes a CPU (Central Processing Unit) 801, a RAM (Random Access Memory) 802, a ROM (Read Only Memory) 803, a storage 804, a display device 805, an input device 806, and a communication device 807. The CPU 801, RAM 802, ROM 803, storage 804, display device 805, input device 806, and communication device 807 are connected to each other via a bus. Note that the display device 805 does not have to be included in the data processing device 1, and may be a peripheral device of the data processing device 1.

[0138] The CPU 801 is a processor that executes arithmetic processing, control processing, etc. according to a program. The CPU 801 uses a predetermined area of ​​the RAM 802 as a working area and executes various processes as the acquisition unit 101, the first calculation unit 102, the second calculation unit 106, and the output unit 103 described above in cooperation with programs stored in the ROM 803, the storage 804, etc.

[0139] The RAM 802 is a memory such as an SDRAM (Synchronous Dynamic Random Access Memory), and operates as a work area for the CPU 801. The ROM 803 is a memory that stores programs and various information in a non-rewritable manner.

[0140] The storage 804 is a device that writes and reads data to a semiconductor storage medium such as a flash memory, a magnetically recordable storage medium such as a hard disk drive (HDD), or an optically recordable storage medium. The storage 804 writes and reads data to the storage medium in accordance with control from the CPU 801. The storage 804 can operate as the manufacturing database 104, the manufacturing process database 108, the user database 109, and the analysis database 110. The manufacturing database 104, the manufacturing process database 108, the user database 109, and the analysis database 110 may be stored in a storage medium separate from the storage 804.

[0141] The display device 805 is a display device such as an LCD (Liquid Crystal Display), etc. The display device 805 operates as the display unit 105, and displays various display screens such as the display screen shown in FIG.

[0142] The input device 806 is an input device such as a mouse, a keyboard, etc. The input device 806 operates as the operation unit 107, receives information input by a user as an instruction signal, and outputs the instruction signal to the CPU 801.

[0143] The communication device 807 communicates with external devices via a network under the control of the CPU 801 .

[0144] The instructions shown in the processing procedures described in the above-described embodiments can be executed based on a software program. A general-purpose computer system can store this program in advance and, by loading this program, achieve effects similar to those of the data processing device described above. The instructions described in the above-described embodiments are recorded as a computer-executable program on a magnetic disk (flexible disk, hard disk, etc.), an optical disk (CD-ROM, CD-R, CD-RW, DVD-ROM, DVD±R, DVD±RW, Blu-ray (registered trademark) Disc, etc.), semiconductor memory, or similar recording medium. The recording medium may take any storage format as long as it is readable by a computer or embedded system. A computer can achieve operations similar to those of the data processing device described in the above-described embodiments by loading the program from the recording medium and having the CPU execute the instructions described in the program based on the program. Of course, the computer may acquire or load the program via a network. In addition, an OS (operating system), database management software, network middleware, etc. running on a computer may execute some of the processes required to realize this embodiment based on instructions from a program installed on the computer or embedded system from a recording medium. Furthermore, the recording medium in this embodiment is not limited to a medium independent of a computer or an embedded system, but also includes a recording medium that stores or temporarily stores a program downloaded via a LAN, the Internet, or the like. Furthermore, the number of recording media is not limited to one, and cases where the processing in this embodiment is executed from multiple media are also included in the recording media in this embodiment, and the media may have any configuration.

[0145] The computer or embedded system in this embodiment is for executing each process in this embodiment based on a program stored on a recording medium, and may be configured as either a device consisting of a single device such as a personal computer or a microcomputer, or a system in which multiple devices are connected to a network. Furthermore, the computer in this embodiment is not limited to a personal computer, but also includes an arithmetic processing unit, a microcomputer, etc. included in information processing equipment, and is a general term for equipment or devices that can realize the functions in this embodiment by a program.

[0146] Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, and are also included in the scope of the invention and its equivalents as defined in the claims. [Explanation of symbols]

[0147] 1 Data processing device, 101 Acquisition unit, 102 First calculation unit, 103 Output unit, 104 Manufacturing database, 105 Display unit, 106 Second calculation unit, 107 Operation unit, 108 Manufacturing process database, 109 User database, 110 Analysis database, 801 CPU, 802 RAM, 803 ROM, 804 Storage, 805 Display device, 806 Input device, 807 Communication device.

Claims

1. an acquisition unit that acquires status data relating to a status of each product and sequence data relating to a manufacturing sequence of each product, the status data being obtained during the manufacturing of each product; a first calculation unit that calculates a first score related to a change in the status data acquired by the acquisition unit based on a first change model that expresses a change in the status data using the order data, using the status data and the order data acquired by the acquisition unit; an output unit that outputs the first score; Equipped with the first calculation unit calculates the first score based on a goodness of fit of the first change model to the state data and the order data and a penalty for complexity of the first change model. Data processing device.

2. The sequence data represents the sequence or time at which the product has undergone a predetermined manufacturing process in the manufacture of the product.

2. The data processing device according to claim 1.

3. the output unit generates visualization data for displaying at least one of the first score and information explaining the first score on a display unit, in accordance with the first score; 3. A data processing device according to claim 1 or 2.

4. the first calculation unit calculates a plurality of first scores for a plurality of pieces of ordinal data; the output unit generates the visualization data so that each of the first scores or information explaining the first scores is displayed in a priority order according to the first scores.

4. The data processing device according to claim 3.

5. the first calculation unit calculates a plurality of first scores for a plurality of pieces of ordinal data; the output unit generates the visualization data such that an amount of information displayed as each of the first scores or information explaining the first scores varies depending on the first scores.

5. A data processing device according to claim 3 or 4.

6. an acquisition unit that acquires status data relating to the status of each product, sequence data relating to the sequence of production of each product, and production condition data relating to the production conditions of each product, which are obtained during the production of each product; a first calculation unit that calculates a first score related to a change in the status data acquired by the acquisition unit based on a first change model that expresses a change in the status data using the order data, using the status data and the order data acquired by the acquisition unit; a second calculation unit that calculates a second score related to a change in the status data acquired by the acquisition unit based on a second change model that expresses a change in the manufacturing condition data using the manufacturing condition data, using the status data and the manufacturing condition data acquired by the acquisition unit; an output unit that outputs the first score and the second score; Equipped with the first calculation unit calculates the first score based on a goodness of fit of the first change model to the state data and the order data and a penalty for complexity of the first change model. Data processing device.

7. The sequence data represents the sequence or time at which the product has undergone a predetermined manufacturing process in the manufacture of the product.

7. The data processing device according to claim 6.

8. the output unit generates visualization data for displaying at least one of the first score and information explaining the first score, and at least one of the second score and information explaining the second score, on a display unit, in accordance with the first score and the second score.

8. A data processing device according to claim 6 or 7.

9. the first calculation unit calculates a plurality of first scores for a plurality of pieces of ordinal data; the second calculation unit calculates a plurality of second scores for a plurality of pieces of manufacturing condition data; the output unit generates the visualization data so that each of the first scores or information explaining the first scores is displayed with a priority according to the first scores, and each of the second scores or information explaining the second scores is displayed with a priority according to the second scores.

9. A data processing device according to claim 8.

10. the first calculation unit calculates a plurality of first scores for a plurality of pieces of ordinal data; the second calculation unit calculates a plurality of second scores for a plurality of pieces of manufacturing condition data; the output unit generates the visualization data such that an amount of information to be displayed as each of the first scores or information explaining the first scores changes depending on the first scores, and an amount of information to be displayed as each of the second scores or information explaining the second scores changes depending on the second scores.

10. A data processing device according to claim 8 or 9.

11. the output unit generates the visualization data according to a reliability calculated based on the first score and the second score.

11. A data processing device according to any one of claims 8 to 10.

12. the output unit generates the visualization data in a manner that emphasizes or deemphasizes at least one of the first score, information explaining the first score, the second score, and information explaining the second score, according to the first score and the second score.

12. A data processing device according to any one of claims 8 to 11.

13. the first calculation unit calculates the first score according to the number of conditions in the manufacturing condition data.

13. A data processing device according to any one of claims 6 to 12.

14. the second calculation unit calculates the second score based on a goodness of fit of the second change model to the state data and the manufacturing condition data and a penalty for complexity of the second change model.

7. The data processing device according to claim 6.

15. obtaining status data relating to a status of each product and sequence data relating to a manufacturing sequence of each product, the status data being obtained during the manufacturing of each product; Using the acquired status data and the order data, calculate a first score related to the change in the acquired status data based on a first change model that expresses a change in the status data using order data; outputting the first score; Equipped with calculating the first score includes calculating the first score based on a goodness of fit of the first change model to the state data and the order data and a penalty for the complexity of the first change model. Data processing methods.

16. Acquiring status data relating to the status of each product, sequence data relating to the order in which each product is manufactured, and manufacturing condition data relating to the manufacturing conditions of each product, which are obtained during the manufacturing of each product; Using the acquired status data and the order data, calculate a first score related to the change in the acquired status data based on a first change model that expresses a change in the status data using order data; Using the acquired status data and the manufacturing condition data, calculate a second score related to the change in the acquired status data based on a second change model that expresses a change in the manufacturing condition data using the manufacturing condition data; outputting the first score and the second score; Equipped with calculating the first score includes calculating the first score based on a goodness of fit of the first change model to the state data and the order data and a penalty for the complexity of the first change model. Data processing methods.

17. obtaining status data relating to a status of each product and sequence data relating to a manufacturing sequence of each product, the status data being obtained during the manufacturing of each product; Using the acquired status data and the order data, calculate a first score related to the change in the acquired status data based on a first change model that expresses a change in the status data using order data; outputting the first score; A data processing program for causing a computer to execute the above, calculating the first score includes calculating the first score based on a goodness of fit of the first change model to the state data and the order data and a penalty for the complexity of the first change model. Data processing program.

18. Acquiring status data relating to the status of each product, sequence data relating to the order in which each product is manufactured, and manufacturing condition data relating to the manufacturing conditions of each product, which are obtained during the manufacturing of each product; Using the acquired status data and the order data, calculate a first score related to the change in the acquired status data based on a first change model that expresses a change in the status data using order data; Using the acquired status data and the manufacturing condition data, calculate a second score related to the change in the acquired status data based on a second change model that expresses a change in the manufacturing condition data using the manufacturing condition data; outputting the first score and the second score; A data processing program for causing a computer to execute the above, calculating the first score includes calculating the first score based on a goodness of fit of the first change model to the state data and the order data and a penalty for the complexity of the first change model. Data processing program.

Images