Manufacturing process control method

Abstract

This invention provides a manufacturing process control method that can improve the manufacturing process by effectively using the acquired data. [Solution] The manufacturing process control method comprises four steps. In the first step, a data master is created by analyzing first and second training data, and join conditions for combining the first and second training data are extracted. In the second step, first and second measured data are cleansed based on the data master, and combined data is created by combining the first and second measured data based on the join conditions. In the third step, first and second correlations are calculated by analyzing multiple factors and multiple characteristics included in the combined data. In the fourth step, characteristics requiring improvement are determined, improvement measures to improve the characteristics requiring improvement are formulated based on the first and second correlations, the impact of the improvement measures on other characteristics is estimated, and information on the improvement measures and their impacts is displayed.

Description

Technical Field

[0001] Embodiments of the present invention relate to a manufacturing process control method.

Background Art

[0002] There is a manufacturing process control method for controlling a manufacturing process by analyzing data acquired in the manufacturing process. In such a manufacturing process control method, it is required to effectively use the acquired data to improve the manufacturing process.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The problem to be solved by the present invention is to provide a manufacturing process control method capable of effectively using the acquired data to improve the manufacturing process.

Means for Solving the Problems

[0005] The manufacturing process control method according to the embodiment comprises first to fourth steps. In the first step, a data master is created by analyzing first training data and second training data. The first training data is the training data for the first manufacturing process. The second training data is the training data for the second manufacturing process. In the first step, join conditions for combining the first training data and the second training data are extracted. In the second step, first measured data and second measured data are cleansed based on the data master. The first measured data is the measured data for the first manufacturing process. The second measured data is the measured data for the second manufacturing process. In the second step, combined data is created by combining the first measured data and the second measured data based on the join conditions. In the third step, a first correlation and a second correlation are calculated by analyzing a plurality of factors and a plurality of characteristics included in the combined data. The first correlation is the correlation between the plurality of factors and the plurality of characteristics. The second correlation is the correlation between the plurality of factors. In the fourth step, the characteristics requiring improvement are determined. These characteristics are those among the multiple characteristics that require improvement. In the fourth step, improvement measures are formulated to improve the characteristics requiring improvement based on the first and second correlations. In the fourth step, the impact of the improvement measures on at least one characteristic other than the characteristics requiring improvement among the multiple characteristics is estimated based on the first and second correlations. In the fourth step, information about the improvement measures and their impacts is displayed. [Brief explanation of the drawing]

[0006] [Figure 1] This is a block diagram schematically representing the manufacturing process control system according to the embodiment. [Figure 2] This is a flowchart illustrating the manufacturing process control method according to the embodiment. [Figure 3] This is a flowchart representing the first step of the manufacturing process control method according to the embodiment. [Figure 4]Figures 4(a) to 4(c) are tables showing an example of the extraction of coupling conditions for the first step of the manufacturing process control method according to the embodiment. [Figure 5] This is a flowchart representing the second step of the manufacturing process control method according to the embodiment. [Figure 6] This table shows an example of actual measurement data before cleansing. [Figure 7] This table shows an example of actual measurement data after cleansing. [Figure 8] This is a flowchart representing the third step of the manufacturing process control method according to the embodiment. [Figure 9] This is a flowchart showing the fourth step of the manufacturing process control method according to the embodiment. [Figure 10] This is a flowchart illustrating the determination of the characteristics requiring improvement in the fourth step of the manufacturing process control method according to the embodiment. [Modes for carrying out the invention]

[0007] Each embodiment of the present invention will be described below with reference to the drawings. Drawings are schematic or conceptual, and the relationships between the thickness and width of each part, as well as the ratios of the sizes of different parts, are not necessarily identical to those of reality. Even when representing the same part, the dimensions and ratios may be depicted differently in different drawings. In this specification and in each figure, elements similar to those described above are denoted by the same reference numerals with respect to previously shown figures, and detailed explanations are omitted as appropriate.

[0008] Figure 1 is a block diagram schematically representing a manufacturing process control system according to an embodiment. As shown in Figure 1, the manufacturing process control system 100 according to this embodiment includes a control device 10, a storage device 20, and a display device 30. The manufacturing process control system 100 is capable of executing the manufacturing process control method described later. The manufacturing process control system 100 is capable of executing, for example, the first to fourth steps described later.

[0009] The control device 10 is capable of performing various calculation processes in the first to fourth steps. For example, the control device 10 can perform the creation of a data master and the extraction of joining conditions in the first step. For example, the control device 10 can perform the cleansing of measured data and the joining of measured data (creation of joined data) in the second step. For example, the control device 10 can perform the analysis of factors and characteristics of the joined data and the calculation of first and second correlations in the third step. For example, the control device 10 can perform the determination of characteristics requiring improvement, the formulation of improvement measures, and the estimation of the impact on other characteristics in the fourth step. The control device 10 includes, for example, a CPU (Central Processing Unit). The control device 10 may, for example, perform all calculation processes in one area, or it may be divided into multiple areas for each executable calculation process.

[0010] The storage device 20 is capable of storing information related to various arithmetic processes in the first to fourth steps. For example, the storage device 20 can store information such as training data, data master, combination conditions, measured data before cleansing, measured data after cleansing, combined data, first and second correlations, characteristics requiring improvement, improvement measures, and influence on other characteristics. The storage device 20 is electrically connected to the control device 10. The storage device 20 can output the stored information to the control device 10. The storage device 20 can store information input from the control device 10. The storage device 20 includes, for example, memory or storage. The storage device 20 may, for example, store all information in a single area, or it may be divided into multiple areas according to the type of information that can be stored.

[0011] The display device 30 is capable of displaying information related to various calculation processes in the first to fourth steps. The display device 30 can display information such as the first and second correlations, improvement measures, and the impact on other characteristics. The display device 30 is electrically connected to the control device 10. The display device 30 is capable of displaying information input from the control device 10. The display device 30 includes, for example, a display. The display device 30 may, for example, display all information in one area, or it may be divided into multiple areas according to the information that can be displayed.

[0012] Figure 2 is a flowchart illustrating the manufacturing process control method according to the embodiment. As shown in Figure 2, the manufacturing process control method according to this embodiment comprises a first step, a second step, a third step, and a fourth step. The first to fourth steps are performed in the order of the first step, the second step, the third step, and the fourth step.

[0013] First, the control device 10 performs the first step (step S101). In the first step, the control device 10 creates a data master by analyzing the first training data and the second training data. Also in the first step, the control device 10 extracts the joining conditions for combining the first training data and the second training data. The first training data is the training data for the first manufacturing process. The second training data is the training data for the second manufacturing process. The training data is, for example, the data for each manufacturing process when a good product is manufactured. The first manufacturing process and the second manufacturing process are each one of several manufacturing processes for manufacturing one item. The second manufacturing process is performed after the first manufacturing process. The second manufacturing process may be a process that is continuous with the first manufacturing process, or it may be a process that is not continuous with the first manufacturing process. The first step will be described later.

[0014] Note that in the first step, the control device 10 may analyze other teacher data in addition to the first teacher data and the second teacher data. The other teacher data is teacher data of manufacturing processes other than the first manufacturing process and the second manufacturing process (other manufacturing processes). The other manufacturing process may be a manufacturing process performed before the first manufacturing process, a manufacturing process performed between the first manufacturing process and the second manufacturing process, or a manufacturing process performed after the second manufacturing process. Also, in the first step, the control device 10 may extract a combination condition for combining the first teacher data and the second teacher data with the other teacher data in addition to the combination condition for combining the first teacher data and the second teacher data.

[0015] Next, the control device 10 performs the second step (step S102). In the second step, the control device 10 performs cleansing of the first measured data and the second measured data based on the data master. Also, in the second step, the control device 10 creates combined data by combining the first measured data and the second measured data based on the combination condition. The first measured data is the measured data of the first manufacturing process. The second measured data is the measured data of the second manufacturing process. The second step will be described later.

[0016] Note that in the second step, the control device 10 may perform cleansing of other measured data in addition to cleansing of the first measured data and the second measured data. Also, in the second step, the control device 10 may create combined data by combining the first measured data, the second measured data, and the other measured data.

[0017] Next, the control device 10 performs the third step (step S103). In the third step, the control device 10 calculates the first correlation relationship and the second correlation relationship by analyzing a plurality of factors and a plurality of characteristics included in the combined data. The first correlation relationship is the correlation relationship between the plurality of factors and the plurality of characteristics. The second correlation relationship is the correlation relationship between the plurality of factors.

[0018] The combined data includes data on multiple factors and data on multiple characteristics. Here, "factors" are numerical values ​​that indicate, for example, the operating conditions of the equipment used in manufacturing (e.g., applied pressure, heating temperature, rotation speed, etc.) or the environmental conditions during manufacturing (e.g., temperature, humidity, etc.). "Characteristics" are numerical values ​​that indicate, for example, the performance of the manufactured goods. In other words, the first correlation is, for example, the correlation between the operating conditions of the equipment used in manufacturing or the environmental conditions during manufacturing and the performance of the manufactured goods. The second correlation is, for example, the correlation between the operating conditions of the equipment used in manufacturing and the environmental conditions during manufacturing and each other. The third process will be described later.

[0019] Next, the control device 10 performs the fourth step (step S104). In the fourth step, the control device 10 determines the characteristics that need improvement. The characteristics that need improvement are those characteristics among several characteristics that require improvement. Also in the fourth step, the control device 10 formulates improvement measures based on the first and second correlations. The improvement measures are measures to improve the characteristics that need improvement. Improvement measures include, for example, changing factors related to the characteristics that need improvement (change factors described later). Also in the fourth step, the control device 10 estimates the impact of the improvement measures on at least one characteristic other than the characteristics that need improvement among several characteristics, based on the first and second correlations. Also in the fourth step, the control device 10 displays information about the formulated improvement measures and the estimated impact on the display device 30. The fourth step will be described later.

[0020] The first step will be explained below. Figure 3 is a flowchart showing the first step of the manufacturing process control method according to the embodiment. As shown in Figure 3, in the first step, the control device 10 first acquires training data (first training data and second training data) from the storage device 20 (step S201).

[0021] Next, the control device 10 determines whether the acquired training data is in database format (step S202). If the acquired training data is not in database format (step S202: No), the control device 10 converts the acquired training data into database format (step S203). Database format refers to relational databases or tabular data. The conversion to database format varies depending on the format of the training data. For example, if the database is not a relational database, it is converted between databases; if it is an image, it is converted by a program with optical character recognition capabilities. Database conversion is performed, for example, when migrating data between different database systems. Database conversion is performed by methods such as exporting and importing data, using ETL (Extract, Transform, Load) tools, using database links, and using custom scripts. In the exporting and importing method, for example, data exported from a source database is imported into a target database. In the exporting and importing method, for example, intermediate formats such as CSV or JSON are used. In the method using ETL tools, for example, the ETL tool automates data extraction, transformation, and loading. ETL tools such as Talend, Apache iFi, and Informatica can be used. The database linking method allows for direct data exchange between different databases. Examples of database linking features include Oracle Database's DB linking function. The custom scripting method allows for direct data exchange between different databases using data transformation scripts created with programming languages ​​such as Python or SQL.

[0022] If the acquired training data is in database format (Step S202: Yes), or if the acquired training data has been converted to database format (Step S203), the control device 10 analyzes the training data (Step S204). For example, in analyzing the training data, the control device 10 analyzes the column classification, format, key column, and numerical distribution of the training data. For column classification, for example, it classifies the data into date columns (inspection date, measurement date, etc.), numerical columns (tolerance, dimension, etc.), identification columns (product name, order number, etc.). For column classification, for example, if the most frequent format is a date, it can be classified as a date column; if the most frequent format is not a date and the most frequent format is a numerical value without characters, it can be classified as a numerical column; and if the most frequent format is not a date and the most frequent format is not a numerical value without characters, it can be classified as an identification column.

[0023] The control device 10 creates a data master based on the analysis results of the training data (step S205). The data master is a set of data that will be used as a reference and basis for comparison with the measured data in the second step. The control device 10 also extracts joining conditions for combining the first training data and the second training data based on the analysis results of the training data (step S206).

[0024] The extraction of the coupling conditions in the first step will be explained below. Figures 4(a) to 4(c) are tables showing an example of the extraction of coupling conditions for the first step of the manufacturing process control method according to the embodiment. Figure 4(a) is a table representing the first training data. Figure 4(b) is a table representing the second training data. Figure 4(c) is a table representing the combined data obtained by combining the first training data and the second training data.

[0025] As shown in Figure 4(a), the first training data includes the columns "Inspection Date," "Dimension 1," and "Individual Number." The "Inspection Date" and "Dimension 1" columns may contain duplicate values. On the other hand, the "Individual Number" column does not contain duplicate values. In other words, the "Individual Number" column is a key column with unique values. Thus, the first training data includes a first key column with unique values.

[0026] As shown in Figure 4(b), the second training data includes the columns "Test Date," "Test Result," "Individual Number," and "Serial Number." The "Test Date" and "Test Result" columns may contain duplicate values. On the other hand, the "Individual Number" and "Serial Number" columns do not contain duplicate values. In other words, the "Individual Number" and "Serial Number" columns are key columns with unique values. Thus, the second training data includes a second key column with unique values.

[0027] When extracting join conditions, the control device 10 obtains the first key array of the first training data and the second key column of the second training data. Then, the control device 10 extracts join conditions based on the similarity between the first key column and the second key column. In this example, the similarity between the "individual number" column in the second key column and the first key column ("individual number" column) is higher than the similarity between the "manufacturing number" column in the second key column and the first key column ("individual number" column). Therefore, the control device 10 extracts the "individual number" column as the join condition. By joining the first training data and the second training data based on the extracted join condition ("individual number" in this example), it is possible to create one combined data set by combining the two training data sets, as shown in Figure 4(c). The join condition extracted in this way is used to join the first measured data and the second measured data in the second step.

[0028] The second step will be explained below. Figure 5 is a flowchart showing the second step of the manufacturing process control method according to the embodiment. As shown in Figure 5, in the second step, the control device 10 first acquires measured data (first measured data and second measured data) from the storage device 20 (step S301).

[0029] Next, the control device 10 determines whether the acquired measured data is in database format (step S302). If the acquired measured data is not in database format (step S302: No), the control device 10 converts the acquired measured data into database format (step S303). The conversion to database format is performed, for example, in the same manner as in step S203 described above.

[0030] If the acquired measured data is in database format (Step S302: Yes), or if the acquired measured data has been converted to database format (Step S303), the control device 10 cleans the measured data based on the data master (Step S304). For example, during the cleansing of the measured data, the control device 10 handles abnormal values ​​and corrects column correspondences.

[0031] The control device 10 creates combined data by combining the first measured data and the second measured data based on the combination conditions (step S305).

[0032] The following describes the data cleansing process for the second step, which involves measuring actual data. Figure 6 is a table showing an example of actual measurement data before cleansing. Figure 7 is a table showing an example of measured data after cleansing.

[0033] As shown in Figure 6, in this example, the measured data before cleansing includes data from three groups: GR1, GR2, and GR3. In GR1, numerical values ​​are entered in the correct columns. In GR2 and GR3, numerical values ​​are entered in the wrong columns. More specifically, in GR2, "OK" is entered in the "Dimension 1" column, the data for "Dimension 1" is entered in the "Dimension 2" column, the data for "Dimension 2" is entered in the "Dimension 3" column, and the data for "Dimension 3" is entered in the "Dimension 4" column. In GR3, the data for "Dimension 1" is entered in the "Dimension 3" column, the data for "Dimension 2" is entered in the "Dimension 1" column, and the data for "Dimension 3" is entered in the "Dimension 2" column.

[0034] For example, suppose in the data master, "Dimension 1" is a value of 100.00 ± 0.05, "Dimension 2" is a value of 150.00 ± 0.05, and "Dimension 3" is a value of 50.00 ± 0.05. In this case, the control device 10 determines that the correct values ​​have been entered in the correct columns in GR1 because "Dimension 1", "Dimension 2", and "Dimension 3" in GR1 satisfy the ranges of "Dimension 1", "Dimension 2", and "Dimension 3" in the data master. On the other hand, the control device 10 determines that the incorrect values ​​have been entered in the wrong columns in GR2 and GR3 because "Dimension 1", "Dimension 2", and "Dimension 3" in GR2 and GR3 do not satisfy the ranges of "Dimension 1", "Dimension 2", and "Dimension 3" in the data master.

[0035] Then, because the numerical value entered in the "Dimension 2" column of GR2 satisfies the range of "Dimension 1" in the data master, the control device 10 inputs the numerical value that was entered in the "Dimension 2" column of GR2 in Figure 6 into the "Dimension 1" column of GR2, as shown in Figure 7. Similarly, because the numerical value entered in the "Dimension 3" column of GR2 satisfies the range of "Dimension 2" in the data master, the control device 10 inputs the numerical value that was entered in the "Dimension 3" column of GR2 in Figure 6 into the "Dimension 2" column of GR2, as shown in Figure 7. Similarly, because the numerical value entered in the "Dimension 4" column of GR2 satisfies the range of "Dimension 3" in the data master, the control device 10 inputs the numerical value that was entered in the "Dimension 4" column of GR2 in Figure 6 into the "Dimension 3" column of GR2, as shown in Figure 7.

[0036] Similarly, since the numerical value entered in the "Dimension 3" column of GR3 satisfies the range of "Dimension 1" in the data master, the control device 10 inputs the numerical value that was entered in the "Dimension 3" column of GR3 in Figure 6 into the "Dimension 1" column of GR3, as shown in Figure 7. Similarly, since the numerical value entered in the "Dimension 1" column of GR3 satisfies the range of "Dimension 2" in the data master, the control device 10 inputs the numerical value that was entered in the "Dimension 1" column of GR3 in Figure 6 into the "Dimension 2" column of GR3, as shown in Figure 7. Similarly, since the numerical value entered in the "Dimension 2" column of GR3 satisfies the range of "Dimension 3" in the data master, the control device 10 inputs the numerical value that was entered in the "Dimension 2" column of GR3 in Figure 6 into the "Dimension 3" column of GR3, as shown in Figure 7.

[0037] The third step will be explained below. Figure 8 is a flowchart showing the third step of the manufacturing process control method according to the embodiment. As shown in Figure 8, in the third step, the control device 10 first calculates statistical quantities of the factors and characteristics of the combined data (step S401). Examples of statistical quantities that are calculated include the mean, minimum, maximum, and standard deviation.

[0038] Next, the control device 10 calculates the first correlation and the second correlation based on the values ​​of the factors and characteristics (step S402).

[0039] Next, the control device 10 causes the display device 30 to display a figure showing at least one of the first correlation and the second correlation (step S403). The figure showing at least one of the first correlation and the second correlation may be, for example, a scatter plot, a histogram, or a correlation matrix. Step S403 is optional.

[0040] The fourth step will be explained below. Figure 9 is a flowchart showing the fourth step of the manufacturing process control method according to the embodiment. As shown in Figure 9, in the fourth step, the control device 10 first determines the characteristics that need improvement (step S501). The determination of the characteristics that need improvement will be described later.

[0041] Next, the control device 10 estimates multiple related factors based on the first and second correlations (step S502). The multiple related factors are factors among the multiple factors that are related to the characteristics that need improvement. The multiple related factors include, for example, directly related factors and indirectly related factors. Directly related factors are factors that correlate with the characteristics that need improvement. Directly related factors are estimated, for example, based on the degree of correlation with the characteristics that need improvement, using the first correlation. Indirectly related factors are factors that correlate with the directly related factors. Indirectly related factors are estimated, for example, based on the degree of correlation with the directly related factors, using the second correlation.

[0042] Next, the control device 10 determines one change factor (step S503). The change factor is, for example, one factor among several related factors that is easier to change. Whether or not a change is easy is determined based on, for example, the cost of the change, the time required for the change, and the impact of the change on other factors. For example, if the cost of the change is small, it is determined that the change is easier. For example, if the time required for the change is short, it is determined that the change is easier. For example, if the impact of the change on other factors is small, it is determined that the change is easier.

[0043] Next, the control device 10 formulates improvement measures (step S504). More specifically, the control device 10 decides to improve the characteristics that need improvement by changing the numerical values ​​of the change factors. The control device 10 also decides how much to change the numerical values ​​of the change factors.

[0044] Next, the control device 10 estimates the change characteristics based on the first and second correlations (step S505). The change characteristics are those that change by changing the numerical values ​​of the change factors among several characteristics.

[0045] Next, the control device 10 estimates the impact on the change characteristics caused by changing the numerical values ​​of the change factors based on the first and second correlations (step S506).

[0046] Next, the control device 10 displays information on the formulated improvement measures and their impact on the estimated change characteristics on the display device 30 (step S507).

[0047] Furthermore, the possibility of changing multiple factors may be predetermined. In this case, for example, in step S503, the control device 10 determines which factors to change from among the multiple related factors that can be changed. Alternatively, for example, if all of the multiple related factors are factors that cannot be changed, the control device 10 may omit steps S504 to S507 and display on the display device 30 a measure suggesting "stopping production" or "significantly changing the manufacturing conditions (factors)".

[0048] The following explains how to determine the characteristics that need improvement in the fourth process. Figure 10 is a flowchart showing the determination of the characteristics requiring improvement in the fourth step of the manufacturing process control method according to the embodiment. As shown in Figure 10, in the fourth step, the control device 10 selects one of several characteristics (step S601) and determines whether the difference between the target value and the measured value of the selected characteristic is greater than or equal to a threshold (step S602).

[0049] If the difference between the target value and the measured value of the selected characteristic is greater than or equal to a threshold (Step S602: Yes), the control device 10 determines that the selected characteristic is a characteristic that needs improvement (Step S603).

[0050] If the difference between the target value and the measured value of the selected characteristic is less than the threshold (step S602: No), the control device 10 selects another characteristic from the multiple characteristics (step S604) and returns to step S602.

[0051] The effects and benefits of the manufacturing process control method according to this embodiment will be explained below.

[0052] In manufacturing, there are problems with the difficulty of improving processes due to management of manufacturing processes that are not based on data (relying on intuition) and the personalization of data analysis results. Furthermore, even when analyzing data, errors in reading the data (e.g., errors in optical character recognition) or errors by the person recording the data (e.g., errors in recording data in the wrong place or recording incorrect data) can prevent the data acquired in the manufacturing process from being fully utilized for improvement. Manufacturing process control methods require the effective use of acquired data to improve the manufacturing process.

[0053] In contrast, the manufacturing process control method according to this embodiment allows for the correction of errors in reading data and errors by the person who entered the data, by creating a data master based on training data in the first step and cleaning the measured data based on the data master in the second step. This increases the amount of data available for analysis and allows for the effective use of acquired data (measured data) to improve the manufacturing process. Furthermore, the manufacturing process control method according to this embodiment allows for the more accurate combination of data from multiple processes by extracting combination conditions based on training data in the first step and creating combined data by combining the measured data based on the combination conditions in the second step. In addition, the manufacturing process control method according to this embodiment allows for the formulation of improvement measures in the fourth step, and by estimating and displaying information on the impact of these improvement measures on other characteristics, it becomes easier for the manufacturing process manager to decide whether or not to implement the improvement measures.

[0054] Furthermore, in the manufacturing process control method according to this embodiment, in the first step, data from multiple steps can be more accurately combined by extracting joining conditions based on the similarity between the first key column and the second key column.

[0055] Furthermore, in the manufacturing process control method according to the embodiment, the first and second correlations can be presented in an easily understandable manner to the manufacturing process manager by displaying a diagram showing at least one of the first and second correlations in the third step.

[0056] Furthermore, in the manufacturing process control method according to this embodiment, in the fourth step, the characteristics that need improvement are determined based on the difference between the target value and the measured value of multiple characteristics, thereby improving the characteristics that require further improvement.

[0057] Furthermore, in the manufacturing process control method according to this embodiment, in the fourth step, multiple related factors related to the characteristics requiring improvement are estimated from among multiple factors, one change factor that is easier to change is determined from among the multiple related factors, and an improvement measure is formulated to improve the characteristics requiring improvement by changing the numerical value of the change factor, thereby enabling the formulation of an improvement measure that can improve the characteristics requiring improvement more easily and effectively.

[0058] Furthermore, in the manufacturing process control method according to the embodiment, in the fourth step, by estimating the change characteristics that change by changing the numerical value of a change factor among multiple characteristics, and by estimating the impact of changing the numerical value of the change factor on the change characteristics, it is possible to inform the manufacturing process manager of more accurate information about the impact of improvement measures on other characteristics.

[0059] The embodiment may include the following configurations.

[0060] (Composition 1) The first step involves creating a data master by analyzing the first training data, which is the training data for the first manufacturing process, and the second training data, which is the training data for the second manufacturing process, and extracting join conditions for combining the first training data and the second training data. A second step involves cleansing the first measured data, which is the measured data of the first manufacturing process, and the second measured data, which is the measured data of the second manufacturing process, based on the data master, and creating combined data by combining the first measured data and the second measured data based on the combination conditions. A third step involves analyzing multiple factors and multiple characteristics included in the combined data to calculate a first correlation, which is the correlation between the multiple factors and the multiple characteristics, and a second correlation, which is the correlation between the multiple factors. A fourth step involves determining which of the aforementioned multiple characteristics require improvement, formulating improvement measures to improve the aforementioned characteristics based on the first and second correlations, estimating the impact of the improvement measures on at least one characteristic other than the aforementioned characteristics based on the first and second correlations, and displaying information about the improvement measures and the impact. A manufacturing process control method comprising the following features.

[0061] (Configuration 2) A manufacturing process control method according to Configuration 1, wherein in the first step, the join condition is extracted based on the similarity between a first key column having unique values ​​included in the first training data and a second key column having unique values ​​included in the second training data.

[0062] (Composition 3) A manufacturing process control method according to configuration 1 or 2, wherein in the third step, a diagram showing at least one of the first correlation and the second correlation is displayed.

[0063] (Composition 4) A manufacturing process control method according to any one of configurations 1 to 3, wherein in the fourth step, the characteristic requiring improvement is determined based on the difference between the target value and the measured value of the plurality of characteristics.

[0064] (Composition 5) A manufacturing process control method according to any one of configurations 1 to 4, wherein in the fourth step, based on the first correlation and the second correlation, a plurality of related factors related to the characteristic requiring improvement are estimated from among the plurality of factors, one change factor that is easier to change is determined from among the plurality of related factors, and an improvement measure is formulated to improve the characteristic requiring improvement by changing the numerical value of the change factor.

[0065] (Composition 6) The manufacturing process control method according to configuration 5, wherein in the fourth step, based on the first correlation and the second correlation, the change characteristic among the plurality of characteristics is estimated to change by changing the numerical value of the change factor, and the effect of changing the numerical value of the change factor on the change characteristic is estimated.

[0066] As described above, according to the embodiment, it is possible to provide a manufacturing process control method that can improve the manufacturing process by effectively using the acquired data.

[0067] The embodiments of the present invention have been illustrated above, but these embodiments are presented as examples only and are not intended to limit the scope of the invention. These novel embodiments can be implemented 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 variations are included in the scope and spirit of the invention, as well as in the claims of the invention and its equivalents. [Explanation of Symbols]

[0068] 10: Control device 20: Storage device 30:Display device 100: Manufacturing process control system

Claims

1. A first step involves creating a data master by analyzing the first training data, which is the training data for the first manufacturing process, and the second training data, which is the training data for the second manufacturing process, and extracting join conditions for combining the first training data and the second training data. A second step involves cleansing the first measured data, which is the measured data of the first manufacturing process, and the second measured data, which is the measured data of the second manufacturing process, based on the data master, and creating combined data by combining the first measured data and the second measured data based on the combination conditions. A third step involves analyzing multiple factors and multiple characteristics included in the combined data to calculate a first correlation, which is the correlation between the multiple factors and the multiple characteristics, and a second correlation, which is the correlation between the multiple factors. A fourth step involves determining which of the aforementioned multiple characteristics require improvement, formulating improvement measures to improve the aforementioned characteristics based on the first and second correlations, estimating the impact of the improvement measures on at least one characteristic other than the aforementioned characteristics based on the first and second correlations, and displaying information about the improvement measures and the impact. A manufacturing process control method comprising:

2. The manufacturing process control method according to claim 1, wherein in the first step, the joining conditions are extracted based on the similarity between a first key column having unique values ​​included in the first training data and a second key column having unique values ​​included in the second training data.

3. The manufacturing process control method according to claim 1, wherein in the third step, a diagram showing at least one of the first correlation and the second correlation is displayed.

4. The manufacturing process control method according to claim 1, wherein in the fourth step, the characteristic requiring improvement is determined based on the difference between the target value and the measured value of the plurality of characteristics.

5. The manufacturing process control method according to claim 1, wherein in the fourth step, based on the first correlation and the second correlation, a plurality of related factors related to the characteristic requiring improvement are estimated from among the plurality of factors, one change factor that is easier to change is determined from among the plurality of related factors, and an improvement measure is formulated to improve the characteristic requiring improvement by changing the numerical value of the change factor.

6. The manufacturing process control method according to claim 5, wherein in the fourth step, based on the first correlation and the second correlation, the change characteristic among the plurality of characteristics is estimated to change by changing the value of the change factor, and the effect of changing the value of the change factor on the change characteristic is estimated.

Images