JP6794283B2 - Production process analyzer and production control system using it - Google Patents

Description

The present invention relates to a production process analyzer and a production control system using the same.

In order to efficiently produce products, that is, to improve the yield, a system that analyzes the production process based on various factors related to production has been proposed.

For example, Patent Document 1 describes a data processing method capable of selecting an explanatory variable having a high influence on the objective variable from a plurality of explanatory variables of the continuous quantity associated with the objective variable of the continuous quantity.

JP-A-2007-329415

The method described in Patent Document 1 identifies conditions affecting production by analyzing various factors and detecting explanatory variables having a high degree of influence. However, even if an explanatory variable that affects production is identified, it may be difficult to adjust improvement measures corresponding to the explanatory variable.

The present invention solves the above-mentioned problems, efficiently executes an analysis of production factors of a product, and outputs an analysis result that makes it easy to consider improvement measures, and a production process analyzer capable of outputting an analysis result, and a production using the same. The purpose is to provide a management system.

The present invention for achieving the above object is a production process analyzer, which is a production process acquisition unit that acquires information on a production flow including the order of production processes and the timing of inspection, and production based on the production flow. It has a history data acquisition unit that acquires history data that is information on inspection results of a plurality of products, and a processing unit that analyzes a production process based on the production flow and the inspection results. Is a specific processing unit that identifies the target inspection that is the inspection to be evaluated from the production flow, and the production process, inspection result, and production that are executed before the target inspection and the target inspection of the product from the historical data. A data extraction unit that extracts information on the final measurement result of a product and a permissible range are set for the final measurement result, and each product is classified into a conforming product that meets the permissible range and a nonconformity that does not meet the permissible range based on the setting. Classify into products, execute decision tree analysis using the information extracted by the data extraction unit as explanatory variables, and execute decision tree analysis to calculate the occurrence probability of nonconforming products for each case classified in each decision tree. It is characterized by having a decision tree analysis execution unit.

Further, it is preferable that the explanatory variables are classified by a threshold value set for the test result.

Further, it is preferable that the decision tree analysis execution unit changes the threshold values of the explanatory variables and executes the decision tree analysis for each threshold value.

Further, it is preferable that the processing unit further includes an analysis result output unit that outputs the condition when a condition in which the occurrence rate of nonconforming products is equal to or higher than a predetermined probability is detected as a result of execution by the decision tree analysis execution unit.

Further, the present invention is a production control system, wherein the production process analyzer described in any of the above and a production flow changing unit that changes the production flow based on the analysis result of the production process analyzer are provided. It is characterized by having.

According to the present invention, the analysis of explanatory variables can be performed based on the order of the steps of producing the product. This makes it possible to analyze how much nonconforming products occur in the final product based on the results of the evaluation target, and eliminate nonconforming products by determining whether the final product is conforming or nonconforming based on the evaluation target results. You can judge the possibility of being able to do it. As a result, it is possible to make an efficient judgment based on the results of each inspection, and it is possible to adjust each condition of the production process to a more appropriate condition.

FIG. 1 is a block diagram showing a schematic configuration of a production control system having the production process analyzer of the present embodiment. FIG. 2 is a conceptual diagram showing a part of the production flow. FIG. 3 is a conceptual diagram showing an example of historical data. FIG. 4 is a block diagram showing a schematic configuration of a production process analyzer. FIG. 5 is a flowchart showing an example of processing of the production process analyzer. FIG. 6 is a flowchart showing an example of processing by the decision tree analysis processing unit. FIG. 7 is a conceptual diagram showing an example of the analysis result. FIG. 8 is a conceptual diagram showing an example of the analysis result. FIG. 9 is a flowchart showing an example of processing of the production process analyzer. FIG. 10 is a flowchart showing an example of processing in the production management system.

Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that the present invention is not limited to this embodiment, and when there are a plurality of embodiments, the present invention also includes a combination of the respective embodiments.

FIG. 1 is a block diagram showing a schematic configuration of a production control system having the production process analyzer of the present embodiment. The production control system 10 manufactures a product. The product can be a variety of articles. The product may be a part or a final product. The production control system 10 includes a production line 12, a measuring device 14, a production control device 16, and a production process analysis device 20.

The production line 12 performs processing on the object and produces the product. The production line 12 has various devices for processing. Here, the processing includes various machining such as casting, cutting, welding, polishing, and heat treatment, and chemical processing such as synthesis and film formation. The production line 12 executes processing of a plurality of steps to produce a product.

The measuring device 14 executes an inspection during processing of the object and an inspection of the product produced on the production line 12. The measuring device 14 measures various measurement parameters executed as inspections of an object or a product. Specifically, the dimensions, roundness, surface roughness, composition, crystal structure, etc. of the object or product are exemplified. The measuring device 14 has a measuring device capable of measuring the parameters to be measured. The measuring device 14 may be a part of the production line 12.

The production control device 16 controls the production by the production line 12 and the measurement by the measuring device 14. The production management device 16 is a device having an arithmetic processing function, a storage function, and an input / output function, for example, a personal computer. The production control device 16 has a production flow data 22, a history data 24, and a production flow changing unit 26.

FIG. 2 is a conceptual diagram showing a part of the production flow. The production flow data 22 is data on the production process for producing the product and the order of inspection. The production flow data 22 shown in FIG. 2 indicates that the processes are executed in the order of the first step, the second step, the first measurement, the third step, the second measurement, and the fourth step. The first step, the second step, the third step, and the fourth step are processing processes executed on the production line 12. Each step may include a plurality of processing processes or may be one processing process. The first measurement and the second measurement are measurements performed by the measuring device 14. Each measurement may include a plurality of measurements or may be one measurement. That is, the result detected in each measurement may be a plurality of measurement results or one measurement result. The production flow 22 shown in FIG. 2 is a part of the production flow, and the measurement and the process may be continued after the fourth process.

FIG. 3 is a conceptual diagram showing an example of historical data. As shown in FIG. 3, the history data 24 stores data of measurement results (inspection results) executed when the products are processed based on the production flow 22 for each product. The history data 24 separates the measurement parameters performed in each measurement as factors, and includes the measurement results of each factor (measurement parameter). In the example shown in FIG. 3, in the first measurement, three parameters of the first factor, the second factor, and the third factor are measured, and in the second measurement, the two parameters of the fourth factor and the fifth factor are measured. To. The history data 24 also includes the results measured in the inspection of the product. In the measurement of the product (finished product), three parameters of the Ath factor, the Bth factor, and the Cth factor are measured. The number of measurement parameters in each measurement is an example, and more measurement parameters may be measured and included in the history data in the actual measurement. In addition, the product includes a conforming product that satisfies the permissible range and a non-conforming product that does not meet the permissible range as a result of measurement in the inspection of the product. In addition, the historical data may further include information indicating whether the product is a conforming product or a non-conforming product.

The production flow changing unit 26 performs a process of changing the production flow data 22 based on the result of analysis by the production process analyzer 20. For example, the production flow changing unit 26 changes the criteria for determining conforming products and non-conforming products based on the results of the first measurement and the second measurement, changes the measurement parameters executed in each measurement, and processes the processing conditions executed in each process. Make changes to.

FIG. 4 is a block diagram showing a schematic configuration of a production process analyzer. The production process analyzer 20 is a device having an arithmetic processing function, a storage function, and an input / output function, for example, a personal computer. The production process analyzer 20 may be integrated with the production control device 16. The production management device 16 includes a production process acquisition unit 30, a history data acquisition unit 32, a processing unit 34, a storage unit 36, an input unit 38, and a display unit 40.

The production process acquisition unit 30 acquires the production flow data 22 stored in the production control device 16. The history data acquisition unit 32 acquires the history data 24 stored in the production management device 16. Here, the history data acquisition unit 32 may acquire only the data of the portion of the history data 24 necessary for analysis, or may acquire all the data. Data transfer between the production process analyzer 20 and the production control device 16 may be performed by wired or wireless communication, or may be performed via a recording medium.

The processing unit 34 has an arithmetic unit such as a CPU. The processing unit 34 reads the program and data corresponding to the processing from the storage unit 36 and executes the program. The processing unit 34 includes a specific processing unit 50, a data extraction unit 52, a decision tree analysis execution unit 54, and an analysis result output unit 56.

The specific processing unit 50 specifies the process and measurement to be analyzed in the production flow. Specifically, the specific processing unit 50 identifies the last measurement in the range to be analyzed, and specifies the process and measurement to be performed prior to the measurement, including the measurement, as the analysis target. Even if the specific processing unit 50 specifies the last measurement of the range to be analyzed on the production flow based on the user's operation, the specific processing unit 50 specifies the last measurement of the range to be analyzed based on the preset conditions. You may.

The data extraction unit 52 extracts from the history data 24 the measurement data included in the range specified by the specific processing unit 50 and the measurement result for the product (finished product) for each product. To do. For example, when the analysis target is up to the first measurement, the results of the first factor, the second factor, the third factor, the A factor, the B factor, and the C factor are extracted for each product. When the analysis targets up to the second measurement, the results of the first factor, the second factor, the third factor, the fourth factor, the fifth factor, the A factor, the B factor, and the C factor are obtained for each product. Extract.

The decision tree analysis execution unit 54 executes the decision tree analysis on the data extracted by the data extract 52. Specifically, the decision tree analysis execution unit 54 classifies each product into conforming products that satisfy the permissible range and nonconforming products that do not meet the permissible range based on the settings, and explains the information extracted by the data extraction unit. A decision tree analysis is performed as a variable, and the probability of occurrence of nonconforming products is calculated for each case classified by each decision tree. The decision tree analysis execution unit 54 can also adjust the threshold value of the explanatory variable that becomes the decision tree. Further, when the conforming product and the non-conforming product are determined in advance, the decision tree analysis execution unit 54 classifies each product into a conforming product satisfying the permissible range and a non-conforming product not satisfying the permissible range based on the setting. It is not necessary to execute the processing to be performed.

The analysis result output unit 56 causes the display unit 40 to display the result executed by the decision tree analysis execution unit 54. The analysis result output unit 56 may process and display the result executed by the decision tree analysis execution unit 54. The analysis result output unit 56 displays as an alert rule a combination of decision trees (combination of determination conditions) in which the probability of occurrence of nonconforming products (nonconformity risk) is equal to or higher than a predetermined probability as a result of execution by the decision tree analysis execution unit 54. It may be displayed on the unit 40.

The storage unit 36 has a storage device such as a RAM, a ROM, and an auxiliary storage device (for example, a hard disk drive). The storage unit 36 stores the program and data read out by the processing unit 34, and also functions as a storage area used for processing by the processing unit 34. The input unit 38 has an input device such as a keyboard and a mouse. The input unit 38 performs input according to the operation by the user who performs the input operation. The display unit 40 has a display device such as a liquid crystal display. The display unit 40 outputs a display according to the processing content of the processing unit 34.

Next, the process executed by the production process analyzer will be described with reference to FIGS. 5 to 8. FIG. 5 is a flowchart showing an example of processing of the production process analyzer. FIG. 6 is a flowchart showing an example of processing by the decision tree analysis processing unit. 7 and 8 are conceptual diagrams showing an example of the analysis results, respectively. The treatments of FIGS. 5 and 6 can be realized by the production process analyzer 20 executing the processing in each part.

The production process analyzer 20 acquires the production flow data 22, and the specific processing unit 50 identifies the timing of the inspection to be evaluated from the acquired production flow (step S12). That is, it is specified which measurement time point is to be analyzed among the plurality of measurement of the production flow.

After the production process analyzer 20 specifies the timing of the inspection (measurement) of the evaluation target, the data extraction unit 52 uses the data extraction unit 52 to display the inspection result (measurement data) before the inspection timing of the evaluation target and the final result data (finished product). (Measurement data of) is extracted (step S14).

After extracting the data, the production process analyzer 20 determines a threshold value that is a boundary between the conforming product and the non-conforming product (step S16). If conforming products and non-conforming products are determined, this process does not have to be executed.

After the threshold value is determined, the production process analyzer 20 performs a decision tree analysis by the decision tree analysis execution unit 54 (step S18). Hereinafter, an example of decision tree analysis will be described with reference to FIG. 7. First, the decision tree analysis execution unit 54 sets the threshold value of the explanatory variable (step S32). Here, the explanatory variables are each factor in each measurement of the analysis result. The factor of the finished product is not included in the explanatory variables. The explanatory variable sets a threshold value for dividing the measurement result of the measurement parameter, which is each factor, into two. Here, the threshold value may be a range. In this case, it is a decision tree whether it is included or not included in the range.

After setting the threshold value of the explanatory variable, the decision tree analysis execution unit 54 executes the decision tree analysis (step S34). Specifically, a branch of the decision tree is created based on the threshold value of the explanatory variable, and the generation rate of nonconforming products of the product passing through each branch is calculated. The decision tree analysis execution unit 54 changes the order and combinations of the decision trees, and calculates the occurrence rate of nonconforming products of the product for the plurality of combinations.

For example, the decision tree analysis execution unit 54 outputs the analysis result 60 shown in FIG. 7 when the first measurement is specified as the inspection timing. The analysis result 60 includes the decision trees 62a and 62b62c to be branched, and the branch results 64a and 64b including the ratio of nonconforming products, the number of products, and the ratio of products that are the ends of the branches and satisfy the branching conditions of the decision tree. 64c, 64d and so on. In the analysis result 60, the decision tree 62a, which is the first branch, is the second factor, one branch becomes the branch result 64a, and the other branch becomes the decision tree 62b. The decision tree 62b is the first factor, and one branch becomes the branch result 64b and the other branch becomes the decision tree 62c. The decision tree 62c is the third factor, and one branch becomes the branch result 64c and the other branch becomes the decision tree 62d.

For example, the decision tree analysis execution unit 54 outputs the analysis result 70 shown in FIG. 8 when the second measurement is specified as the inspection timing. The analysis result 70 is a branching result including the decision trees 72a, 72b72c, 72d, 72e to be branched, the nonconforming product ratio of the products that satisfy the branching condition of the decision tree, and the number and ratio of the products. 74a, 74b, 74c, 74d, 74e, 74f and the like. In the analysis result 70, the decision tree 72a, which is the first branch, is the fifth factor, one branch becomes the decision tree 72b, and the other branch becomes the decision tree 72c. The decision tree 72b is the third factor, and one branch has a branch result 74a and the other branch has a branch result 74b. The decision tree 72c is the first factor, and one branch becomes the decision tree 72d and the other branch becomes the decision tree 72e. The decision tree 72d is the second factor, and one branch has a branch result 74c and the other branch has a branch result 74d. The decision tree 72e is the fourth factor, and one branch has a branch result 74e and the other branch has a branch result 74f.

The decision tree analysis execution unit 54 evaluates the results of various combinations of the design variables to be analyzed, and outputs the analysis results of any combination. Therefore, the decision tree analysis execution unit 54 combines the decision trees regardless of the order of measurement.

When the decision tree analysis is executed, the decision tree analysis execution unit 54 determines whether or not there is a condition (branch result) in which the proportion of nonconforming products is above a certain level (step S36). When the decision tree analysis execution unit 54 determines that the ratio of nonconforming products does not have a certain condition (branch result) (No in step S36), the decision tree analysis execution unit 54 proceeds to step S39. When the decision tree analysis execution unit 54 determines that the ratio of nonconforming products has a certain condition (branch result) (Yes in step S36), the decision tree analysis execution unit 54 stores the target condition (step S38) and proceeds to step S39.

The decision tree analysis execution unit 54 determines whether the convergence condition is satisfied when it is determined as No in step S36 or when the process of step S38 is executed (step S39). Convergent conditions can be set in various ways. For example, whether the threshold value of the explanatory variable has been changed a set number of times, or whether the number of times that a condition equal to or more than the non-conforming product has been extracted is a certain number or more. When the decision tree analysis execution unit 54 determines that the convergence condition is not satisfied (No in step S39), the decision tree analysis execution unit 54 returns to step S32, changes the threshold value of the explanatory variable, and executes the processes from step S32 to step S39. When the decision tree analysis execution unit 54 determines that the convergence condition is satisfied (Yes in step S39), the decision tree analysis execution unit 54 ends this process.

After executing the decision tree analysis, the production process analyzer 20 displays the analysis result (step S20). For example, when the production process analyzer 20 specifies the first measurement as the inspection timing, the production process analyzer 20 displays the analysis result 60 of the decision tree analysis shown in FIG. When the second measurement is specified as the inspection timing, the production process analyzer 20 displays the analysis result 70 of the decision tree analysis shown in FIG. Further, the image to be displayed as the analysis result is not limited to displaying the decision tree, and the condition that the number of nonconforming products increases may be displayed. Further, the production process analyzer 20 may display a condition for detecting a nonconforming product, that is, a combination of threshold values of explanatory variables in which the possibility of the nonconforming product is equal to or higher than the threshold value as an alert rule.

The production process analyzer 20 can execute the analysis of the explanatory variables based on the order of the processes for producing the product. This makes it possible to analyze how much nonconforming products occur in the final product based on the results of the evaluation target, and eliminate nonconforming products by determining whether the final product is conforming or nonconforming based on the evaluation target results. You can judge the possibility of being able to do it. As a result, it is possible to make an efficient judgment based on the results of each inspection, and it is possible to adjust each condition of the production process to a more appropriate condition.

As described above, the production process analyzer 20 specifies the timing of the evaluation target on the production flow, the timing of measurement in the present embodiment, extracts the measurement results before that point, and determines the result. Perform the analysis. This makes it possible to analyze the correspondence between the factors (measurement results) that can be determined up to the measurement at that time and the probability that a nonconforming product will occur. In other words, if you perform a decision tree analysis for all the factors in the production flow, even if you can grasp the factors that are the explanatory variables, further analysis will be performed to understand when and how to judge in the production flow. Although necessary, the production process analyzer 20 of the present embodiment can obtain analysis results according to the state at each time point of the production flow process. As a result, it is possible to grasp whether or not there is a high possibility of a nonconforming product based on the results up to a specific process in the production flow. As a result, the measurement result in the middle of the production process can be reflected in the production with higher accuracy.

Further, the production process analyzer 20 can extract the threshold value of the explanatory variable capable of detecting the nonconforming product by adjusting the threshold value of the explanatory variable and executing the decision tree analysis for each case. As a result, by setting the threshold value as a criterion for determining whether or not the product is non-conforming at the time of measurement, it is possible to eliminate the non-conforming product during manufacturing with a high probability. In other words, it is possible to find a standard that can eliminate nonconforming products in the middle of manufacturing with a high probability with higher accuracy.

FIG. 9 is a flowchart showing an example of processing of the production process analyzer. The process shown in FIG. 9 is the same as that in FIG. 6 except for a part of the process. The same step numbers are assigned to the same processes as those in FIG. 6, and detailed description thereof will be omitted.

The production process analyzer 20 acquires the production flow data 22, and the specific processing unit 50 identifies the timing of the inspection to be evaluated from the acquired production flow (step S12). After the production process analyzer 20 specifies the timing of the inspection (measurement) of the evaluation target, the data extraction unit 52 uses the data extraction unit 52 to display the inspection result (measurement data) before the inspection timing of the evaluation target and the final result data (finished product). (Measurement data of) is extracted (step S14). After extracting the data, the production process analyzer 20 determines a threshold value that is a boundary between the conforming product and the non-conforming product (step S16).

After the threshold value is determined, the production process analyzer 20 performs a decision tree analysis by the decision tree analysis execution unit 54 (step S18). After executing the decision tree analysis, the production process analyzer 20 determines whether or not the analysis has been executed at all the inspection timings (step S52).

When the production process analyzer 20 determines that the analysis is not executed at all the inspection timings (No in step S52), the production process analyzer 20 returns to step S12 and does not perform the decision tree analysis on the inspection timing of the evaluation target. Set the timing and execute the same process. When the production process analyzer 20 determines that the analysis has been executed at all the inspection timings (Yes in step S52), the production process analyzer 20 displays the analysis result (step S20).

By executing the decision tree analysis at each measurement time point as described above, the production process analyzer 20 can analyze the conditions that can be determined as nonconforming products at each measurement time. This makes it possible to efficiently analyze the production flow.

Next, the processing of the production control system 10 will be described with reference to FIG. FIG. 10 is a flowchart showing an example of processing in the production management system. The production control system 10 acquires the result of the decision tree analysis performed by the production process analyzer 20 by the production flow changing unit 20 (step S62). The production management system 10 determines whether or not there is an improvement target in the production flow changing unit 20 (step S64). When the production management system 10 determines that there is no improvement target (No in step S64), the production control system 10 ends this process. When the production control system 10 determines that there is an improvement target (Yes in step S64), the production control system 10 sets a condition for determining a nonconforming product with respect to the inspection result (step S66). That is, the production flow changing unit 20 changes the condition for determining the nonconforming product at the time of measuring the object based on the condition of the object having the combination of the inspection results to be improved.

As described above, the production control system 10 changes the production flow based on the result of the decision tree analysis performed by the production process analyzer 20, and specifically, the nonconforming product at the inspection stage during the manufacturing process. By setting the criteria for determining that, it is possible to determine with higher accuracy that a product that will eventually become a nonconforming product is produced as a nonconforming product in the middle of production. As a result, the object that finally becomes a nonconforming product can be eliminated in the middle of the process, and the production yield can be increased.

Further, in the present embodiment, the conditions for determining nonconforming products with respect to the inspection result are adjusted, but the conditions for each process of the production flow may be adjusted based on the result of the decision tree analysis. Further, the timing of measurement may be adjusted based on the result of the decision tree analysis. Specifically, based on the measurement results, the measurement with many objects that can be judged as non-conforming products during manufacturing continues, and the measurement with few objects that can be judged as non-conforming products even if measured is continued. , The flow may be changed so that the measurement is not performed.

10 Production control system 12 Production line 14 Measuring device 16 Production control device 20 Production process analyzer 22 Production flow data 24 History data 26 Production flow change section 30 Production process acquisition section 32 History data acquisition section 34 Processing section 36 Storage section 38 Input section 40 Display unit 50 Specific processing unit 52 Data extraction unit 54 Decision tree analysis execution unit 56 Analysis result output unit

Claims (5)

The production process acquisition department that acquires production flow information including the order of production processes and the timing of inspection,
A history data acquisition unit that acquires history data that is information on inspection results of a plurality of products produced based on the production flow, and a history data acquisition unit.
It has a processing unit that analyzes the production process based on the production flow and the inspection result.
The processing unit includes a specific processing unit that identifies a target inspection, which is an inspection to be evaluated, from the production flow.
A data extraction unit that extracts information on the target inspection of the product, the production process executed before the target inspection, the inspection result, and the final measurement result of the product from the historical data.
An allowable range is set for the final measurement result, each product is classified into a conforming product that satisfies the allowable range and a non-conforming product that does not meet the allowable range based on the setting, and the information extracted by the data extraction unit is used. It is characterized by having a decision tree analysis execution unit that executes a decision tree analysis as an explanatory variable and executes a decision tree analysis that calculates the probability of occurrence of nonconforming products for each case classified by each decision tree. Production process analyzer. The production process analyzer according to claim 1, wherein the explanatory variables are classified according to a threshold value set for the inspection result. The production process analysis apparatus according to claim 2, wherein the decision tree analysis execution unit changes the threshold values of the explanatory variables and executes the decision tree analysis for each threshold value. The claim is characterized in that the processing unit further includes an analysis result output unit that outputs the condition when a condition in which the occurrence rate of nonconforming products is equal to or higher than a predetermined probability is detected as a result of execution by the decision tree analysis execution unit. The production process analyzer according to any one of items 1 to 3. The production process analyzer according to any one of claims 1 to 4.
A production management system characterized by having a production flow changing unit that changes a production flow based on an analysis result of the production process analyzer.

Images