Progress forecasting system

Description: TECHNICAL FIELD [Detailed description of the invention]

【0001】

[Technical field to which the invention belongs]

The present invention predicts a production process / production shop in which an in-process lot in a certain process is in the middle of production after a lapse of a predetermined time in a factory that produces a product, and predicts the time required to complete a product in the middle of production. Alternatively, it relates to a prediction technique for easily calculating the index of the neck process using a computer.

【0002】

[Conventional technology]

As a conventional technique for predicting the working time until the product is completed, for example, the technique described in Japanese Patent Application Laid-Open No. 10-111963 is known.

【0003】

With this technology, data consisting of the processes required for product production and standard working hours in each process is prepared in advance, and the total time required for production is calculated from the data, and the current work in process for each process is performed. The completion time is calculated by adding the time required for the work in progress.

【0004】

Further, as a method of predicting a production process and a production shop in which a lot in process during production is considered to be advanced after a lapse of a predetermined time, a technique using a linear graph of a linear line is known. Specifically, the production process y is calculated by expressing the x-axis as the time and the y-axis as the product production process as a linear equation of y = a × x and substituting the predetermined time into the variable x. is there. Here, the coefficient a is obtained by calculating the total number of production processes of the product by the total working time required for the production of the product.

【0005】

Further, as a technique for indexing the neck process, a method is known in which the work load is piled up for each production shop and the results are compared to determine the neck process. The work time of each work required in the process of calculating the work time of a pile of work loads is calculated using the standard work time.

【0006】

[Problems to be Solved by the Invention]

In production shops that focus on manual work, or in production shops where there are many equipment setup changes and sudden failures, there are large variations in the capabilities of workers and equipment utilization rates, and the standard work time and actual work time Is often different. Therefore, in the conventional technique described in JP-A-10-111963 described above, when there are many production processes, errors between the standard working time and the actual working time are accumulated for each process, and the expected completion time (estimated lot progress). ) And the actual completion time of the work (actual lot progress) may be significantly different. In addition, it cannot be used unless standard working hours are prepared in advance.

【0007】

In addition, as a method of predicting the production process and the production shop in which the in-process lot in the middle of production is considered to be advanced after the lapse of a predetermined time, there is a method of using a linear graph of a linear line, which is constant after a certain period of time. This is an effective technique when the number of processes advances, but if there is a large gap in the working time of each production process, it cannot be approximated correctly.

【0008】

Furthermore, according to the conventional technique of indexing the neck process, since the standard working time is used, if the working time deviates from the actual working time, the load pile result does not match the actual situation, and the neck process is properly calculated. I can't.

【0009】

An object of the present invention is to provide a progress prediction system capable of further improving the accuracy of lot progress and the accuracy of neck process indexing.

【0010】

[Means for solving problems]

In order to solve the above problems, the progress prediction system of the present invention is provided with the means (1) to (8) shown below.

【0011】

The production order of the production process of the product and the production shop of the product are registered in advance. (1) The work procedure master file, the production shop where each in-process lot was worked, and the completion time of the work are recorded for each process ( 2) Calculate the average work time for each production shop and the standard deviation showing the variation of the work time from the work history file and the work history file. (3) The production shop capacity calculation means and the production shop capacity calculation unit. Store the calculation result (4) Work time master file and the current process of each in-process lot, store the production shop (5) Extract the current process of each in-process lot from the in-process lot master file and the in-process lot master file Then, the remaining process of the in-process lot and the working time of the remaining process are extracted from the working procedure master file and the working time master file, and the working time when the working time is short from the average working time and the standard deviation of the working time. Outputs the sum of the total work time and the total work time when the work time is long (6) Progress prediction result output means, the variation management limit value of the work time for each production shop, and the variation of the work time of the shop is the control limit value. (7) The abnormality diagnosis master file in which the warning message and the message of the countermeasure procedure are registered in advance when the abnormality exceeds the above, and the variation management limit value registered in the abnormality diagnosis master file are read, and the variation management limit value is set. It is composed of (8) work abnormality detection and diagnosis means that detects the production shop corresponding to the exceeding work time, warns the production shop of the abnormality, and presents the countermeasure procedure.

【0012】

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, an embodiment of the progress prediction system according to the present invention will be described.

【0013】

FIG. 1 shows the configuration of the progress prediction system 1 according to the present embodiment.

【0014】

As shown in the figure, the progress prediction system 1 has a work history file 21 in which work history for each production shop is registered, a work in process lot master file 22 in which the current locations and processes of all work in process are registered, and a work in process lot master file 22 for producing products. Input the work procedure master file 3 in which the work procedure is registered and output the progress prediction information 14 of the work-in-process lot. The work history file 21 and the work-in-process lot master file 22 can be obtained from the progress management system 2.

【0015】

In addition, the progress prediction system 1 includes a production shop capacity calculation means 11 that calculates the capacity (work time) of the production shop from the work history file 21, a work time master file 13 that stores the calculated work time, and a work-in-progress lot master file 22. And work procedure master file 3 and prediction period The progress of the in-process lot is predicted from the master file 16 and output to the progress prediction information 14. The progress prediction result output means 12 and the variation of the work time for each production shop. Abnormality warning and countermeasure procedure to be output when the value is exceeded Read the variation management limit value from the abnormality diagnosis master file 18 registered in advance, detect the production shop that exceeds the variation management limit value, and notify the production shop of the abnormality. It is equipped with a work abnormality detection and diagnosis means 16 that outputs abnormality diagnosis and diagnosis information 17 in which countermeasure procedures are described.

【0016】

In practice, the progress prediction system 1 can be realized as a process on the computer 10 having the network configuration shown in FIG. In addition, the work history file 21 of each production shop 41 is additionally registered and the work-in-process lot master file 22 is updated. The progress management system 2 can be realized as a process on the computer 20. The progress prediction system 1 can acquire the work history file 21 and the work-in-process lot master file 22 via the network 5.

【0017】

Hereinafter, the process of predicting the time required for the progress prediction system 1 to complete the product in the middle of production will be described by taking Examples 2 to 10 as an example. The case where the progress prediction system 1 is implemented as the process on the computer shown in FIG. 2 and the product X is produced will be described as an example. Work procedure 4 in FIG. 2 shows the processes and production shops in the order in which product X is produced. In other words, product X means to work in the order of shop A, shop B, shop C, shop B, and shop C. The contents of this work procedure 4 are registered in the computer 20 in advance as the work procedure master file 3 shown in FIG. Next, the work history file 21 in the process in the middle of production of the product X will be described with reference to FIG. FIG. 4 is a Gantt chart showing the work-in-process lots (x1 to x3) and processes required by the production shop to produce the product X, the actual work start time, and the actual work completion time. For example, work in process lot x1 works in shop A from 1:00 to 2:00, works in shop B from 2:30 to 3:00, and works in shop C from 3:00 to 4:00. It means that it was done. The same content as this is the work history file 21 shown in FIG. 5, which is similarly registered in the computer 20 in advance.

【0018】

The processing procedure of the production shop capacity calculation means 11 is shown in the flowchart of FIG. As shown in the figure, the production shop capacity calculation means 11 first takes in the work history file 21 (step 1). Find the working time for each work-in-process lot and each production shop (step 2). Here, the working time is the difference between the time when the process is completed and the time when the process one step before the process is completed. In the case of the work history file 21 in FIG. 5, the value is shown in the work time 14 in the figure. The unit of working time in this example is [min].

【0019】

Next, based on the work time obtained in step 3, the average work time (X'), the standard deviation of the work time (σ), the short work time (X'-σ), and the long work time (X'+ σ) Is calculated for each production shop, and the result is registered in the working time master 13 (step 3). In the case of this embodiment, the values are as shown in FIG. For example, the working time at shop A is about 50 [min] on average working time, about 17 [min] standard deviation of working time, and about 33 [min] when working time is short, based on the normal distribution. , It means that it is about 67 [min] when the working time is long. In addition, the work time required for shop A is about 13% for less than 33 [min], about 68% for 33 to 67 [min], and about 13 for more than 67 [min]. Means that it is%.

【0020】

The processing procedure of the progress prediction result output means 12 is shown in the flowchart of FIG. As shown in the figure, the progress prediction result output means 12 first takes in the work-in-process lot master file 22 (step 1). In this embodiment, the values shown in FIG. 9 are stored in the work-in-process lot master file 22. This value means that the factory's current work in process is only lot x0 of product X, work from step 1 remains, and the location is shop A.

【0021】

In step 2, the remaining process of the lot taken in in step 1 and the working time thereof are acquired from the working procedure master 3 and the working time master 13.

【0022】

Finally, the time when the work time is short and the time when the work time is long are accumulated in the order of processes, and the result is output to the progress prediction information 14 (step 3). Here, the progress prediction information 14 of the lot x0 when the current time is 1:00 is as shown in FIG. This shows the production shops required for the production of lot x0 and the scheduled completion times for each probability arranged in the order of processes. For example, the scheduled work completion time of shop A in process 1 is before 1:33 in area 62 at the earliest, between 1:33 and 2:05 in area 61 at normal times, and after 2:05 at late times. Means to be. It also means that the probability that the actual work completion time will be area 61 is about 68%, and the probability that it will be area 62 or area 63 is 32%. As a result, the progress of the in-process lot can be predicted in consideration of the variation in the actual working time, so that the lot progress can be predicted more accurately.

【0023】

Examples of the process in which the progress prediction system 1 predicts the location (production shop) and the process of the in-process lot in the middle of production after a lapse of a predetermined time will be described below by taking FIGS. 11 to 14 as an example.

【0024】

In this embodiment, the configuration of the progress prediction system 1 is the same as that of the above-described embodiment, but since the processing of the progress prediction result output means 12 is different, only the processing flow of the progress prediction result output means 12 is used with reference to FIG. explain. The contents of the various files are the same as those in the above-described embodiment.

【0025】

In the progress prediction result output means 12 shown in FIG. 11, first, information on the process and the production shop where the current work-in-process lot is located is taken in from the work-in-process lot master file 22 (step 1). As a result of importing, it can be seen that the current work-in-process lot of the factory is only the lot x0 of the product X as shown in the work-in-process lot master of FIG. 9, and the work from step 1 remains. Next, from the work procedure master file 3, the remaining processes (after the first process) of the work-in-process lot (x0) imported in step 1 and the production shop are extracted, and the work time of the production shop extracted from the work time master is acquired. (Step 2). The correspondence table 71 of FIG. 12 shows the correspondence result of the remaining process acquired in this step, the corresponding shop, and the working time. Thus, for example, the production shop in step 1 is shop A, and its average working time is 50 [min], 33 [min] for short working time, and 67 [min] for long working time. I understand. In step 3, the cumulative values of the average working time, the short working time, and the long working time in the correspondence table 71 acquired in step 2 are obtained. The cumulative value of the obtained average work time in the process order is the cumulative value 72, the cumulative value of the process order of the short work time is the cumulative value 73, and the cumulative value of the process order of the long work time is the cumulative value 74.

【0026】

Next, the elapsed time is fetched from the scheduled period master file 16, and the last process is extracted from the processes in which the cumulative working time values (cumulative value 72, cumulative value 73, cumulative value 74) correspond to the elapsed time or less. (Steps 5 and 6). In this embodiment, the elapsed time is 215 [min] as shown in the scheduled period master 16 in FIG. 13, so the maximum process among the cumulative values of 215 [min] or less is the cumulative value 72 and the cumulative value 73. , Extract from the cumulative value 74, respectively. At the cumulative value of 72, the maximum value is 170 [min] at 215 [min] or less, so the corresponding step is step 3. In other words, if the work is performed in an average working time, it means that the work-in-process lot in step 1 proceeds to step 3 after 215 [min]. For the cumulative value 73 and the cumulative value 74, the corresponding steps are extracted by the same procedure. As a result, the step corresponding to the cumulative value 73 (in the case of a short working time) is the step 5, and the step corresponding to the cumulative value 74 (in the case of a long working time) is the step 2. This result is output to the progress prediction information file 14 (step 6). The progress prediction information file 14 can be represented as FIG. That is, the work-in-process lot x0 in step 1 is 215 [min]. It can be seen that the process to be followed is process 5 when it is early, process 3 when it is average, and process 1 when it is late. As a result, the progress of the process in which the work-in-process lot progresses to the specified period can be known in consideration of the variation in the work, so that it can be expected that the progress can be predicted more accurately.

【0027】

Finally, an example of the process of determining the neck process by the progress prediction system 1 will be described below by taking FIG. 15 as an example.

【0028】

The abnormality diagnosis master file 18 in FIG. 15 is a file in which the control limit value of the work variation range and the warning message and the countermeasure message when the control limit value is exceeded are described for each production shop. In the work abnormality detection diagnosis procedure 16, the calculation result of the sum of the average working time (X') and the standard deviation (σ) of the production shop described in the abnormality diagnosis master file 18 and the abnormality diagnosis master file 18 are described. Compare with the control limit value, determine the production shop that exceeds the control limit value as the neck process, and output the warning / countermeasure message of the abnormality diagnosis master file 18 to the abnormality diagnosis and diagnosis information file 17. In this embodiment, the calculation result of the sum of the average working time (X') and the standard deviation (σ) of shop B in FIG. 7 is 62 [min], whereas the management limit value of shop B in FIG. 15 is 60. Since the control limit value was exceeded, shop B was decided as the neck process. Figure 16 shows the output results of this work abnormality detection and diagnosis procedure 16. This means that the neck process is shop B, and it means that the neck process can be performed by taking measures against equipment failure. In this method, the neck process is determined based on the statistical results of the variation in actual work time, so it is possible to extract production shops where sudden failures occur, and a more accurate neck process indexing method is realized. it can.

【0029】

Finally, FIG. 17 shows an embodiment different from that of the present embodiment. This is an example of a distributed system in which the progress prediction system operates as a process of multiple computers 2. Further, it may be operated as the same process of the computer 200 as the progress management system.

【0030】

[Effect of the invention]

According to the above-mentioned progress prediction system, it is possible to obtain cases such as when the work is completed early and when the work is completed late in consideration of the probability distribution of the work variation, so that more accurate prediction can be performed.

[Simple explanation of drawings]

[Figure 1]

The configuration explanatory view of the progress prediction system which concerns on this embodiment.

[Figure 2]

An operation explanatory diagram of the progress prediction system 1 explaining the case where the product X is produced and implemented as a process on the computer 10 in the network configuration.

[Fig. 3]

Work procedure master file contents explanatory diagram.

[Fig. 4]

Work history file contents explanatory diagram.

[Fig. 5]

Work history file detailed explanatory diagram.

[Fig. 6]

The flowchart which shows the processing procedure of the production shop capacity calculation means.

[Fig. 7]

Work master file content explanatory diagram.

[Fig. 8]

The flowchart which shows the processing procedure of the progress prediction result output means.

[Fig. 9]

Work-in-process lot master file content explanatory diagram.

[Fig. 10]

Diagram of progress prediction information for lot x0 when the current time is 1:00.

[Fig. 11]

Progress prediction result output means content explanatory diagram.

[Fig. 12]

Progress prediction result Output means Detailed explanatory diagram of the contents of the output means (The corresponding results of the remaining processes, corresponding shops, and working hours are explained in this figure).

[Fig. 13]

Scheduled period master file content explanatory diagram.

[Fig. 14]

The figure which shows an example of the progress prediction information file output result.

[Fig. 15]

Abnormality diagnosis master file contents explanatory diagram.

[Fig. 16]

The figure which shows an example of the output result of the abnormality detection diagnosis procedure.

[Fig. 17]

The figure which shows the operation explanatory example as a distributed system as a process of a plurality of electronic computers of a progress prediction system.

[Explanation of symbols]

1 ... Progress prediction system, 10 ... Computer with progress prediction system installed, 11 ... Production shop capacity calculation means, 12 ... Progress prediction result output means, 13 ... Working time master , 14 ... Progress prediction information, 16 ... Scheduled period file, 17 ... Abnormality diagnosis and diagnosis information, 18 ... Abnormality diagnosis master file, 2 ... Progress management system, 20 ... Progress management Computer with system installed, 200 ... Computer with progress prediction system and progress management system installed, 21 ... Work history file, 22 ... In-process lot master, 3 ... Work procedure Master, 4 ... work procedure, 41 ... each production shop, 5 ... network, 61 ... area, 62 ... area, 63 ... area.