Automatic process model generation system and automatic process model generation method

Abstract

To automatically generate a process model in which a flow shop and a job shop are mixed, by automatically determining a production system between processes by using work result data obtained at a manufacturing site.SOLUTION: A process model automatic generation system 210 reads work result data from work result data 120 and extracts a process flow for each product (130). Next, a synthetic flow obtained by adding a plurality of process flows is created, and a production system between processes corresponding to a closed path is changed to a job shop, thereby automatically determining the production system (140). Finally, a process model in which a flow shop and the job shop are mixed is generated (150) and displayed on a display screen 160.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a process model creation system and a method for automatically generating a process model using the same.

Background Art

[0002] In each operation in the manufacturing industry, on-site data is generated and collected upon completion of an operation or occurrence of a predetermined event. On-site workers and data analysts analyze this on-site data to improve work efficiency in each operation. However, when the departments managing each operation are different, it may be difficult to share information between departments. This is presumably due to differences in the systems for business management among different departments. Although analyzing the collected on-site data aims to improve work efficiency within a department, in order to improve work efficiency across an entire series of business processes, cooperation among departments is essential. For this purpose, it is necessary to manage on-site data generated across the entire business process, not only within a department, by associating them with each other and connecting them to data analysis aimed at improving the efficiency of the entire business process.

[0003] Patent Document 1 discloses an information collection display system in which a data generation device that generates on-site data and a storage device that stores the on-site data generated by the data generation device are connected. The information collection display system includes a relevance data storage unit that stores relevance data defining the relevance of each of a plurality of information included in the on-site data, a relevance data search unit that searches for second information related to first information included in the plurality of information based on the relevance data stored in the relevance data storage unit, and a user interface unit that displays the connection relationship of the plurality of information associated by the relevance data. The relevance data search unit searches for second information related to the first information based on the relevance data in response to selection of the first information in the connection relationship of the plurality of information displayed on the user interface unit, and displays the first information and the second information together with the connection relationship of the plurality of information on the user interface unit.

Prior Art Documents

[0004] [Patent Document 1] Japanese Patent Publication No. 2019-153051 [Patent Document 2] Japanese Patent Publication No. 2002-62922 [Patent Document 3] US2013 / 0208315 publication [Overview of the Initiative] [Problems that the invention aims to solve]

[0005] When generating process models from on-site data generated in the manufacturing process, there are two main classifications of production methods: flow shops and job shops. A flow shop is a production method that allows processes to be executed sequentially (continuous flow) and also allows for processes that are not executed (line flow). Flow shops have the advantages of being able to visualize the process sequence and simplifying input data, but they also have the disadvantages that the process sequence becomes more complex and it becomes difficult to add process sequence patterns as the amount of on-site data used to generate the process model increases. A job shop is a production method that allows processes to be executed in any order and also allows for processes that are not executed. Job shops have the advantages of a simpler display screen and easier addition of process sequence patterns, but they also have the disadvantages that the process sequence cannot be visualized and that the input data requires information on the connections between processes, increasing the effort and the possibility of input errors. In this context, a process model refers to a model composed of data that includes a process model identifier, a previous process identifier, a next process identifier, and the production method between processes.

[0006] The information collection system disclosed in Patent Document 1 proposes a relationship data model that shows the structure of relationship data defining the relationships between each of the multiple pieces of information contained in on-site data generated from operations included in the manufacturing process, and a method for managing information of the manufacturing process using relationship data, but it does not mention a method for determining the production method and generating a process model.

[0007] Patent Document 2 discloses a method for selecting a production line system that determines the most efficient production system among flow shops, GT shops, and job shops based on the route and cycle time for each product. However, it does not address the determination of a production system that combines multiple production systems.

[0008] Patent Document 3 discloses a process model generation method that generates the most efficient process model from planning data, including both flow shop and job shop processes. However, it does not describe a method for generating a process model from actual data.

[0009] This invention has been made in view of the above-mentioned problems, and one of its objectives is to automatically generate a process model that combines a flow shop and a job shop using work performance data. Another objective of the present invention is to automatically generate a process model in which flow shops and job shops coexist by creating a process flow as a flow shop from the work process of each product included in the work performance data, and then modifying the parts of the process flow where closed loops exist into job shops. [Means for solving the problem]

[0010] To achieve the above objective, this invention automatically generates process flows for each product from work performance data collected at the manufacturing site. Here, when generating process models for each product, the system automatically generates process models according to the basic rule that parts where closed circuits are formed are determined to be job shops, and the rest are determined to be flow shops. The automatic process model generation system includes a process flow generation unit that generates process flows for each product from work performance data collected at the manufacturing site, a composite flow generation unit that generates a composite flow by adding up multiple process flows, an automatic production method determination unit that determines the production method between each process from the structure of the composite flow, and a composite flow restoration unit that generates a process model consisting of flow shops and job shops based on the determined production method. Furthermore, it includes a work performance data storage unit that stores the collected work performance data and a work performance data selection unit that selects target work performance data from the work performance data to be used as material for generating process models, and the composite flow generation unit generates a composite flow using the work performance data selected from the work performance data. The process flow generation unit, when generating a process flow, performs pre-processing to delete a portion of the flow if there is a closed loop in the flow, so that there is no closed loop in the process flow. The deleted flow is then restored after being synthesized by the composite flow restoration unit. The production method automatic determination unit determines that it is a job shop if there is a closed loop in the synthesized flow.

[0011] According to another feature of the present invention, the process model generated by the synthetic flow restoration unit is displayed on the operating terminal as a process model consisting of a flow shop and a job shop. The displayed process model is receptive to manual modification by the user, and an interface is provided to enable such modification. The functions of each part of the automated process model generation system described above are realized by the processor executing a specific computer program. [Effects of the Invention]

[0012] According to the present invention, by aggregating work performance data acquired from the manufacturing site, the data entry work by workers can be significantly simplified, and an automated process model generation system has been realized that can automatically generate process models combining flow shops and job shops. Furthermore, the conventional problem of inefficiency in constructing process models using only line-only process models or job shop-only process models has been resolved, and a system has been realized that can mix both types of processes and is compatible with various real-world processes. In addition, it is possible to easily add work sequence patterns to the automatically generated process model, and modifications such as changes to job shops can be easily made, resulting in a user-friendly system. [Brief explanation of the drawing]

[0013] [Figure 1] This figure shows an overview of an embodiment of the present invention. [Figure 2] This block diagram shows an example of a network configuration to which the overall system 200 for automatic process model generation according to the first embodiment is applied. [Figure 3] This block diagram shows an example of the functional configuration of the overall system 200 shown in Figure 2. [Figure 4] Figure 2 shows an example of the operation procedure of the overall system 200, from creating a model from work performance data to registering it in the process model storage unit. [Figure 5] Figure 2 shows an example of work performance data stored in the work performance data storage unit. [Figure 6] This figure shows examples of closed-circuit data stored in the closed-circuit storage unit for each product in Figure 2. [Figure 7] Figure 2 shows an example of process model data stored in the process model storage unit. [Figure 8] This diagram shows examples of a flow shop and a job shop. [Figure 9] Figure 2 is a flowchart showing the procedure by which the automated process model generation system 210 generates a process model consisting of a flow shop and a job shop. [Figure 10] A diagram that divides the procedure for generating a process model composed of a flow shop and a job shop in FIG. 9 into three steps. [Figure 11] A diagram showing a specific example of automatically discriminating the process model of the process model automatic generation system 210 in FIG. 2 (Part 1). [Figure 12] A diagram showing a specific example of automatically discriminating the process model of the process model automatic generation system 210 in FIG. 2 (Part 2). [Figure 13] A diagram showing a specific example of automatically discriminating the process model when there is a closed loop within one product of the process model automatic generation system 210 in FIG. 2 (Part 1). [Figure 14] A diagram showing a specific example of automatically discriminating the process model when there is a closed loop within one product of the process model automatic generation system 210 in FIG. 2 (Part 2). [Figure 15] A diagram showing an example of a screen for displaying a process model composed of a flow shop and a job shop of the process model automatic generation system 210 in FIG. 2. [Figure 16] In the process model automatic generation system according to the second embodiment, it is a block diagram showing a functional configuration example of the process model automatic generation system equipped with a production method abnormality detection function. [Figure 17] In the process model automatic generation system according to the third embodiment, it is a diagram showing examples of a flow shop and a job shop. [Figure 18] It is a block diagram showing a functional configuration example of the process model automatic generation system according to the third embodiment. [Figure 19] A diagram showing an example of a user interface that provides a function of filtering a process model during the period of the process model automatic generation system according to the fourth embodiment.

Mode for Carrying Out the Invention

Examples

[0014] Hereinafter, embodiments of the present invention will be described with reference to the drawings. Note that the embodiments described below are not intended to limit the scope of the claims, and not all of the elements and combinations described in the embodiments are necessarily essential to the solution of the invention.

[0015] Figure 1 shows an overview of one embodiment of the present invention. The computer 110 extracts a process flow from work performance data 120 accumulated at a manufacturing site or the like (130). Next, it automatically determines the production method between each process from the structure of the extracted process flow (140) and generates a process model consisting of a flow shop and a job shop (150). It also has an interface that distinguishes and displays flow shops and job shops by drawing the job shop as a single node (160). Steps 130 to 160, which are performed using the computer 110, are implemented in software by executing a program.

[0016] [Block diagram] Figure 2 is a block diagram showing an example configuration of the overall system 200 to which the process model automatic generation system 210 is applied in the first embodiment. Here, the process model automatic generation system 210 corresponds to the core part of the computer 110 shown in Figure 1. In Figure 2, the process model automatic generation system 210 is connected to the network 220. The network 220 is also connected to one or more work performance data generators 230, one or more work performance data storage units 240, a closed-loop storage unit 250 for each product, and a process model storage unit 260. These form the overall system 200. The process model automatic generation system 210, the work performance data generators 230, the work performance data storage units 240, the closed-loop storage units 250 for each product, and the process model storage unit 260 are all connected to the information collection system 280 via the network 220. In this specification, the term "automatic process model generation system," which is the title of the invention, basically refers to part 210, but the entire system 200 may also be referred to as the "automatic process model generation system" in a broader sense.

[0017] The process model automatic generation system 210 is connected to an operation terminal 270 for operating the process model automatic generation system 210. The operation terminal 270 is used by various users such as administrators who manage the process model automatic generation system 210, monitors who monitor it, and users who utilize it, and consists of one or more information terminal devices. In the example in Figure 2, the overall system 200 and another point Inside the line Although the diagram shows the operation terminal 270 as a separate configuration from the overall system 200, part or all of the operation terminal 270 may be provided within the overall system 200. Alternatively, the process model automatic generation system 210 and the operation terminal 270 may be connected via the network 220, allowing access to the process model automatic generation system 210 from any point outside the overall system 200.

[0018] The work performance data generator 230 is a device that monitors work in the manufacturing site and generates output indicating the progress status and progress results. The work performance data generator 230 may be, for example, a barcode reader, PC (Personal Computer), or server that acquires worker work logs, or a machine that processes parts or assembles finished products, or a sensor that collects inspection information of RFID (Radio Frequency Identifier) ​​attached to parts or finished products. Numerous such work performance data generators are installed in the manufacturing site, and the collected performance data is stored in a system (not shown) that manages the manufacturing site. Furthermore, the work performance data collected or generated by the work performance data generator 230 is transmitted to the work performance data storage unit 240 via the network 220 and stored there.

[0019] The work performance data storage unit 240 is, for example, a storage device such as a server or memory, and stores work performance data received from the work performance data generator 230 via the network 220. The closed-loop storage unit 250 for each product is, for example, a storage device such as a server or memory, and stores closed-loop data of the process model for each product.

[0020] The process model storage unit 260 is, for example, a storage device such as a server or memory, and stores the process model data generated by the process model automatic generation system 210. The information collection system 280 collects information related to the execution of the process, such as process execution records.

[0021] [System Configuration] Next, the main functions of the automatic process model generation system 210 will be described. The automatic process model generation system 210 includes a processor 211 such as a central processing unit (CPU) that controls the entire system, a storage device 212 that stores various processing programs for realizing the functions of the automatic process model generation system 210, a network interface (I / F) 213, and the like. The storage device 212 is implemented using known storage devices such as ROM (Read Only Memory) for storing various processing programs, Random Access Memory (RAM) for temporarily storing information, and Hard Disk Drives (HDD). The functions of the present invention, described below, are realized when the processor 211 executes the various processing programs stored in the storage device 212. Note that the configuration of the automatic process model generation system 210 is not limited to the example shown, and some or all of the programs may be introduced from other devices via non-temporary storage media or communication lines.

[0022] Figure 3 is a block diagram illustrating the functions of the automatic process model generation system 210. The automatic process model generation system 210 is implemented by a computer program and includes the functions of a work performance data selection unit 310, a process flow generation unit 315 for each product, a composite flow generation unit 320, an automatic production method determination unit 325, a composite flow restoration unit 330, a process model display unit 335, and a process model modification unit 340. As mentioned above, the automatic process model generation system 210 is connected via a network 220 to a work performance data generator 230, a work performance data storage unit 240, a closed-circuit storage unit 250 for each product, and a process model storage unit 260.

[0023] The process model automatic generation system 210 is connected to the operation terminal 270. Here, the flow of information through the connection path is indicated by the direction of the arrows. The operation terminal 270 provides the user with the following functions: a work performance data selection unit 310 for selecting work performance data to be used for process model creation; a process model display unit 335 for displaying the generated process model; and a process model modification unit 340 for modifying the process model.

[0024] First, the work performance data selection unit 310 of the process model automatic generation system 210 acquires work performance data for generating a process model specified by the user using the operation terminal 270, and selects the work performance data to be used from the work performance data registered in the work performance data storage unit 240. The process flow generation unit 315 for each product generates the process flow for each product by reading the work performance data selected by the work performance data selection unit 310 from the work performance data storage unit 240. At this time, the production method of the process flow is always set to flow shop when the process flow is generated. In addition, if there are closed circuits in the process flow for each product, the product identifier and closed circuit data are stored in the closed circuit storage unit 250 for each product, and preprocessing is performed so that there are no closed circuits in the process flow for each product.

[0025] The composite flow generation unit 320 combines the process flows for each product generated by the process flow generation unit 315 for each product to generate a composite flow. Next, the production method automatic determination unit 325 determines from the structure of the composite flow generated by the composite flow generation unit 320 whether the production method between each process corresponds to a flow shop or a job shop.

[0026] The composite flow restoration unit 330 restores the process flow pre-processed by the process flow generation unit 315 for each product as a flow shop, based on the process model generated by the automatic production method determination unit 325. It is preferable to perform this restoration using the data stored in the closed-loop storage unit 250 for each product. The process model display unit 335 transmits the process model, consisting of the flow shop and job shop restored by the composite flow restoration unit 330, to the operation terminal 270. The operation terminal 270 displays the display screen 1300 (described later in Figure 15), etc.

[0027] The process model modification unit 340 provides a function to manually modify the process model after the user has confirmed the process model generated by the process model display unit 335, if it differs from the intended process model. Once the manual modification of the process model is complete, the process model data is sent to the process model storage unit 260.

[0028] [Operating Procedure] Figure 4 shows an example of the operation procedure from when the process model automatic generation system 210 automatically generates a process model consisting of a flow shop and a job shop, until it is registered in the process model storage unit 260. First, the user using the process model automatic generation system 210 issues an instruction to the process model automatic generation system 210 to automatically generate a process model using the operation terminal 270 (steps 415, 420). Next, the process model automatic generation system 210 selects the work performance data to be used and sends a command to read the corresponding work performance data to the work performance data storage unit 420 (step 426), and the process model automatic generation system 210 acquires the corresponding work performance data (step 427).

[0029] Next, the automatic process model generation system 210 performs automatic process model generation processing using the selected work performance data (step 430). Here, the automatic process model generation 430 represents the combined function of the process flow generation unit 315 for each product, the composite flow generation unit 320, the automatic production method determination unit 325, the composite flow restoration unit 330, and the process model display unit 335 shown in Figure 3. The automatic process model generation 430 transmits the closed loops for each product to the closed loop storage unit 435 for each product at the process flow generation unit 315 for each product (step 431). The automatic process model generation 430 acquires the closed loop data from the closed loop storage unit 435 for each product at the composite flow restoration unit 330 (see Figure 3) (step 432) and performs composite flow restoration processing. The process model generated by the automatic process model generation 430 is output to the operation terminal 270 (step 435).

[0030] A user utilizing the automatic process model generation system 210 edits the process model displayed on the operation terminal 270, i.e., the result generated by the automatic process model generation 430 (step 440). The process model edited by the user is sent to the automatic process model generation system 210 (step 436) and registered in the process model storage unit 260 (steps 450, 460).

[0031] [Data Contents] Figure 5 shows an example of work performance data 500 registered in the work performance data storage unit 240. The work performance data 500 includes a product identifier 510 that uniquely identifies a product, a process identifier 520 that uniquely identifies a process, and a work date and time 530 that indicates the date and time the process was performed. In Figure 5, the product identifier 510, in the example of an automobile manufacturer, means a specific model (car name), and an identifier is assigned to identify the model (car name) and the individual production unit. In this example, product 1 indicates the first production of that model, product 2 indicates the second production of that model, and product 3 indicates the third production of that model. The process identifier 520 is an identifier (process name, identification number, etc.) for identifying the process of the production process. The work date and time 530 stores the date and time the work was performed, but it may contain multiple pieces of information, such as a work start date and time and a work end date and time.

[0032] Figure 6 shows the closed-loop data 600 for each product registered in the closed-loop storage unit 250 for each product. The closed-loop data 600 for each product includes a product identifier 610 that uniquely identifies the product and a process identifier 620 that uniquely identifies the process.

[0033] Figure 7 shows the process model data 700 registered in the process model storage unit 260. The process model data 700 includes a process model identifier 710 that uniquely identifies the process model, a previous process identifier 720 that uniquely identifies the previous process, a next process identifier 730 that uniquely identifies the next process, and a production method 740 between processes that indicates the production method between the previous and next processes.

[0034] [Flow shops and job shops] Figure 8 shows examples of a flow shop and a job shop in this embodiment. A characteristic of a flow shop is that processes are executed sequentially, and it is permissible for some processes not to be executed. In this embodiment, a production method that takes into account both continuous flow 810 and line flow 820 is defined as a flow shop. In the example in Figure 8, possible process sequences for the flow shop include 1->2->3->4->5 and 1->3->4->5. A characteristic of a job shop is that processes are executed in any order, and it is permissible for some processes not to be executed. In the example in Figure 8, possible process sequences for the job shop 830 include 1->2->4->5->3 and 3->2->5->4->1.

[0035] [flowchart] Figures 9 and 10 are flowcharts showing the procedure for automatic process model generation by the process model automatic generation system 210. The procedure shown in these flowcharts is realized by the computer constituting the process model automatic generation system 210 executing a computer program. Here, the process flow for each product is generated as a flow shop, and then the production method between processes included in the closed loops within the composite flow is changed to a job shop to generate the process model. Alternatively, the process flow for each product may be generated entirely as a job shop, and then the production method between processes other than the closed loops within the composite flow is changed to a flow shop to generate the process model. The procedure shown in Figures 9 and 10 corresponds to the series of processes from the work performance data selection unit 310 to the process model display unit 335 of the process model automatic generation system 210 shown in Figure 3.

[0036] The flowchart is broadly composed of three step groups 1010, 1020, and 1030. Step group 1010 is the process applied to the process flow of each product, and is the part that performs preprocessing to prevent closed circuits within the process flow of a single product from being identified as job shops. Step group 1020, shown in Figure 10, is the process applied to the composite flow obtained by adding up the process flows of all products, and is the part that automatically determines whether the production method between each process is a flow shop or a job shop from the path structure of the composite flow obtained by adding up all the process flows of each product. Step group 1030 is the process applied to the composite flow obtained by adding up all the process flows of each product, and is the part that restores the paths that were removed in the preprocessing of step group 1010 as flow shops.

[0037] In Figure 9, the process model automatic generation system 210 first inputs the work performance data selected by the work performance data selection unit 310 (step 905). Next, the product-specific process flow generation unit 315 uses the work performance data selected by the work performance data selection unit 310 to generate all product-specific process flows as flow charts (step 910).

[0038] Next, the process flow generation unit 315 performs cycle detection in the process flow for each product (step 915). For cycle detection, for example, topological sorting can be used. Topological sorting is a known method in graph theory that orders each node of a directed acyclic graph so that every node comes before the node that is the end of its output edge. Topological sorting can efficiently perform process flow sorting and cycle detection, so it is suitable for performing cycle detection in the process flow for each product. Next, it is determined whether or not a cycle exists in the process flow for each product (step 920). If no cycle exists, the process proceeds to step 940. If a cycle exists, a cycle search is performed (step 925), and after storing the cycle data in the cycle storage unit 250 for each product, a part of the cycle is deleted so that the process flow for each product becomes an acyclic path (step 930). In this embodiment, the last edge that constitutes the cycle is deleted. This process is a pre-processing step to prevent a cyclical path within a single product from being identified as a job shop. Steps 925 to 935 are repeated until the process flow for each product becomes an acyclic graph.

[0039] Next, the synthesis flow generation unit 320 generates a synthesis flow by adding up the process flows for each product (step 940). In this way, the processes from steps 910 to 940 are repeated for product 1, product 2, etc., thereby generating a synthesis flow for each product. Note that in Figure 9, the upper corner of the rectangular box shown in step 910 is cut diagonally, and the lower corner of the box in step 940 is cut diagonally, which means that steps 910 to 940 are repeated multiple times.

[0040] Moving on to Figure 10, the automatic production method determination unit 325 performs cycle detection in the composite flow using topological sorting (step 945). Topological sorting is not necessarily required for cycle detection; other methods for sorting directed acyclic graphs may also be used. Next, it is determined whether or not a cycle has been detected based on the results of the topological sorting (step 950). If no cycles exist in the composite flow, the process proceeds to step 970. If cycles exist in the composite flow, a cycle search is performed (step 955). Then, the production method between processes included in the cycle is changed from a flow shop to a job shop (step 960). This process in steps 955 and 960 is repeated until the search for all cycles is complete (step 965).

[0041] Next, the synthesized flow restoration unit 330 determines whether all the paths that existed in the process flow of each product are present in the synthesized flow (step 970). If the paths that existed in the process flow of each product are present in the synthesized flow, the process ends as is. If they are not present, the unit restores the corresponding edges, i.e., the missing routes, as flow shops (step 975), and then terminates the process.

[0042] [Automatic detection of production method] Figures 11 and 12 illustrate specific examples of automatic process model identification. First, the automatic process model generation system 210 reads the work performance data 1110 of the products for which the process model is to be created from the work performance data storage unit 240 (see Figure 2) and generates a process flow for each product (1120) as shown in Figure 11(b). The work performance data 1110 shown in Figure 11(a) contains the processes of the first product (product 1), the second product (product 2), and the third product (product 3) of the same product.

[0043] In the generation of process flows for each product (step 1120), process flows for product 1, product 2, and product 3 are created according to the work performance data 1110. As can be seen in step 1120, products 1 through 3 can all be displayed as flow shops, but their execution order is different.

[0044] Next, by adding up the process flows of products 1 to 3, these combined flows are generated as shown in Figure 11(c) (step 1130). By generating step 1130, it is possible to confirm the rules of execution order, such as process 2 always being executed after process 1, process 3 always being executed after process 2, process 2 sometimes being executed after process 4, process 5 sometimes being executed, ... (and so on).

[0045] Next, as shown in Figure 12(d), the system automatically identifies closed loops in the composite flow as job shops and the rest as flow shops (step 1140). Here, closed loops are detected by performing a topological sort. Figure 12(d) shows the composite flow created in step 1130 as a linear flow. By rewriting the composite flow in this way, it can be detected that the part indicated by the dotted line, i.e., process 5 → process 6 → process 3 → process 5..., is a closed loop. The process model automatic generation system 210 then transforms the process model as shown in Figure 12(e), assuming that processes 5, 6, and 3, which are included in this closed loop, are tasks performed in job shops (step 1150). Following the procedures assisted in steps 1120 to 1150, the system automatically generates a process model combining flow shops and job shops as shown in step 1150.

[0046] Next, we will explain a specific example of automatically determining the process model for another product using Figures 13 and 14. This example demonstrates the automatic determination of the production method in a process model when a closed loop exists within a single product. When a closed loop exists within a single product, such a path is determined to be a rework / redo operation, and a process model is generated as a flow shop.

[0047] First, the process model automatic generation system 210 reads the work performance data 1210 for the product for which the process model is to be created from the work performance data storage unit 240 (see Figure 2) and generates a process flow for each product (step 1220). The work performance data 1210 shown in Figure 13(a) contains the processes for the first product (product 1) and the second product (product 2) of the same product.

[0048] In the generation of process flows for each product (step 1220) shown in Figure 13(b), the process flows for product 1 and product 2 are created according to the work performance data 1210. As can be seen from these flows, product 1 and product 2 can be displayed as process models. Here, nodes are shown as circles and the connections between each node are shown as edges indicated by arrows, but if not shown, calculations may be performed on the data in a computer-processed state. As can be seen in Figure 13(b), the execution order of product 1 and product 2 is significantly different.

[0049] Next, as shown in Figure 13(c), if a closed loop exists in the process flow for product 1 and product 2, the last edge of the path constituting the closed loop is deleted so that the process flow becomes a non-cyclic path (step 1230). In addition, the closed loop data is stored in the closed loop storage unit 250 for each product. Next, Figure 1 4 As shown in (d), a composite flow is generated by summing up all the process flows from which the closed loops in the process flow for each product have been removed (1240).

[0050] Next, the edges deleted in step 1230 are restored as flow shops (1250). In this case, if all product-specific process flows exist in the composite flow, the restoration process does not need to be performed. In the example shown in Figure 14(d), although the edge from process 6 to process 3 existed in the process flow of product 1, the edge from process 6 to process 3 does not exist in the composite flow. Only in such cases does Figure 14( e As shown in the diagram, the edges are restored as a flow shop. By performing the above process, it is possible to prevent flow shop operations, such as rework, from being identified as job shops.

[0051] As illustrated in Figures 11 to 14 above, the production method between each task is automatically determined from the work performance data. Then, a business flow is generated for each product from the work performance data, and the production method between each task is determined as either a "line" or a "job shop" from the configuration of the model formed by summing up each business flow. Here, the routes that are visited are detected from the model formed by summing up each business flow. Routes that are visited are designated as job shops, and routes that are not visited are designated as lines. If a cyclical path exists within a single product, it is classified as a "rework" and designated as a "line" rather than a "job shop." Based on this classification, a model can be generated using both "lines" and "job shops." The process model created in this way is then rendered in a way that allows for the identification of "lines" and "job shops," and displayed to the user.

[0052] [Display screen] Figure 15 shows an example of a screen displaying a process model consisting of a flow shop and a job shop. By operating the display screen 1300, the user can obtain the functions to display and edit the process model. On the process model display screen 1300, the database (DB) list button 1330 and the table list button 1340 are displayed first. Here, the table list button 1340 is pressed, and the process model list ("Model 1", "Model 2", "Model 3", etc.) 1310 contained in the selected table (not shown) is displayed in icon format. The user who creates the process model selects the target model from the process model list 1310 and clicks the display screen icon 1320, and the automatically created process model is displayed in the display screen window 1360 below (or on another screen). If the user clicks the edit screen icon 1350 displayed to the right of the display screen icon 1320, the display screen window 1360 switches to an editing screen (not shown). In the editing screen, the display window 1360 can show a screen equivalent to the automatically created process model, or one or all of the screens shown in Figures 13(b), (c), 14(d), and (e). The operator then edits the production method of the selected process model. Editing methods include dragging and dropping process nodes to move them, clicking on edges to change the production method between processes, etc. Once the operator has finished modifying the process model, they click the registration icon 1370. The registered process model is then stored in the process model storage unit 260. [Examples]

[0053] Next, the automatic process model generation procedure according to the second embodiment of the present invention will be explained using Figure 16. Figure 16 shows a system configuration that adds a production method master data storage unit 1410 and a production method anomaly detection unit 1420 to the system configuration of the process model automatic generation system 210 in Figure 3. The system 210 stores master data in which the production methods between each process are predetermined in the production method master data storage unit 1410. The production method master data storage unit 1410 is, for example, a storage device such as a server or memory. The same reference numerals are used for the same parts as in Figure 3 in the other configurations and each step showing the processing procedure, so repeated explanations will be omitted.

[0054] The production method anomaly detection unit 1420 detects anomalies in the production method by comparing the production method between each process in the master data stored in the production method master data storage unit 1410 and the process model output via the process model modification unit 340. If an anomaly is detected, the unit may be configured to notify the operator of the anomaly via the operation terminal 270. Furthermore, the production method anomaly detection unit 1420 can be used not only for detecting anomalies in the production method but also for planning purposes, such as formulating more efficient production methods. For example, by displaying a proposed modification of the process model created using the master data on the operation terminal 270, it is possible to assist the user in making a decision on whether to make the modification. [Examples]

[0055] Figure 17 shows an example in which the process model automatic generation system 210 according to the third embodiment adds a batch flow 1510 as a type of flow shop. In the third embodiment, the continuous flow 810, the line flow 820, and the batch flow 1510 are defined as flow shops, and in order to appropriately distinguish between these flow shops and job shops, a product passage count storage unit 1610 is added in addition to the system configuration in Figure 3. A batch flow is a production method in which there is a dominant route pattern through which the majority of products pass, as well as a subordinate route pattern through which a small number of products pass.

[0056] In Figure 17, the fixed route pattern 810 is a production method in which processes are executed sequentially, for example, from process 1 to process 5, and is also called a continuous flow. The line flow 820 is a fixed route pattern, but steps such as moving from process 1 to process 3 are also possible. The batch flow 1510 is a production method in which processes are executed sequentially from process 1 to process 5, but with different flows in batches, for example, a flow that returns from process 3 to process 1, or a flow that moves from process 1 to process 3. Here, the numbers above the arrows in the batch flow 1510 (for example, "100" or "5") indicate the number of times the corresponding arrow sequence was executed based on the aggregated work performance data. In other words, the number of times the process moved from process 1 to process 2 to process 3 to process 4 to process 5 is 100, while the number of times it returned from process 3 to process 1 is 5, and the number of times it moved from process 1 to process 3 is 5. The flow that returns from process 3 to process 1 and the flow that moves from process 1 to process 3 are recognized as batch processing. In this way, in the second embodiment, continuous flows, line flows, and batch flows are determined to be flow shops.

[0057] Figure 18 is a block diagram showing a functional configuration example of the process model automatic generation system 210 according to the third embodiment when it corresponds to the production method shown in Figure 17. The product passage count storage unit 1610 is, for example, a storage device such as a server or memory, and stores the number of products passing between each process. The product passage count storage unit 1610 stores the number of products passing between each process in the synthetic flow in the synthetic flow generation unit 320. The process model modification unit 340 refers to the number of products passing that has been stored in the product passage count storage unit 1610 and changes the production method from a job shop to a flow shop for routes that have been determined to correspond to a batch flow. The threshold for the number of products passing that determines the dominant route and the inferior route may be automatically determined from the data, or it may be made selectable by the user.

[0058] In the third embodiment, a product passage count storage unit 1610 is provided, and the batch flow 1510 is also configured to be automatically created, making it possible to automatically generate a variety of process models. [Examples]

[0059] Next, a fourth embodiment of the present invention will be described using Figure 19. The process model automatic generation system 210 according to the fourth embodiment provides a user interface equipped with a function to filter work performance data for a specific period. Figure 19 is a diagram showing an example of a screen that displays a process model consisting of a flow shop and a job shop in the fourth embodiment. The display screen 1900 of this embodiment provides a user interface that adds a work start time 1970, a work end time 1980, and a filtering button 1990 to the display screen of Figure 15. The user selects the target model from the process model list 1910, and after selecting the process model, clicks the display screen icon 1320 to display the process model in the display screen window 1960. Here, the worker limits the target period of work performance data for creating or displaying the process model by entering the work start time 1970 and the work end time 1980, and then clicks the filtering button 1990, which switches the display of the process model in the display screen window 1960 to a process model regenerated only from work performance data from the work start time 1970 to the work end time 1980. In this way, by specifying a work start time of 1970 and a work end time of 1980, the process model can be filtered by period, and a new model can be generated using only the work performance data performed within the specified period. [Explanation of Symbols]

[0060] 200 Overall System 210 Model Automatic Generation System 230 Work Performance Data Generator 240 Work Performance Data Storage Unit 250 Closed-circuit storage unit for each product 260 Process Model Storage Unit 270 Operating terminals 310 Work Performance Data Selection Department 315 Process flow generation unit for each product 320 Synthesis Flow Generation Unit 325 Automatic production method determination unit 330 Synthetic Flow Restoration Unit 335 Process Model Display Unit 340 Process Model Modification Section 500 work performance data 600 cycle data 810 continuous flow 820 Lineflow 830 Job Shop 1300, 1900 Process Model Display Screen

Claims

1. A work performance data storage unit that stores work performance data collected at the manufacturing site, A work performance data selection unit selects from the work performance data stored in the aforementioned work performance data storage unit to be used as material for generating a process model, A process flow generation unit that generates process flows for each product, A composite flow generation unit generates a composite flow by adding together a plurality of process flows selected by the aforementioned work performance data selection unit, An automatic production method determination unit that determines the production method between each process from the structure of the aforementioned synthesis flow, It includes a synthetic flow restoration unit that generates a process model consisting of a flow shop and a job shop based on the determined production method, When generating the process flow, if there is a closed loop in the flow, the process flow generation unit performs preprocessing to delete a part of the flow so that there are no closed loops in the process flow. The production method automatic determination unit generates the process model by determining that a job shop exists when a closed loop exists in the synthesis flow. The process model automatic generation system is characterized in that the composite flow restoration unit completes the process model by combining a composite flow that restores part or all of the composite flow that was deleted by the preprocessing.

2. The process model automatic generation system according to Claim 1, characterized in that the work performance data includes a product identifier that identifies a product, a process identifier that identifies a process, and the work date and time on which the process was executed.

3. A process model display unit that displays the process model generated by the aforementioned synthetic flow restoration unit as a process model composed of a flow shop and a job shop, The process model automatic generation system according to claim 1 or 2, further comprising a process model modification unit that provides an interface enabling the user to modify the displayed process model.

4. A master data storage unit that stores master data in which the production methods between each process are predetermined, The process model automatic generation system according to claim 3, further comprising: an anomaly detection unit that, when master data corresponding to the master data storage unit exists, compares the difference between the process model generated by the process model automatic generation system and information in which the production methods between each process are predetermined, thereby performing at least one of anomaly detection and production method planning.

5. A method for automatically generating a process model, which is realized by having a computer having a processor and a work performance data storage unit that stores work performance data collected at a manufacturing site, Using the work performance data stored in the aforementioned work performance data storage unit, a process flow is generated for each product. A composite flow is generated by adding up multiple process flows. The production method between each process is automatically determined from the structure of the created synthesis flow. A method for automatically generating a process model, characterized by automatically generating a process model in which flow shops and job shops coexist by replacing a flow shop with a job shop when a flow shop is included in the closed loop of the composite flow based on the results of the automatic determination.

6. The process model automatic generation method according to claim 5, characterized in that in the step of generating the process flow, the process flow is generated using all or part of the work performance data of the product stored in the work performance data storage unit.

7. In the step of generating the process flow for each product, if a circular path exists within a single product, the last edge is removed. The process model automatic generation method according to claim 6, characterized in that, after a process model in which the flow shop and the job shop are mixed is automatically generated, the removed edges are restored.

8. The system includes a display means for displaying the generated process model and an input means for modifying the displayed process model. When displaying the process model as a mixture of the flow shop and the job shop, the job shop is drawn as a single node. The process model automatic generation method according to claim 7, characterized in that it provides an interface that allows a user to arbitrarily modify a process model in which the production method has been automatically determined.

9. The process model automatic generation method according to claim 8, characterized in that, if master data exists which defines the production methods between each process, the process model automatic generation system compares the difference between the process model it generates and the information which defines the production methods between each process in advance, thereby performing at least one of anomaly detection and production method planning.

10. Continuous flow, line flow, and batch flow are identified as flow shops. The process model automatic generation method according to claim 9, characterized in that it automatically generates the process model from work performance data that occurred within a period specified by the user.

Images