Data creating device

Abstract

This data creating device comprises: an input unit for inputting scheduled board information relating to production-scheduled boards; and a calculating unit that, on the basis of the input scheduled board information, creates configuration data for a scheduled mounting line for producing the production-scheduled boards. The configuration data includes, at least, component mounting machine information and, for each component mounting machine, component holding unit information relating to a component holding unit that holds components to be disposed inside the component mounting machine. The calculating unit creates the configuration data for the scheduled mounting line using a learning device generated by machine learning using, as teacher data, actual component information relating to the type and number of components mounted on actual component mounting boards produced in the past, actual configuration data indicating the configuration of the mounting line on which the actual component mounting boards were produced, and actual production time information relating to the actual production time required to produce the actual component mounting boards. This improves productivity, thereby enabling a contribution to the realization of what is known as a smart factory.

Description

Data creation device

[0001] The technology disclosed in this specification relates to a data creation device. In particular, it relates to a data creation device that creates configuration data indicating the configuration of a mounting line for producing a component mounting substrate.

[0002] In facilities such as component mounters installed on the mounting line, for example, various parameters are set according to the material of the components and the like. The line facility setting support device of Patent Document 1 stores the configuration of lines with production records. When information regarding the configuration of a line scheduled to newly produce components is input, for example, the line facility setting support device outputs the parameters used on a line with a production record having a configuration similar to the configuration indicated by the input information.

[0003] Japanese Patent Application Laid-Open No. 2018-18416

[0004] The configuration of the mounting line is determined by various conditions, for example, the type and number of component mounters installed on the mounting line. Conventionally, after an operator sets each condition based on their own experience and the like, calculates the production time based on the condition, and repeats the adjustment of each condition until the calculated production time is below the allowable production time, the configuration of the mounting line (that is, the configuration data of the mounting line) is determined. In this specification, a technology that can shorten the working time until the configuration data of the mounting line is determined is provided.

[0005] The data creation device disclosed herein creates configuration data indicating the configuration of a mounting line for producing component-mounted substrates. The data creation device includes an input unit for inputting planned substrate information relating to substrates to be produced, and a calculation unit for creating configuration data for a planned mounting line for producing the substrates to be produced based on the input planned substrate information. The planned substrate information includes at least planned component information relating to the types and number of components to be mounted on the substrates to be produced, and planned production time information relating to the production time allowed for the production of the substrates to be produced. The configuration data includes at least component mounting machine information relating to the types and number of component mounting machines equipped on the mounting line, and component holding unit information relating to the types and number of component holding units that hold the components arranged in the component mounting machine for each of the component mounting machines equipped on the mounting line. The calculation unit creates the configuration data for the planned assembly line using a learner generated by machine learning, with the following as training data: actual component information relating to the types and number of components mounted on past actual component mounting boards; actual configuration data indicating the configuration of the assembly line that produced the actual component mounting boards; and actual production time information relating to the actual production time required to produce the actual component mounting boards.

[0006] (Effects of the Invention) The data creation device described above creates configuration data for a planned assembly line using a learning model generated by machine learning with actual parts information, actual configuration data, and actual production time information as training data. Therefore, compared to conventional technology that creates configuration data based on the experience of workers, it is possible to create appropriate configuration data for a planned assembly line without relying on the experience of workers. As a result, the time required to determine the configuration data for a planned assembly line can be shortened. This improves productivity and can contribute to the realization of a so-called smart factory.

[0007] Configuration diagram of the data creation device in the embodiment. Flowchart of the data creation process.

[0008] The main features of the embodiments described below are listed below. Note that the technical elements described below are independent technical elements that exhibit technical usefulness individually or in various combinations, and are not limited to the combinations described in the claims at the time of filing.

[0009] (Feature 1) In the data creation device described above, the planned component information may include at least one of the following: information indicating the name of the component to be mounted on the planned production board, an image of the component to be mounted on the planned production board, and information indicating the size of the component to be mounted on the planned production board.

[0010] With this configuration, the size of the component to be mounted on the production board can be determined based on at least one of the following pieces of information: information indicating the name of the component, an image of the component, and information indicating the size of the component.

[0011] (Feature 2) In the data creation device described above, the types of component mounting machines equipped on the mounting line may include a first type indicating a first component mounting machine capable of mounting a first number of components per unit time, and a second type indicating a second component mounting machine capable of mounting a second number of components greater than the first number per unit time.

[0012] With this configuration, the types of components that should be installed on the planned mounting line can be reflected in the configuration data, given that the first and second component mounting machines have different numbers of components that can be mounted per unit time.

[0013] (Feature 3) In the data creation device described above, the types of component mounting machines equipped on the mounting line may include a third type, which indicates a third component mounting machine that receives the supply of components by a feeder, and a fourth type, which indicates a fourth component mounting machine that receives the supply of components by a tray.

[0014] With this configuration, it is possible to reflect in the configuration data the types of third and fourth component mounting machines, which have different component supply methods, that should be installed on the planned mounting line.

[0015] (Feature 4) In the data creation device described above, the data creation device may further calculate the production time required for the production of the planned production board based on the configuration data of the planned mounting line that it has created.

[0016] With this configuration, the production time on the planned implementation line is calculated based on the created configuration data, so for example, an operator can easily determine whether the configuration data is appropriate or not.

[0017] (First Embodiment) As shown in Figure 1, the data creation device 10 of this embodiment includes a display unit 12, an operation unit 14, a communication interface 16, and a control unit 20. The data creation device 10 is installed on the Internet 4 by, for example, a component mounting machine manufacturer. The data creation device 10 is connected to a plurality of component mounting factories F1 to F3 via the communication interface 16. In a modified example, the data creation device 10 may be installed on the Internet 4 by a company other than a component mounting machine manufacturer.

[0018] The display unit 12 displays, for example, various information related to component mounting factories F1 to F3. The display unit 12 also displays the input screen SC1 (see Figure 2), which will be described later. The operation unit 14 receives various operations from the operator (not shown). The operation unit 14 receives, for example, input of component information, which will be described later.

[0019] Each of the component mounting factories F1 to F3 is equipped with mounting lines L10 to L30. Each mounting line L10 to L30 consists of multiple component mounting machines. Although not shown in the diagram, each component mounting machine is a device that mounts components (e.g., electronic components) onto a circuit board to produce component-mounted boards, and is also called an electronic component mounting device or chip mounter. In addition to component mounting machines, mounting lines L10 to L30 may be installed together with other board work machines such as solder printing machines and board inspection machines.

[0020] The control unit 20 includes a CPU 22 and a memory 24. The CPU 22 is a hardware processor that performs various processes based on the information stored in the memory 24. The memory 24 is composed of non-volatile memory and volatile memory. The memory 24 stores the learner 30.

[0021] The learning device 30 is equipped with a neural network and is generated by machine learning based on training data T1 for creating the configuration data D1 (see Figure 2), which will be described later. The training data T1 includes various information about the mounting lines that have previously mounted components and produced component-mounted boards. Hereafter, mounting lines that have a track record of mounting components and producing component-mounted boards will be specifically referred to as "proven mounting lines."

[0022] In this embodiment, as shown in Figure 1, the training data T1 includes, for example, actual configuration data A1, actual component information A2, and actual production time A3 from actual mounting lines L10, L20, and L30. The data creation device 10 receives, for example, the actual configuration data A1, actual component information A2, and actual production time A3 from the component mounting factory F1 when production of a component mounted board is completed at mounting line L10 (i.e., actual mounting line L10) at the component mounting factory F1. As a result, the data creation device 10 updates (stores) the training data T1.

[0023] The learner 30 learns the weights of the neural network, for example, using training data T1. The learner 30 inputs the input values ​​of the training data T1 into the input layer and performs training of the neural network weights using backpropagation to minimize the error between the output values ​​of the neural network and the output values ​​of the training data T1. Furthermore, the learner 30 stores the learned weights in memory 24.

[0024] Actual configuration data A1 shows the configuration of the actual mounting line. Actual configuration data A1 includes mounting machine information regarding the component mounting machine equipped on the actual mounting line, and nozzle information regarding the nozzle that picked up the component by the component mounting machine.

[0025] In this embodiment, component mounting machines of models M1 and M2 are installed on the actual mounting line. The component mounting machine of model M1 is characterized by mounting components onto circuit boards at a relatively low speed, and is suitable for producing small quantities of various component-mounted boards by mounting multiple types of components onto multiple types of circuit boards. In contrast, the component mounting machine of model M2 can mount components onto circuit boards at a higher speed than the component mounting machine of model M1, and is suitable for producing large quantities of identical component-mounted boards by mounting the same type of components onto the same type of circuit boards. The mounting machine information includes information on the number of each component mounting machine of models M1 and M2 installed on the actual mounting line. For example, the actual mounting line L10 is equipped with three component mounting machines of model M1.

[0026] Furthermore, in this embodiment, regardless of the component mounting machine models M1 and M2, nozzles of types N1, N2, and N3 are arranged inside the component mounting machine. Nozzle type N1 is suitable for picking up relatively small components. Nozzle type N3 is suitable for picking up relatively large components, and nozzle type N2 is suitable for picking up components of sizes between types N1 and N3. The nozzle information includes quantity information indicating how many of each type N1, N2, and N3 nozzle was used in the production of component mounted boards on the actual mounting line. For example, in the first component mounting machine equipped on the actual mounting line L10, 12 nozzles of type N1 are arranged, in the second component mounting machine, 5 nozzles of type N2 are arranged, and in the third component mounting machine, 4 nozzles of type N3 are arranged. In this embodiment, one type of nozzle is arranged in each component mounting machine, but for example, multiple types of nozzles may be arranged in each component mounting machine. In that case, the component mounting machine may, for example, after the component is picked up and mounted using a nozzle of type N1, remove the nozzle of type N1 from the mounting header (not shown) and attach a nozzle of type N2.

[0027] Actual component information A2 is information regarding the types and number of components mounted on component-mounted boards produced on the actual mounting line, and includes the name of the component and the number of that component. For example, in the first component mounting machine equipped on the actual mounting line L10, 120 components with name P1 were mounted on the circuit board, in the second component mounting machine, 5 components with name P2 were mounted on the circuit board, and in the third component mounting machine, 72 components with name P3 were mounted on the circuit board.

[0028] Actual production time A3 indicates the production time required for past production of component-mounted boards on the actual mounting line. In this embodiment, actual production time A3 is calculated by a management device (not shown) for each actual mounting line. The management device, for example, calculates the production time under multiple mounting conditions when mounting a predetermined number of components of a predetermined name on the actual mounting line to which it is connected, and identifies the mounting condition that allows for the shortest production time. Here, the mounting conditions include, for example, the mounting order of components and the mounting position of components on the circuit board. The management device calculates the production time under the identified mounting condition as actual production time A3. That is, actual production time A3 is the time required to produce a component-mounted board produced by mounting components of a predetermined name and number indicated by actual component information A2 on an actual mounting line having the configuration indicated by actual configuration data A1. In a modified example, the management device may measure the actual time required to produce a component-mounted board on the actual mounting line L10 instead of the production time calculated under the mounting condition that allows for the shortest production time, and calculate this time as actual production time A3.

[0029] Referring to Figure 2, the configuration data generation process executed by the CPU 22 of the control unit 20 of the data creation device 10 will be described. The configuration data creation process is a process for generating configuration data D1 that indicates the configuration of the mounting line for producing component mounting boards, based on information about component mounting boards input by the worker before the component mounting boards are produced. In other words, the configuration data creation process is a process that automatically creates a mounting line configuration that is considered suitable for the production of component mounting boards, and the worker determines whether the created mounting line configuration is appropriate or not. Conventionally, workers set the mounting line configuration based on their own experience, for example, and calculated the production time for that configuration. As a result, it was difficult for less experienced workers to set an appropriate mounting line configuration, and it sometimes took a long time to adjust the mounting line configuration until the calculated production time fell below the acceptable production time. Hereinafter, a mounting line having the configuration indicated by the configuration data D1 created by the configuration data creation process will be referred to as the "planned mounting line," and a component mounting board produced on the planned mounting line will be referred to as the "planned production board." For example, when the CPU 22 receives a data creation instruction from an operator via the operation unit 14, it executes the data creation process.

[0030] In S10, the CPU 22 displays the input screen SC1 on the display unit 12. The input screen SC1 is a screen for receiving information from the operator that is a prerequisite for creating the configuration data D1 of the planned assembly line. The input screen SC1 includes a component information input unit C11, a planned production time input unit C12, and a configuration data creation button B11. The component information input unit C11 receives information from the operator regarding the components to be mounted on the planned production board produced on the planned assembly line. The component information input unit C11 includes an input field for the name of the component to be mounted on the planned production board and an input field for the number of the component. The planned production time input unit C12 also receives information from the operator regarding the target production time for the production of the planned production board, that is, the production time that is allowed for the production of the planned production board. The configuration data creation button B11 is a button for accepting the start of the creation of configuration data D1 based on the information entered in each input unit C11 and C12.

[0031] In S20, the CPU 22 determines whether information has been entered into each input unit C11 and C12. The CPU 22 continues to display the input screen SC1 until information has been entered into each input unit C11 and C12. When information has been entered into each input unit C11 and C12 and the selection of the configuration data creation button B11 is accepted, the CPU 22 determines that the information has been entered (YES in S20) and proceeds to S22.

[0032] In S22, the CPU 22 performs calculations based on the information input to each input unit C11 and C12. The CPU 22 inputs the names and quantities of the parts input to each input unit C11 and C12, as well as the planned production time, to the learner 30 (see Figure 1) in the memory 24. The learner 30 inputs each piece of information into the input layer of the neural network and performs calculations using the neural network, which takes into account the stored weights described above. This generates the configuration data D1 for the planned assembly line.

[0033] In S24, the CPU 22 further calculates the production time. Specifically, the CPU 22 calculates the implementation conditions that result in the shortest production time based on the configuration data D1 of the planned implementation line created in the calculation process of S22, and calculates the production time under those implementation conditions.

[0034] In S30, the CPU 22 displays the output screen SC2 on the display unit 12. The output screen SC2 includes the configuration data D1 of the planned mounting line created in S22 and the production time PT1 calculated in S24. As shown in Figure 2, the configuration data D1 includes mounting machine information and nozzle information. The mounting machine information includes the model (e.g., M1) and number (e.g., 3) of the component mounting machines to be equipped on the planned mounting line. In this way, the configuration data D1 can display the model to be equipped on the planned mounting line from among models M1 and M2. The mounting machine information further includes the supply type. The supply type is information regarding the supply of components in the component mounting machine. The supply type includes supply type F, which accepts component supply by a feeder, and supply type T, which accepts component supply by a tray. The supply type is determined regardless of the model M1 and M2 of the component mounting board, for example, according to the type and number of components. Thus, the configuration data D1 can display the supply type that should be installed on the planned mounting line from among supply types F and T. Furthermore, for example, a component mounting machine of supply type T tends to have a longer production time compared to a component mounting machine of supply type F. For this reason, for example, if the production time PT1 is longer than expected, the operator can check whether or not there is a component mounting machine of supply type T. Also, since the output screen SC2 includes the production time PT1, the operator can determine whether the production time based on the configuration data D1 is less than the planned production time entered in S10. For this reason, it is easy to determine whether or not the configuration data D1 generated in S22 is appropriate.

[0035] In S32, the CPU 22 displays the modification screen SC3 on the display unit 12. The modification screen SC3 is a screen for the operator to modify the information that is the basis for creating the configuration data D1, based on the configuration data D1 and production time PT1 displayed on the output screen SC2. The modification screen SC3 includes a parts information input unit C21, a planned production time input unit C22, a configuration data creation button B21, and a cancel button B22, similar to the input screen SC1 described above. For example, if the production time PT1 displayed on the output screen SC2 exceeds the planned production time entered in S10, the operator enters the modified information into the respective input units C21 and C22 and selects the configuration data creation button B21.

[0036] In S40, the CPU 22 determines that correction information has been entered (YES in S40) when information has been input to each input unit C21, C22 and the selection of the configuration data creation button B21 is accepted, and returns to S22 to execute the calculation process again. As a result, new configuration data D1 is created. If the cancel button B22 on the correction screen SC3 is selected, the CPU 22 determines that no correction information has been entered (NO in S40) and terminates the process shown in Figure 2.

[0037] (Effects of this embodiment) As described above, the data creation device 10 of this embodiment creates configuration data D1 of the planned assembly line using a learner 30 generated by machine learning using training data T1 which includes actual parts information A2, actual configuration data A1, and actual production time A3 (S22 in Figure 2). Therefore, compared to conventional technology that creates configuration data based on the experience of the worker, it is possible to create appropriate configuration data D1 of the planned assembly line without relying on the experience of the worker, and the time required for subsequent correction work can be shortened. As a result, the work time required to create the configuration data D1 of the planned assembly line can be shortened. Furthermore, since it is prevented, for example, from generating configuration data D1 based on the biased experience of the worker, appropriate configuration data D1 can be created stably.

[0038] The correspondence in this embodiment is as follows. The nozzle is an example of the "component holding part". Model M1 is an example of the "first type", and model M2 is an example of the "second type". Supply type F is an example of the "third type", and supply type T is an example of the "fourth type".

[0039] The process of S10 in FIG. 2 is an example of the process executed by the "input unit". The process of S22 is an example of the process executed by the "arithmetic unit".

[0040] Points to note regarding the data creation device 10 described in the embodiment are described. The data creation device 10 does not have to be installed on the Internet 4. For example, it may be directly connected to the mounting line L10 of the component mounting factory F1.

[0041] The component mounter may be provided with a chuck for holding components instead of a nozzle. In that case, the teacher data T1 may include information regarding the type of chuck instead of the nozzle types N1 to N3.<0OO0083>

[0042] The component information input unit C11 of the input screen SC1 in FIG. 2 may be provided with a field for inputting an image of the component instead of or in addition to the field for inputting the name of the component. In that case, in the arithmetic processing of S22, the learning device 30 may identify the components to be mounted on the production-scheduled substrate based on the input image of the component and generate the configuration data D1. In a further modification, the component information input unit C11 may be provided with a field for inputting the size of the component. In that case, the learning device 30 may identify the components to be mounted on the production-scheduled substrate based on the input size of the component and generate the configuration data D1.

[0043] The planned mounting line may be equipped only with the component mounters of model M1 or only with the component mounters of model M2. In a further modification, the planned mounting line may be further equipped with the component mounters of model M3.

[0044] The mounting machine information in the configuration data D1 does not have to include the supply types F and T.

[0045] The output screen SC2 does not have to include the production time PT1.

[0046] The technical elements described herein or in the drawings demonstrate technical usefulness individually or in various combinations, and are not limited to the combinations described in the claims at the time of filing. Furthermore, the technologies illustrated herein or in the drawings achieve multiple objectives simultaneously, and achieving even one of these objectives constitutes technical usefulness in itself.

[0047] For example, this specification also discloses a technical concept in which "the data creation device described in claim 1" is changed to "the data creation device described in claim 1 or 2" in claim 3. Similarly, a technical concept in which "the data creation device described in claim 1" is changed to "the data creation device described in any one of claims 1 to 3" in claim 4, and a technical concept in which "the data creation device described in claim 1" is changed to "the data creation device described in any one of claims 1 to 4" in claim 5.

[0048] 4: Internet 10: Data creation device 12: Display unit 14: Operation unit 16: Communication I / F 20: Control unit 22: CPU 24: Memory 30: Learning unit A1: Actual configuration data A2: Actual component information A3: Actual production time C11, C21: Component information input unit C12, C22: Planned production time input unit D1: Configuration data F, T: Supply type F1, F2, F3: Component mounting factory L10, L20, L30: Actual mounting line M1, M2: Model N1, N2, N3: Type P1, P2, P3: Name T1: Training data

Claims

1. A data creation device for creating configuration data indicating the configuration of a mounting line for producing component mounting boards, comprising: an input unit for inputting planned board information relating to planned production boards; and a calculation unit for creating configuration data for a planned mounting line for producing the planned production boards based on the input planned board information, wherein the planned board information includes at least planned component information relating to the types and number of components to be mounted on the planned production boards, and planned production time information relating to the production time allowed for the production of the planned production boards, and the configuration data includes at least component mounting machine information relating to the types and number of component mounting machines equipped on the mounting line, and component holding unit information relating to the types and number of component holding units that hold the components arranged in the component mounting machine for each of the component mounting machines equipped on the mounting line, The calculation unit is a data creation device that creates the configuration data for the planned assembly line using a learner generated by machine learning, with the following as training data: actual component information relating to the types and number of components mounted on actual component mounting boards produced in the past; actual configuration data indicating the configuration of the assembly line that produced the actual component mounting boards; and actual production time information relating to the actual production time required to produce the actual component mounting boards.

2. The data creation apparatus according to claim 1, wherein the planned component information includes at least one of the following: information indicating the name of the component to be mounted on the planned production board; an image of the component to be mounted on the planned production board; and information indicating the size of the component to be mounted on the planned production board.

3. The data creation apparatus according to claim 1, wherein the types of component mounting machines equipped on the mounting line include: a first type indicating a first component mounting machine capable of mounting a first number of the components per unit time; and a second type indicating a second component mounting machine capable of mounting a second number of the components greater than the first number per unit time.

4. The data creation apparatus according to claim 1, wherein the types of component mounting machines equipped on the mounting line include a third type indicating a third component mounting machine that receives the supply of components by a feeder, and a fourth type indicating a fourth component mounting machine that receives the supply of components by a tray.

5. The data creation device according to claim 1, further calculating the production time required for the production of the planned production board based on the configuration data of the planned assembly line that has been created.

Images