Component mounting system and component mounting method

Abstract

To provide a component mounting system capable of determining the existence / absence of foreign substances at the mounting position without requiring preparation of a reference image for each mounting position to be inspected, and a component mounting method.SOLUTION: The component mounting system executes the following steps: picking up an image of a component mounting position on a circuit board before mounting the component on the circuit board (ST3); determining whether a foreign substance is present at the mounting position from the images taken at the mounting position by using a learning model that has learned multiple images of foreign substances on the circuit board as teacher data (ST4); and mounting the component onto the mounting position (ST6) after determining that no foreign substance is present in the mounting position (No in ST4).SELECTED DRAWING: Figure 5

Description

[Technical field]

[0001] The present invention relates to a component mounting apparatus and a component mounting method for mounting components on a board. [Background technology]

[0002] When a component mounting device mounts a component on a board, the component may be dropped onto the board by mistake. If a component is mounted when there is a foreign object such as a component at the mounting position on the board, there is a risk of a mounting failure in which the component is not properly mounted on the board. Therefore, immediately after the board is carried in or immediately before mounting the component, the mounting position on the board is inspected for the presence of foreign objects, and the component is mounted only after confirming that there is no foreign object. Patent Document 1 discloses a component mounting device (surface mounter) that compares a reference image of the mounting position without foreign objects, which is captured in advance by a board recognition camera, with an image of the mounting position immediately before mounting the component, to determine the presence or absence of foreign objects. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2008-60249 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, in conventional technologies including Patent Document 1, it is necessary to prepare reference images of all mounting positions to be inspected for each type of mounting board. This poses a problem in that when there are many inspection targets, the workload of the worker preparing the reference images is heavy.

[0005] Therefore, an object of the present invention is to provide a component mounting device and a component mounting method that can determine the presence or absence of foreign matter at a mounting position without the need to prepare a reference image for each mounting position to be inspected. [Means for solving the problem]

[0006] The component mounting apparatus of the present invention comprises an imaging means for imaging the mounting position of a component on a board before the component is mounted on the board, a determination means for determining whether or not a foreign object is present at the mounting position from the image captured by the imaging means using a learning model trained using multiple images of foreign objects on the board as teacher data, and a component mounting means for mounting the component at the mounting position when the determination means determines that no foreign object is present at the mounting position.

[0007] The component mounting method of the present invention captures an image of the mounting position of the component on a board before mounting the component on the board, and uses a learning model that has been trained using multiple images of foreign objects on the board as training data to determine whether or not a foreign object is present at the mounting position from the captured image of the mounting position, and if it is determined that no foreign object is present at the mounting position, the component is mounted at the mounting position. Effect of the Invention

[0008] According to the present invention, it is possible to determine the presence or absence of a foreign object at a mounting position without the need to prepare a reference image for each mounting position to be inspected. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a plan view of a component mounting system according to an embodiment of the present invention; [Diagram 2] FIG. 1 is a block diagram showing a configuration of a component mounting system according to an embodiment of the present invention. [Diagram 3] FIG. 1 is a plan view showing an example of a board on which components are mounted by a component mounting apparatus according to an embodiment of the present invention; [Figure 4] 1A, 1B, 1C, and 1D are explanatory diagrams of an example of a captured image of a mounting position used for learning a learning model in a component mounting system according to an embodiment of the present invention. [Diagram 5] 1 is a flow diagram of a component mounting method according to an embodiment of the present invention; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] An embodiment of the present invention will be described in detail below with reference to the drawings. The configurations, shapes, etc. described below are examples for explanation and can be changed as appropriate according to the specifications of the component mounting system, the component mounting device, and the management computer. In the following, the same reference numerals are used for corresponding elements in all the drawings, and duplicated explanations will be omitted. In FIG. 1 and in some parts described later, the X-axis (left-right direction in FIG. 1) in the board transport direction and the Y-axis (up-down direction in FIG. 1) perpendicular to the board transport direction are shown as two axes perpendicular to each other in the horizontal plane. In FIG. 1 and in some parts described later, the Z-axis (perpendicular to the paper surface in FIG. 1) is shown as the height direction perpendicular to the horizontal plane.

[0011] First, the configuration of the component mounting system 1 will be described with reference to Fig. 1. The component mounting system 1 is configured to include a component mounting device M and a management computer CP that are connected via a communication network NT. The component mounting device M has a function of performing a component mounting operation of mounting components supplied from a component supply unit onto a board. The management computer CP has a function of supporting the manufacture of a mounted board by the component mounting device M, etc. The management computer CP of this embodiment has a function of generating a learning model for determining the presence or absence of a foreign object at a mounting position used by the component mounting device M, based on a plurality of captured images of the mounting positions on the board.

[0012] Next, the configuration of component mounting apparatus M will be described with reference to Fig. 1. In component mounting apparatus M, a board transport mechanism 3 is arranged along the X-axis at the center of a base 2. The board transport mechanism 3 transports a board KB carried in from upstream along the X-axis, and positions and holds the board at a mounting work position by a mounting head, which will be described below. The board transport mechanism 3 also carries out the board KB downstream after component mounting work has been completed. A component supply unit 4 is installed on both sides of the board transport mechanism 3 (front and rear directions of the Y-axis).

[0013] A plurality of tape feeders 5 are attached in parallel along the X-axis to both component supply units 4. The tape feeders 5 feed the carrier tape, on which pockets for storing components are formed, in a direction (tape feed direction) from the outside of the component supply unit 4 toward the board transport mechanism 3, thereby supplying the components to the component removal position where the mounting head picks up the components. In addition, one of the component supply units 4 is attached with a tray feeder 8 that supplies a tray pallet 7, on which trays 6 containing components D such as large components, connectors, and communication modules are placed, to the component removal position.

[0014] In Fig. 1, a Y-axis table 9 equipped with a linear drive mechanism is arranged along the Y-axis at both ends in the X-axis direction on the upper surface of the base 2. A beam 10 similarly equipped with a linear mechanism is coupled to the Y-axis table 9 so as to be freely movable along the Y-axis. A mounting head 11 is attached to the beam 10 so as to be freely movable along the X-axis. The mounting head 11 is equipped with a plurality of suction units (not shown) that can hold a component D and move up and down in the Z-axis direction, and can rotate around a vertical axis (Z-axis). A suction nozzle (not shown) that suctions and holds the component D is attached to the lower end of each suction unit.

[0015] The Y-axis table 9 and the beam 10 constitute a mounting head moving mechanism that moves the mounting head 11 in the horizontal direction (X-axis direction, Y-axis direction). The mounting head moving mechanism and the mounting head 11 constitute a component mounting means 12 that performs a component mounting operation of taking out components D from the component supply unit 4 and mounting them at the mounting positions on the board KB. In the component mounting operation, the mounting head 11 moves above the component supply unit 4 and picks up predetermined components D with each suction nozzle. Next, the mounting head 11 moves above the board KB and rotates the components D held by each suction nozzle in a predetermined direction, repeating a series of mounting turns to mount them at their respective mounting positions.

[0016] 1, beam 10 is fitted with head camera 13 which is located on the underside of beam 10 and moves integrally with mounting head 11. As mounting head 11 moves, head camera 13 moves above board KB which is positioned at the mounting operation position of board transport mechanism 3, and captures images of board marks (not shown) provided on board KB, the mounting position of component D, and the like.

[0017] A component recognition camera 14 is disposed between the component supply unit 4 and the board transport mechanism 3. When the mounting head 11, which has taken out a component D from the component supply unit 4, is positioned above the component recognition camera 14, the component recognition camera 14 images the component D held by the suction nozzle from below. From the image capture result, the holding posture of the component D is recognized. During the component mounting operation of the mounting head 11 to the board KB, the mounting position is corrected taking into account the image capture result of the board KB captured by the head camera 13 and the image capture result of the component D captured by the component recognition camera 14.

[0018] A touch panel 15 operated by the worker is provided at the position where the worker works on the front side of the component mounting device M. The touch panel 15 displays various information on its display section, and the worker inputs data and operates the component mounting device M using operation buttons and the like displayed on the display section.

[0019] Next, the configuration of the control system of the component mounting system 1 will be described with reference to Fig. 2. Here, the description will focus on the function of the component mounting device M capturing an image of the mounting position on the board KB, and mounting the component D after confirming that there is no foreign object such as the component D at the mounting position. The board transport mechanism 3, the component supply unit 4, the component mounting means 12, the head camera 13, the component recognition camera 14, and the touch panel 15 are connected to the control device 20 provided in the component mounting device M. The control device 20 also includes a mounting memory unit 21, a mounting control unit 22, a determination means 23, and a mounting communication unit 24. The mounting memory unit 21 is a storage device, and stores mounting data 21a, a learning model 21b, and the like.

[0020] The mounting data 21a stores the production model name of the mounting board, the component number (type of component) specifying the type of component D to be mounted on the board KB, the supply position of the component D in the component supply unit 4 (tape feeder 5, tray feeder 8, etc.), the mounting position (XY coordinates), the mounting direction, the mounting order, etc. Also, the mounting data 21a stores information indicating that the component D, which is to be determined whether or not a foreign object exists in the mounting position before mounting, is a foreign object determination target component. The learning model 21b is generated as teacher data by the management computer CP or the like as described later, using a plurality of captured images in which a foreign object exists in the mounting position on the board KB, and is stored in the mounting storage unit 21.

[0021] Here, with reference to Fig. 3, an example of the board KB on which the component D is mounted by the component mounting apparatus M will be described. Fig. 3 shows a part of the board KB. The board KB has electrodes E (lands) formed at positions corresponding to the terminals of the components D1 to D6 to be mounted on the board KB. The board KB has a plurality of electrodes E formed at mounting positions P1 to P6 on which the components D1 and D2, which are BGAs with bumps formed on their bottom surfaces, and the components D3 to D6, which are chip components such as resistors and capacitors, are mounted. Here, the mounting positions P1 to P6 are positions on the board KB that are targets for mounting the components D1 to D6 on the board KB so that the centers of the components D1 to D6 are aligned. In this specification, the mounting positions P1 to P6 may also refer to areas where the components D1 to D6 mounted on the board KB exist.

[0022] Component D1 is a BGA with 9 bumps. Component D2 is a BGA with 16 bumps. If large components D1 and D2 are mounted at mounting positions P1 and P2 where foreign objects such as chip components have fallen, there is a risk of a mounting failure in which the bumps and electrodes E are not properly connected. For this reason, the mounting data 21a indicates that the large components D1 and D2 are foreign object detection target components, which are used to determine whether or not there is a foreign object at mounting positions P1 and P2 before mounting.

[0023] 2, the mounting control unit 22 controls the component supply unit 4 and the component mounting means 12 based on the mounting data 21a to mount the component D supplied from the component supply unit 4 at the mounting position on the board KB. Before mounting the components D1 and D2 set in the mounting data 21a as foreign object detection target components on the board KB, the mounting control unit 22 causes the head camera 13 to capture images of the mounting positions P1 and P2. The determination means 23 uses the learning model 21b stored in the mounting storage unit 21 to determine whether or not a foreign object exists at the mounting positions P1 and P2 from the captured images of the mounting positions P1 and P2 to be inspected captured by the head camera 13.

[0024] When the determination means 23 determines that no foreign object such as component D or dust is present at the mounting position, the mounting control unit 22 controls the component mounting means 12 to mount the component for which foreign object detection is to be performed on the board KB. Furthermore, when the determination means 23 determines that a foreign object is present at mounting positions P1, P2, the mounting control unit 22 interrupts the component mounting operation, causes the touch panel 15 to display a warning that a foreign object is present at mounting positions P1, P2, and causes the captured image in which it is determined that a foreign object is present to be transmitted to the management computer CP via the mounting communication unit 24.

[0025] In this way, head camera 13 is an imaging means that images mounting positions P1, P2 of components D1, D2 on board KB before components D1, D2 are mounted on board KB. Determination means 23 determines whether or not a foreign object exists at mounting positions P1, P2 from the images captured by imaging means (head camera 13) using learning model 21b that has been trained using multiple images of foreign objects on board KB as teacher data. Then, when determination means 23 determines that no foreign object exists at mounting positions P1, P2, component mounting means 12 mounts components D1, D2 at mounting positions P1, P2.

[0026] 2, the management computer CP includes a management storage unit 31, a management communication unit 32, a collection unit 33, and a learning unit 34. The management storage unit 31 is a storage device, and stores captured image data 31a, a learning model 31b, etc. The implementation communication unit 24 and the management communication unit 32 are communication devices, and are connected to each other via a communication network NT, and transmit and receive various types of data to and from each other.

[0027] The collecting unit 33 collects captured images in which it has been determined that a foreign object exists during component mounting work from the component mounting device M, and stores them as captured image data 31a in the management storage unit 31. The collecting unit 33 also collects captured images in which a foreign object exists at the mounting position of the board KB, captured using the component mounting device M or the like, as teacher data used by the worker when generating the learning model 31b, and stores them as captured image data 31a in the management storage unit 31.

[0028] Here, an example of a captured image captured by a worker will be described with reference to Fig. 4. Fig. 4(a) and Fig. 4(b) are examples of captured images captured by head camera 13 of component mounting apparatus M. That is, the images are images of imaging field of view 13a captured by head camera 13 of mounting positions P7, P8 of component D1, which is a BGA, to be mounted on boards KB1, KB2. Small components D7-D9 that have been dropped in advance by the worker are present at mounting positions P7, P8 on boards KB1, KB2.

[0029] FIG. 4(a) is a captured image in which component D7 is present at mounting position P7 for component D1 on board KB1. FIG. 4(b) is a captured image in which component D8, which is larger than component D7, is present at mounting position P7 for component D1 on board KB1. The position, orientation, and posture of component D8 are different from those of component D7 in FIG. 4(a). FIG. 4(c) is a captured image in which component D9 is present at mounting position P7 for component D1 on board KB1. Component D9 is made of a different type of material, such as resin, from components D7 and D8. FIG. 4(d) is a captured image in which component D7 is present at mounting position P8 for component D1 on board KB2. Board KB2 is made of a different type of material from board KB1.

[0030] In this way, the multiple captured images contained in the captured image data 31a include multiple images in which the positions of foreign objects (components D7 to D9) are different, multiple images in which the types of foreign objects are different, multiple images in which the types of foreign objects are different, multiple images in which the postures of foreign objects (directions, standing postures, etc.) are different, and multiple images in which the types of boards KB1 and KB2 are different.

[0031] In FIG. 2, the learning unit 34 uses a plurality of images included in the stored captured image data 31a as teacher data to generate a learning model 31b by a learning algorithm using machine learning or the like. As the learning algorithm, a neural network (including deep learning using a multi-layered neural network), genetic programming, a decision tree, a Bayesian network, a support vector machine (SVM), or the like can be used. The generated learning model 31b is transmitted to the component mounting device M via the management communication unit 32 and stored in the mounting storage unit 21 as the learning model 21b. The learning model 31b created by the management computer CP may be directly stored (installed) in the mounting storage unit 21 of the component mounting device M without going through the communication network NT.

[0032] By using learning model 31b trained in this way, even if the teacher data does not contain an image that completely matches the shape and arrangement of electrodes E or the type, size, and position of component D that is a foreign object, it is possible to determine whether or not a foreign object exists from the captured image of mounting position P. For example, by using learning model 31b trained using the teacher data that includes the captured image of mounting position P1 of component D1 shown in Fig. 3, it is possible to determine whether or not a foreign object exists from the captured image of mounting position P2 of component D2, even if the teacher data does not include and train component D2, which has a different number of electrodes E.

[0033] Next, a component mounting method for mounting components D on a board KB will be described with reference to the flow of Fig. 5. Here, a single mounting turn in which the mounting head 11 takes out components D from the component supply unit 4 and mounts them at the mounting position on the board KB will be described as an example. Also, it is assumed that a learning model 21b generated by a management computer CP or the like is stored in advance in the mounting storage unit 21 of the component mounting device M.

[0034] First, the mounting control unit 22 of the component mounting device M controls the component mounting means 12 based on the mounting data 21a to take out the component D from the component supply unit 4 and have it held by the mounting head 11 (ST1: component holding step). Next, the mounting control unit 22 determines whether the component D held by the mounting head 11 is a component subject to foreign matter determination based on the mounting data 21a (ST2: foreign matter determination target determination step). If the component D held by the mounting head 11 is a component subject to foreign matter determination (Yes in ST2), the mounting control unit 22 causes the head camera 13 (imaging means) to capture an image of the mounting position on the board KB (ST3: mounting position imaging step).

[0035] For example, when the mounting head 11 is holding a BGA component D1 (see FIG. 3), the component is set as a component subject to foreign object detection in the mounting data 21a, and is therefore determined to be a component subject to foreign object detection (Yes) in the foreign object detection target determination step (ST2). Then, the mounting position P1 is imaged by the head camera 13 (ST3).

[0036] 5, the determination means 23 then uses the learning model 21b to determine whether or not a foreign object is present at the mounting position from the captured image of the mounting position (ST4: determination step). If it is determined that a foreign object is present in the determination step (ST4) (Yes), the mounting control unit 22 causes the touch panel 15 to display a warning that a foreign object is present at the mounting position P (ST5: warning step).

[0037] If the component D held by the mounting head 11 is not a component subject to foreign matter detection (No in ST2), or if it is not determined in the detection process (ST4) that a foreign matter is present (No), the mounting control unit 22 controls the component mounting means 12 based on the mounting data 21a to mount the component D held by the mounting head 11 at the mounting position on the board KB (ST6: component mounting process).

[0038] For example, when the mounting head 11 is holding a chip component D3 (see FIG. 3), it is not set as a component subject to foreign matter detection in the mounting data 21a, and is therefore determined not to be a component subject to foreign matter detection (No) in the foreign matter detection target determination step (ST2). Then, the component D3 held by the mounting head 11 is mounted at the mounting position P3 on the board KB (ST6).

[0039] In this way, the component mounting method of this embodiment captures an image of mounting position P1 of component D1 on board KB before mounting component D1 on board KB (ST3), and uses learning model 21b that has learned from multiple images of foreign objects on board KB as teacher data to determine whether or not a foreign object is present at mounting position P1 from the captured image of mounting position P1 (ST4), and if it is determined that no foreign object is present at mounting position P1 (No in ST4), component D1 is mounted at mounting position P1 (ST6). This makes it possible to determine the presence or absence of a foreign object at a mounting position without the need to prepare a reference image for each mounting position to be inspected.

[0040] As described above, the component mounting apparatus M of this embodiment includes an imaging means (head camera 13) that images the mounting position P1 of the component D1 on the board KB before the component D1 is mounted on the board KB, a determination means 23 that determines whether or not a foreign object exists at the mounting position P1 from the image captured by the imaging means using a learning model 21b that has been trained using multiple images of foreign objects on the board KB as teacher data, and a component mounting means 12 that mounts the component D1 at the mounting position P1 if the determination means 23 determines that no foreign object exists at the mounting position P1. This makes it possible to determine the presence or absence of a foreign object at a mounting position without the need to prepare a reference image for each mounting position to be inspected. [Industrial Applicability]

[0041] The component mounting apparatus and method of the present invention have the advantage of being able to determine the presence or absence of foreign matter at a mounting position without the need to prepare a reference image for each mounting position to be inspected, and are therefore useful in fields where components are mounted on boards. [Explanation of symbols]

[0042] 1 Component mounting system 12 Component mounting means 13 Head camera (imaging means) D, D1~D9 Parts KB, KB1, KB2 board M Component Mounting Equipment P1~P8 Installation position

Claims

1. an imaging means for imaging a mounting position of the component on the board before the component is mounted on the board; a determination means for determining whether or not a foreign object is present at the mounting position from the image captured by the imaging means, using a learning model that has been trained using a plurality of images of foreign objects on a substrate as teacher data; and a component mounting means for mounting the component at the mounting position when the determining means determines that no foreign object is present at the mounting position.

2. The component mounting apparatus according to claim 1 , wherein the foreign matter is a component or dust.

3. The component mounting device according to claim 1 , wherein the plurality of images include a plurality of images in which the positions of foreign matter are different.

4. The component mounting device according to claim 1 , wherein the plurality of images include a plurality of images of different types of foreign matter.

5. The component mounting device according to claim 1 , wherein the plurality of images include a plurality of images of foreign objects having different sizes.

6. The component mounting device according to claim 1 , wherein the plurality of images include a plurality of images in which the foreign matter has different postures.

7. The component mounting device according to claim 1 , wherein the plurality of images include a plurality of images of different types of boards.

8. Taking an image of a mounting position of a component on a board before mounting the component on the board; determining whether or not a foreign object is present at the mounting position from the captured image of the mounting position using a learning model that has been trained using a plurality of images of foreign objects present on the substrate as training data; When it is determined that no foreign matter is present at the mounting position, the component is mounted at the mounting position.

Images