KR20230143414A - Fault prediction apparatus based on surface mounting device, and control method thereof - Google Patents

Abstract

The present invention relates to an SMD-based defect prediction device and its operating method for predicting PCB (Printed Circuit Board) defects that occur during the process of an SMT (Surface Mounter Technology) line before proceeding with the subsequent process.

Description

SMD-based defect prediction device and its operation method {FAULT PREDICTION APPARATUS BASED ON SURFACE MOUNTING DEVICE, AND CONTROL METHOD THEREOF}

The present invention relates to a method for predicting printed circuit board (PCB) defects that occur during the process of an SMT (Surface Mounter Technology) line before proceeding with the subsequent process.

In the SMT (Surface Mounter Technology) line, the process of printing solder paste on a PCB (Printed Circuit Board) through a screen printer and mounting actual components on the PCB on which the solder paste has been printed is generally performed through a mounter.

In addition, in the SMT line, inspection equipment can be used to check PCB defects, and as SMT technology develops, the demand for such inspection equipment is gradually increasing.

However, such PCB defect status inspection can be performed during or after the process. In the case of defect status inspection performed after the process is completed, it is performed after product production, which increases quality costs and makes it difficult to check the possibility of equipment failure. It has limitations.

The present invention was created in consideration of the above-mentioned circumstances, and the goal to be achieved by the present invention is to predict PCB (Printed Circuit Board) defects that occur during the process of the SMT (Surface Mounter Technology) line before proceeding with the subsequent process. I'm doing it.

To achieve the above object, the SMD-based defect prediction device according to an embodiment of the present invention acquires inspection data from at least one inspection equipment that checks the status of the PCB (Printed Circuit Board) in the SMT (Surface Mounting Technology) line. an acquisition department that does; A generator that generates a learning data set by preprocessing the inspection data; and a learning unit that learns a defect prediction model for predicting PCB defect items based on the training data set.

Specifically, the device may further include a prediction unit that predicts PCB defect items that occur in a subsequent process after checking the condition of the PCB through the defect prediction model.

Specifically, the acquisition unit acquires heterogeneous inspection data from the at least one inspection equipment, and the generating unit may generate a learning data set through preprocessing that links the heterogeneous inspection data.

Specifically, the heterogeneous inspection data includes SPI numerical data, which is inspection data of Solder Paste Inspection (SPI) equipment, and judgment results for each AOI defect item, which is inspection data of Automated Optical Inspection (AOI) equipment, and the generator , the SPI numerical data and the determination results for each AOI defective item can be linked to each other through a barcode so that the AOI defective determination result is referenced as a label of the SPI numerical data.

Specifically, the heterogeneous inspection data includes AOI numerical data, which is inspection data of Automated Optical Inspection (AOI) equipment, and determination results for each AOI defect item, and the generator generates the AOI defect determination result as the AOI numerical data. The AOI numerical data and the determination results for each AOI defect item can be linked to each other through a barcode so that they are referred to as labels.

Specifically, the prediction unit may predict the AOI defect item from the SPI numerical data through the defect prediction model.

Specifically, the prediction unit may predict the AOI defect item from the AOI numerical data through the defect prediction model.

A method of operating an SMD-based defect prediction device according to an embodiment of the present invention to achieve the above object includes inspection from at least one inspection device that checks the status of a PCB (Printed Circuit Board) in an SMT (Surface Mounting Technology) line. An acquisition step of acquiring data; A generation step of generating a learning data set by preprocessing the inspection data; And a learning step of learning a defect prediction model for predicting PCB defect items based on the training data set.

Specifically, the method may further include a prediction step of predicting PCB defect items that occur in a subsequent process after checking the condition of the PCB through the defect prediction model.

Specifically, the acquisition step acquires heterogeneous inspection data from the at least one inspection equipment, and the generation step may generate a learning data set through preprocessing that links the heterogeneous inspection data with each other.

Specifically, the heterogeneous inspection data includes SPI numerical data, which is inspection data of SPI (Solder Paste Inspection) equipment, and judgment results for each AOI defect item, which is inspection data of AOI (Automated Optical Inspection) equipment, and the generation step. The SPI numeric data and the determination result for each AOI defect item can be linked to each other through a barcode so that the AOI defect determination result is referenced as a label of the SPI numeric data.

Specifically, the heterogeneous inspection data includes AOI numerical data, which is inspection data of Automated Optical Inspection (AOI) equipment, and judgment results for each AOI defect item, and in the generation step, the AOI defect determination result is the AOI value. To be referred to as a data label, the AOI numerical data and the determination results for each AOI defect item can be linked to each other through a barcode.

Specifically, in the prediction step, the AOI defect item can be predicted from the SPI numerical data through the defect prediction model.

Specifically, in the prediction step, the AOI defect item can be predicted from the AOI numerical data through the defect prediction model.

Accordingly, in the SMD-based defect prediction device and its operating method of the present invention, PCB (Printed Circuit Board) defects that occur during the SMT (Surface Mounter Technology) line process are predicted before proceeding with the subsequent process, and corresponding measures are taken. Through this, quality costs can be reduced, and defective items can be alarmed in advance through PCB defect prediction based on inspection data, greatly reducing the probability of defective products due to equipment failure.

1 is an exemplary diagram of a Surface Mounter Technology (SMT) line according to an embodiment of the present invention.
Figure 2 is an exemplary diagram illustrating a defect prediction environment according to an embodiment of the present invention.
Figure 3 is a diagram illustrating the configuration of a defect prediction device according to an embodiment of the present invention.
Figures 4 and 5 are exemplary diagrams for explaining a learning data set according to an embodiment of the present invention.
Figure 6 is an example diagram for explaining a defect prediction model according to an embodiment of the present invention.
7 is a flowchart illustrating a method of operating a defect prediction device according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described with reference to the attached drawings.

In one embodiment of the present invention, technology related to predicting PCB (Printed Circuit Board) defects in the SMT (Surface Mounter Technology) line is addressed.

Generally, in an SMT line, for example, as shown in FIG. 1, solder paste is printed on a PCB (Printed Circuit Board) through a screen printer, and a PCB on which the solder paste has been printed is completed through a mounter (Chip & Flexible Mounter). The process of mounting the actual components takes place.

In this regard, in the SMT line, inspection equipment can be used to check the defective state of the PCB.

In other words, SPI (Solder Paste Inspection) equipment (3D SPI) checks the condition of lead applied to the PCB through a screen printer, and AOI (Automated Optical Inspection) equipment (3D AOI) checks reflow (N2). After joining components through reflow, the captured images can be used to check whether the components on the PCB are properly mounted, including displacement, distortion, and tilt.

However, as seen above, such PCB defect status inspection can be performed during the process or after the process is completed. In particular, in the case of defect status inspection performed after the process is completed, it is only performed after product production, which may result in an increase in quality costs. , it has the limitation that it is difficult to check the possibility of equipment failure.

Accordingly, in one embodiment of the present invention, we would like to propose a new method for predicting PCB defects during the SMT line process before proceeding with the subsequent process.

In this regard, Figure 2 exemplarily shows a PCB defect prediction environment according to an embodiment of the present invention.

As shown in FIG. 2, the PCB defect prediction environment according to an embodiment of the present invention may include a defect prediction device 100 that predicts PCB defects during the SMT line processing process.

Here, the defect prediction device 100 refers to a device that learns a defect prediction model from inspection data that checks the state of the PCB and predicts PCB defect items through the learned defect prediction model.

This defect prediction device 100 is an SPI device (3D SPI) that checks the condition of lead applied to the PCB in the SMT line, and an AOI device (3D AOI) that checks whether the components on the PCB are properly mounted, or It can be implemented as a separate device from SPI equipment (3D SPI) and AOI equipment (3D AOI).

For reference, if the defect prediction device 100 is implemented as a separate device from SPI equipment (3D SPI) and AOI equipment (3D AOI), the defect prediction device 100 may be implemented in the form of a server. The server may be implemented in the form of, for example, a web server, database server, proxy server, etc., and may be installed with one or more of various software that allows a network load balancing mechanism or service device to operate on the Internet or other networks. It can also be implemented as a computerized system.

In the PCB defect prediction environment according to an embodiment of the present invention, through the above-described configuration, PCB defects that may occur due to previous processes during the SMT line process can be predicted before proceeding with the subsequent process. This will be described below. The configuration of the defect prediction device 100 to be implemented will be described in more detail.

Figure 3 shows a schematic configuration of a defect prediction device 100 according to an embodiment of the present invention.

As shown in FIG. 3, the defect prediction device 100 according to an embodiment of the present invention includes an acquisition unit 110 for obtaining inspection data, a generation unit 120 for generating a learning data set, and a defect prediction model. It may have a configuration including a learning unit 130 that learns.

Additionally, the defect prediction device 100 according to an embodiment of the present invention may further include a prediction unit 140 that predicts defective items in addition to the above-described configuration.

The entire configuration or at least a portion of the defect prediction device 100 including the acquisition unit 110, generation unit 120, learning unit 130, and prediction unit 140 is in the form of a hardware module or a software module. It can be implemented or implemented as a combination of hardware modules and software modules.

Here, the software module can be understood as, for example, an instruction executed by a processor that controls operations within the defect prediction device 100, and these instructions may be mounted in a memory within the defect prediction device 100. There will be.

Through the above-described configuration, the defect prediction device 100 according to an embodiment of the present invention can predict PCB defects that may occur due to previous processes during the SMT line process before proceeding with the subsequent process. Hereinafter, We will continue with a more detailed explanation of the configuration within the defect prediction device 100 to realize this.

The acquisition unit 110 performs the function of acquiring inspection data.

More specifically, the acquisition unit 110 acquires inspection data regarding the status of the PCB from the SMT line.

At this time, the acquisition unit 110 may acquire heterogeneous inspection data from at least one inspection device that inspects the status of the PCB on the SMT line.

Here, at least one inspection equipment from which inspection data is obtained may include SPI equipment (3D SPI), which checks the condition of lead applied to the PCB, and AOI equipment (3D AOI), which checks whether the components on the PCB are properly mounted. there is.

In addition, the heterogeneous inspection data obtained from such inspection equipment includes SPI numerical data, which is inspection data of SPI equipment (3D SPI), AOI numerical data, which is inspection data of AOI equipment (3D AOI), and judgment results for each AOI defect item. may be included.

Meanwhile, heterogeneous inspection data obtained from SPI equipment (3D SPI) and AOI equipment (3D AOI) may be stored, for example, in the form of a CSV (Comma-Separated Values) file.

In addition, as will be explained in detail below, the SPI numerical data, which is the inspection data of the SPI equipment (3D SPI), and the judgment results for each AOI defect item, which is the inspection data of the AOI equipment (3D AOI), are due to the condition of the lead applied to the PCB. It is used to learn a defect prediction model (SPI-AI model) to predict AOI defect items, and the AOI numerical data, which is the inspection data of the AOI equipment (3D AOI), and the judgment results for each AOI defect item are used to learn the defect prediction model (SPI-AI model) to predict AOI defect items. ) It is used to learn a defect prediction model (AOI-AI model) to predict AOI defect items caused by itself.

The generator 120 performs the function of generating a learning data set.

More specifically, when inspection data regarding the state of the PCB is obtained from the SMT line, the generator 120 generates a learning data set by preprocessing the obtained inspection data.

At this time, the generator 120 may generate learning data through preprocessing to link heterogeneous inspection data.

In this regard, the generator 120 is a preprocessing of a learning data set that supports prediction of AOI defect items resulting from the state of lead applied to the PCB, and, for example, determines each SPI numerical data and AOI defect item as shown in FIG. 4. The results are linked to each other through a barcode so that the AOI defect determination result can be referenced as a label for SPI numerical data.

In addition, the generator 120 is a preprocessing of a learning data set that supports prediction of AOI defective items resulting from the AOI equipment (3D AOI) itself. For example, as shown in Figure 5, AOI numerical data and judgment results for each AOI defective item. are linked together through a barcode so that the AOI defect determination result can be referred to as a label of the AOI numerical data.

The learning unit 130 performs a function of learning a defect prediction model.

More specifically, when a learning data set obtained by preprocessing inspection data is generated, the learning unit 130 learns a defect prediction model for predicting PCB defective items based on the generated training data set.

At this time, the learning unit 130 creates a defect prediction model (SPI-AI model) for predicting AOI defect items resulting from the state of lead applied to the PCB from a learning data set that links the SPI numerical data and the judgment results for each AOI defect item. ) can be learned.

Here, the AOI defect items resulting from the state of the lead applied to the PCB may include, for example, Bridge-Lead, Overhang-Lead, and Coplanarity-Lead items, as shown in FIG. 6.

In this regard, for example, One-Side Selection (OSS) and Support Vector Machine (SVM) are used to learn a defect prediction model (SPI-AI model) for predicting Bridge-Lead items according to an embodiment of the present invention. For example, Balanced Bagging Classifier can be used to learn a defect prediction model (SPI-AI model) to predict Overhang-Lead and Coplanarity-Lead items.

In addition, the learning unit 130 creates a defect prediction model (AOI-AI) for predicting AOI defect items resulting from the AOI equipment (3D AOI) itself from a learning data set that links API numerical data and judgment results for each AOI defect item. model) can be learned.

Here, the AOI defect items resulting from the state of the lead applied to the PCB may include Coplanarity-Body and Dimension-Body items, as shown in FIG. 6 previously illustrated.

In this regard, for example, OneClass SVM can be used to learn a defect prediction model (AOI-AI model) for predicting Coplanarity-Body and Dimension-Body items according to an embodiment of the present invention.

The prediction unit 140 performs a function of predicting defective items.

More specifically, when learning of the defect prediction model is completed, the prediction unit 140 predicts PCB defect items that occur in the subsequent process after checking the status of the PCB through the learned defect prediction model and provides an alarm of the prediction result. do.

At this time, the prediction unit 140 predicts AOI defect items (Bridge-Lead, Overhang- Lead, and Coplanarity-Lead) can be predicted.

In addition, the prediction unit 140 uses a defect prediction model (AOI-AI model) to predict AOI defect items resulting from the AOI equipment (3D AOI) itself, and uses AOI numerical data to predict AOI defect items (Coplanarity-Body, and Dimension-Body) can be predicted.

Meanwhile, as described above, when a defective item is predicted through the defect prediction model, the prediction unit 140 searches the component information (board type, defective item, and part location) for which the defective item is predicted from the self-sampling data DB and stores the defective item in the facility DB. The equipment that caused the defective item can be searched, and each search result along with the cause of the defect searched from the defect cause DB can be alerted (displayed) to the user.

As seen above, according to the configuration of the defect prediction device 100 according to an embodiment of the present invention, defects in the PCB are predicted during the SMT line process before proceeding with the post-process, and measures are taken accordingly. Quality costs can be reduced, and defective items can be alerted in advance through PCB defect prediction based on inspection data, thereby greatly reducing the probability of defective products due to equipment failure.

Hereinafter, a method of operating the defect prediction device 100 according to an embodiment of the present invention will be described with reference to FIG. 7.

First, the acquisition unit 110 acquires inspection data regarding the status of the PCB from the SMT line (S110-S120).

At this time, the acquisition unit 110 may acquire heterogeneous inspection data from at least one inspection device that inspects the status of the PCB on the SMT line.

Here, at least one inspection equipment from which inspection data is obtained may include SPI equipment (3D SPI), which checks the condition of lead applied to the PCB, and AOI equipment (3D AOI), which checks whether the components on the PCB are properly mounted. there is.

In addition, the heterogeneous inspection data obtained from such inspection equipment includes SPI numerical data, which is inspection data of SPI equipment (3D SPI), AOI numerical data, which is inspection data of AOI equipment (3D AOI), and judgment results for each AOI defect item. may be included.

Meanwhile, heterogeneous inspection data obtained from SPI equipment (3D SPI) and AOI equipment (3D AOI) may be stored, for example, in the form of a CSV (Comma-Separated Values) file.

In addition, as will be explained in detail below, the SPI numerical data, which is the inspection data of the SPI equipment (3D SPI), and the judgment results for each AOI defect item, which is the inspection data of the AOI equipment (3D AOI), are due to the condition of the lead applied to the PCB. It is used to learn a defect prediction model (SPI-AI model) to predict AOI defect items, and the AOI numerical data, which is the inspection data of the AOI equipment (3D AOI), and the judgment results for each AOI defect item are used to learn the defect prediction model (SPI-AI model) to predict AOI defect items. ) It is used to learn a defect prediction model (AOI-AI model) to predict AOI defect items caused by itself.

Then, when inspection data regarding the state of the PCB is obtained from the SMT line, the generator 120 generates a learning data set by preprocessing the obtained inspection data (S130).

At this time, the generator 120 may generate learning data through preprocessing to link heterogeneous inspection data.

In this regard, the generator 120 is a preprocessing of a learning data set that supports prediction of AOI defect items resulting from the state of lead applied to the PCB, and, for example, determines each SPI numerical data and AOI defect item as shown in FIG. 4. The results are linked to each other through a barcode so that the AOI defect determination result can be referenced as a label for SPI numerical data.

In addition, the generator 120 is a preprocessing of a learning data set that supports prediction of AOI defective items resulting from the AOI equipment (3D AOI) itself. For example, as shown in Figure 5, AOI numerical data and judgment results for each AOI defective item. are linked together through a barcode so that the AOI defect determination result can be referred to as a label of the AOI numerical data.

Furthermore, when the learning data set obtained by preprocessing the inspection data is generated, the learning unit 130 learns a defect prediction model for predicting PCB defective items based on the generated learning data set (S140).

At this time, the learning unit 130 creates a defect prediction model (SPI-AI model) for predicting AOI defect items resulting from the state of lead applied to the PCB from a learning data set that links the SPI numerical data and the judgment results for each AOI defect item. ) can be learned.

Here, the AOI defect items resulting from the state of the lead applied to the PCB may include Bridge-Lead, Overhang-Lead, and Coplanarity-Lead items, as shown in FIG. 6, as illustrated above.

In this regard, for example, One-Side Selection (OSS) and Support Vector Machine (SVM) are used to learn a defect prediction model (SPI-AI model) for predicting Bridge-Lead items according to an embodiment of the present invention. For example, Balanced Bagging Classifier can be used to learn a defect prediction model (SPI-AI model) to predict Overhang-Lead and Coplanarity-Lead items.

In addition, the learning unit 130 creates a defect prediction model (AOI-AI) for predicting AOI defect items resulting from the AOI equipment (3D AOI) itself from a learning data set that links API numerical data and judgment results for each AOI defect item. model) can be learned.

Here, the AOI defect items resulting from the state of the lead applied to the PCB may include Coplanarity-Body and Dimension-Body items, as shown in FIG. 6 previously illustrated.

In this regard, for example, OneClass SVM can be used to learn a defect prediction model (AOI-AI model) for predicting Coplanarity-Body and Dimension-Body items according to an embodiment of the present invention.

Afterwards, when learning of the defect prediction model is completed, the prediction unit 140 predicts PCB defect items that occur in the subsequent process after checking the status of the PCB through the learned defect prediction model, and notifies the user of the prediction result. (S150-S170).

At this time, the prediction unit 140 predicts AOI defect items (Bridge-Lead, Overhang- Lead, and Coplanarity-Lead) can be predicted.

In addition, the prediction unit 140 uses a defect prediction model (AOI-AI model) to predict AOI defect items resulting from the AOI equipment (3D AOI) itself, and uses AOI numerical data to predict AOI defect items (Coplanarity-Body, and Dimension-Body) can be predicted.

Meanwhile, as described above, when a defective item is predicted through the defect prediction model, the prediction unit 140 searches the component information (board type, defective item, and part location) for which the defective item is predicted from the self-sampling data DB and stores the defective item in the facility DB. The equipment that caused the defective item can be searched, and each search result along with the cause of the defect searched from the defect cause DB can be alerted (displayed) to the user.

As discussed above, according to the operating method of the defect prediction device 100 according to an embodiment of the present invention, defects in the PCB are predicted during the SMT line process before proceeding with the subsequent process, and corresponding measures are taken. Through this, quality costs can be reduced, and defective items can be alarmed in advance through PCB defect prediction based on inspection data, thereby significantly reducing the probability of defective products due to equipment failure.

Meanwhile, implementations of the functional operations and subjects described in this specification are implemented as digital electronic circuits, computer software, firmware, or hardware including the structure disclosed in this specification and its structural equivalents, or one or more of these. It can be implemented by combining. Implementations of the subject matter described herein may comprise one or more computer program products, that is, one or more modules of computer program instructions encoded on a tangible program storage medium for processing or execution by the operation of a processing system. It can be implemented.

The computer-readable medium may be a machine-readable storage device, a machine-readable storage substrate, a memory device, or a combination of one or more of these.

In this specification, “system” or “device” includes all instruments, devices, and machines for processing data, including, for example, programmable processors, computers, or multiple processors or computers. In addition to the hardware, the processing system may include code that forms an execution environment for computer programs on demand, such as code making up processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of these. .

A computer program (also known as a program, software, software application, script, or code) may be written in any form of a programming language, including compiled, interpreted, a priori, or procedural languages; as a stand-alone program or module; It can be deployed in any form, including components, subroutines, or other units suitable for use in a computer environment. Computer programs do not necessarily correspond to files in a file system. A program may be stored within a single file that serves the requested program, or within multiple interacting files (e.g., files storing one or more modules, subprograms, or portions of code), or as part of a file that holds other programs or data. (e.g., one or more scripts stored within a markup language document). The computer program may be deployed to run on a single computer or multiple computers located at one site or distributed across multiple sites and interconnected by a communications network.

Meanwhile, computer-readable media suitable for storing computer program instructions and data include semiconductor memory devices such as EPROM, EEPROM, and flash memory devices, magnetic disks such as internal hard disks and external disks, magneto-optical disks, and CDs. -Can include all forms of non-volatile memory, media, and memory devices, including ROM and DVD-ROM disks. The processor and memory may be supplemented by, or integrated into, special-purpose logic circuitry.

Implementations of the subject matter described herein may include backend components, such as a data server, middleware components, such as an application server, or, such as a web browser or graphical user, through which a user may interact with an implementation of the subject matter described herein. It may be implemented in a front-end component, such as a client computer with an interface, or in a computing system that includes any combination of one or more of such back-end, middleware, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication, such as a telecommunications network.

Although this specification contains details of numerous specific implementations, these should not be construed as limitations on the scope of any invention or what may be claimed, but rather as descriptions of features that may be unique to particular embodiments of particular inventions. It must be understood. Likewise, certain features described herein in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination. Furthermore, although features may be described as operating in a particular combination and initially claimed as such, one or more features from a claimed combination may in some cases be excluded from that combination, and the claimed combination may be a sub-combination. It can be changed to a variant of a sub-combination.

Additionally, although operations are depicted in the drawings in a specific order herein, this should not be understood to mean that such operations must be performed in the specific order or sequential order shown or that all illustrated operations must be performed to obtain desirable results. Can not be done. In certain cases, multitasking and parallel processing may be advantageous. Additionally, the separation of various system components in the above-described embodiments should not be construed as requiring such separation in all embodiments, and the described program components and systems may generally be integrated together into a single software product or packaged into multiple software products. You must understand that you can

As such, this specification is not intended to limit the invention to the specific terms presented. Accordingly, although the present invention has been described in detail with reference to the above-described examples, those skilled in the art may make modifications, changes, and variations to the examples without departing from the scope of the present invention. The scope of the present invention is indicated by the claims described below rather than the detailed description above, and all changes or modified forms derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

According to the SMD-based defect prediction device and its operating method according to the present invention, it is possible to predict PCB (Printed Circuit Board) defects that occur during the SMT (Surface Mounter Technology) line process before proceeding with the post-process. It is an invention that has industrial applicability because it overcomes the limitations of technology, not only has the possibility of using related technology, but also markets or sells the applied device, and can clearly be implemented in reality.

100: Defect prediction device
110: acquisition unit 120: creation unit
130: learning unit 140: prediction unit

Claims (14)

An acquisition unit that acquires inspection data from at least one inspection device that checks the status of a PCB (Printed Circuit Board) in an SMT (Surface Mounting Technology) line;
A generator that generates a learning data set by preprocessing the inspection data; and
An SMD-based defect prediction device comprising a learning unit that learns a defect prediction model for predicting PCB defect items based on the learning data set. According to claim 1,
The device is,
An SMD-based defect prediction device further comprising a prediction unit that predicts PCB defect items that occur in a subsequent process after checking the condition of the PCB through the defect prediction model. According to claim 1,
The acquisition department,
Obtaining heterogeneous inspection data from the at least one inspection equipment,
The generating unit,
An SMD-based defect prediction device characterized in that a learning data set is generated through preprocessing that links the heterogeneous inspection data. According to claim 3,
The above heterogeneous inspection data is,
It includes SPI numerical data, which is inspection data of SPI (Solder Paste Inspection) equipment, and judgment results for each AOI defect item, which is inspection data of AOI (Automated Optical Inspection) equipment.
The generating unit,
An SMD-based defect prediction device characterized in that the SPI numeric data and the determination result for each AOI defect item are linked to each other through a barcode so that the AOI defect determination result is referred to as a label of the SPI numeric data. . According to claim 3,
The above heterogeneous inspection data is,
It includes AOI numerical data, which is inspection data of AOI (Automated Optical Inspection) equipment, and judgment results for each AOI defect item.
The generating unit,
An SMD-based defect prediction device characterized in that the AOI numeric data and the determination result for each AOI defect item are linked to each other through a barcode so that the AOI defect determination result is referred to as a label of the AOI numeric data. . According to claim 4,
The prediction unit,
An SMD-based defect prediction device, characterized in that predicting the AOI defect item from the SPI numerical data through the defect prediction model. According to claim 5,
The prediction unit,
An SMD-based defect prediction device, characterized in that predicting the AOI defect items from the AOI numerical data through the defect prediction model. An acquisition step of acquiring inspection data from at least one inspection device that checks the status of a PCB (Printed Circuit Board) in an SMT (Surface Mounting Technology) line;
A generation step of generating a learning data set by preprocessing the inspection data; and
A method of operating an SMD-based defect prediction device, comprising a learning step of learning a defect prediction model for predicting PCB defect items based on the training data set. According to claim 8,
The above method is,
A method of operating an SMD-based defect prediction device, further comprising a prediction step of predicting PCB defect items that occur in a subsequent process after checking the condition of the PCB through the defect prediction model. According to claim 8,
The acquisition step is,
Obtaining heterogeneous inspection data from the at least one inspection equipment,
The creation step is,
A method of operating an SMD-based defect prediction device, characterized in that a learning data set is generated through preprocessing that links the heterogeneous inspection data with each other. According to claim 10,
The above heterogeneous inspection data is,
It includes SPI numerical data, which is inspection data of SPI (Solder Paste Inspection) equipment, and judgment results for each AOI defect item, which is inspection data of AOI (Automated Optical Inspection) equipment.
The creation step is,
An SMD-based defect prediction device characterized in that the SPI numeric data and the determination result for each AOI defect item are linked to each other through a barcode so that the AOI defect determination result is referred to as a label of the SPI numeric data. How it works. According to claim 10,
The above heterogeneous inspection data is,
It includes AOI numerical data, which is inspection data of AOI (Automated Optical Inspection) equipment, and judgment results for each AOI defect item.
The creation step is,
An SMD-based defect prediction device characterized in that the AOI numeric data and the determination result for each AOI defect item are linked to each other through a barcode so that the AOI defect determination result is referred to as a label of the AOI numeric data. How it works. According to claim 11,
The prediction step is,
A method of operating an SMD-based defect prediction device, characterized in that predicting the AOI defect item from the SPI numerical data through the defect prediction model. According to claim 12,
The prediction step is,
A method of operating an SMD-based defect prediction device, characterized in that predicting the AOI defect item from the AOI numerical data through the defect prediction model.