JP7472472B2 - Rework condition learning device, rework condition determination device, and rework device - Google Patents

Description

This invention relates to a rework condition learning device that learns the setting values that are the conditions for reworking an object to be reworked, and a rework condition determination device and rework device that use the same.

Conventionally, rework devices (repair devices) that reprocess objects include devices that heat solder to remove electronic components from the substrate to be reprocessed (see, for example, Patent Documents 1 and 2). Furthermore, some rework devices (repair devices) attach replacement electronic components after removing the electronic components from the substrate (see, for example, Patent Document 3).

On the other hand, there are those that apply learning models based on machine learning using AI (Artificial Intelligence) to rework (repair) equipment (see, for example, Patent Document 4). Patent Document 4 discloses that a learning model (sequential learning) is used to determine whether a product is good or defective, and that the rework (repair) equipment displays the defective location and details of the defect, which are the results of the determination.

JP 2018-148032 A JP 2005-223000 A JP 2006-179556 A JP 2008-185395 A

However, conventional rework equipment requires the input of settings for each rework target, which are the conditions for reworking, such as removing or installing the target part, and there is an issue that the yield is affected by the skill of the person entering the settings.

This invention has been made to solve the above problems, and aims to provide a rework condition learning device that learns, for each rework object, the setting values that are the conditions for reworking the rework object, and a rework condition determination device and rework device that use the same.

The rework condition learning device according to the present invention, and the rework condition determination device and rework device using the same, are characterized in that the rework condition learning device learns setting values that are conditions for reworking a rework object by removing or attaching a target part, and is equipped with a setting value input unit to which the setting value associated with the rework object is input for each device performing reworking, a quality judgment result input unit to which the quality judgment result of the rework object reworked using the setting value input to the setting value input unit is input, and a learning unit that learns the setting value that will result in a quality product for each rework object based on the setting value and the quality judgment result.

The rework condition learning device according to the present invention is a rework condition learning device that learns a setting value that is a removal condition for removing a target part from a rework object in which a target part is attached to a processing object by soldering, and is characterized in that it is equipped with a setting value input unit to which the setting value associated with the rework object is input for each device that performs reworking, a quality judgment result input unit to which a quality judgment result of a quality or defective product of the rework object reworked by the setting value input to the setting value input unit is input, and a learning unit that learns the setting value that will result in a quality product for each rework object based on the setting value and the quality judgment result.

The rework condition learning device according to the present invention is a rework condition learning device that learns a setting value that is an attachment condition for attaching a new target part to a rework object from which a target part has been removed, and is characterized in that it is equipped with a setting value input unit to which the setting value associated with the rework object is input for each device that performs reworking, a quality judgment result input unit to which a quality judgment result of a quality or defective product of the rework object reworked using the setting value input to the setting value input unit is input, and a learning unit that learns the setting value that will result in a quality product for each rework object based on the setting value and the quality judgment result.

The rework condition determination device according to the present invention is a rework condition learning device that learns a setting value that is a removal condition for removing a rework object in which a target part is attached to a processing object by soldering, and is a rework condition determination device that uses the learning results of the rework condition learning device, which includes a setting value input unit to which the setting value associated with the rework object is input for each device that performs reworking, a quality judgment result input unit to which a quality judgment result of the rework object reworked by the setting value input to the setting value input unit is input, and a learning unit that learns the setting value that will be a quality product for each rework object based on the setting value and the quality judgment result, and is characterized in that it includes a rework object input unit to which information of the rework object is input, and a rework condition determination unit that uses the learning result of the learning unit to determine the setting value that will make the rework object, whose information is input to the rework object input unit, a quality product.

The rework condition determination device according to the present invention is a rework condition learning device that learns a setting value that is an attachment condition for attaching a new target part to a rework target from which a target part has been removed, and includes a setting value input unit to which the setting value associated with the rework target is input for each device that performs reworking, a quality judgment result input unit to which a quality judgment result for the rework target reworked using the setting value input to the setting value input unit is input, and a learning unit that learns the setting value that will result in a quality product for each rework target based on the setting value and the quality judgment result, and is characterized in that the rework condition determination device uses the learning results of the rework condition learning device and includes a rework target input unit to which information about the rework target is input, and a rework condition determination unit that uses the learning results of the learning unit to determine the setting value that will result in a quality product for the rework target whose information has been input to the rework target input unit.

The rework device according to the present invention is a rework condition learning device that learns a setting value, which is a removal condition for removing a target part from a rework object in which a target part has been attached to a processing object by soldering, and is equipped with a setting value input unit to which the setting value associated with the rework object is input for each device that performs reworking, a quality judgment result input unit to which a quality judgment result of a quality or defective product of the rework object reworked by the setting value input to the setting value input unit is input, and a learning unit that learns the setting value that will result in a quality product for each rework object based on the setting value and the quality judgment result. A rework condition determination device using the learning results of a condition learning device, the rework device using a rework condition determination device having a rework object input unit to which information of the rework object is input, and a rework condition determination unit that uses the learning results of the learning unit to determine the setting value that makes the rework object whose information is input to the rework object input unit a non-defective product, the rework device is characterized in that it has a removal unit of the device, a condition setting unit to which the setting value determined by the rework condition determination unit is input, and a removal control unit that controls the removal unit to remove the target part under the conditions input to the condition setting unit.

The rework device according to the present invention is a rework condition learning device that learns a setting value that is an attachment condition for attaching a new target part to a rework object from which a target part has been removed, and is equipped with a setting value input unit to which the setting value associated with the rework object is input for each device that performs reworking, a quality judgment result input unit to which a quality judgment result of a quality or defective product of the rework object reworked by the setting value input to the setting value input unit is input, and a learning unit that learns the setting value that will result in a quality product for each rework object based on the setting value and the quality judgment result. A rework condition determination device using the results of the learning is provided with a rework object input unit to which information on the rework object is input, and a rework condition determination unit that uses the learning results of the learning unit to determine the setting value that will make the rework object whose information has been input to the rework object input unit a non-defective product. The rework device is characterized in that it is provided with a soldering unit of the device, a condition setting unit to which the setting value determined by the rework condition determination unit is input, and a soldering control unit that controls the soldering unit to solder the new target part under the conditions input to the condition setting unit.

This invention provides a rework condition learning device capable of learning the set values that result in a good product for each rework object, as well as a rework condition determination device and a rework device that use the same.

1 is a functional block diagram of a rework condition learning device according to a first embodiment of the present invention; 4 is a flowchart illustrating the operation of the rework condition learning device (rework condition learning method) according to the first embodiment of the present invention. 1 is a functional block diagram of a rework condition learning device, a rework condition determining device, and a rework device according to a first embodiment of the present invention; 4 is a flowchart illustrating an operation of the rework condition determining apparatus (rework condition determining method) according to the first embodiment of the present invention. 4 is a flowchart illustrating the operation of the rework apparatus (rework method) according to the first embodiment of the present invention. 4 is a flowchart illustrating the operation of the rework apparatus (rework method) according to the first embodiment of the present invention. FIG. 11 is a functional block diagram of a rework condition learning device according to a second embodiment of the present invention. 11 is a flowchart illustrating the operation of a rework condition learning device (rework condition learning method) according to a second embodiment of the present invention. FIG. 11 is a functional block diagram of a rework condition learning device according to a third embodiment of the present invention. 13 is a flowchart illustrating the operation (good/bad method) of the rework condition learning device according to the third embodiment of the present invention. FIG. 11 is a functional block diagram of a rework condition learning device, a rework condition determining device, and a rework device according to a third embodiment of the present invention. FIG. 11 is a functional block diagram of a rework condition learning device according to a fourth embodiment of the present invention. 13 is a flowchart illustrating the operation of a rework condition learning device (rework condition learning method) according to a fourth embodiment of the present invention. FIG. 11 is a functional block diagram of a rework condition learning device, a rework condition determining device, and a rework device according to a fourth embodiment of the present invention. FIG. 11 is a functional block diagram of a rework condition learning device, a rework condition determining device, and a rework device according to a fourth embodiment of the present invention. 13 is a flowchart illustrating the operation of a rework condition determining apparatus (rework condition determining method) according to a fourth embodiment of the present invention. 13 is a flowchart illustrating the operation of the rework apparatus (rework method) according to the fourth embodiment of the present invention. 13 is a flowchart illustrating the operation of the rework apparatus (rework method) according to the fourth embodiment of the present invention. FIG. 13 is a functional block diagram of a rework condition learning device according to a fifth embodiment of the present invention. 13 is a flowchart illustrating the operation of a rework condition learning device (rework condition learning method) according to a fifth embodiment of the present invention. FIG. 13 is a functional block diagram of a rework condition learning device according to a sixth embodiment of the present invention. 13 is a flowchart for explaining the operation (good/bad method) of the rework condition learning apparatus according to the sixth embodiment of the present invention. FIG. 13 is a functional block diagram of a rework condition learning device, a rework condition determining device, and a rework device according to a sixth embodiment of the present invention. FIG. 13 is a functional block diagram of a rework condition learning device, a rework condition determining device, and a rework device according to a sixth embodiment of the present invention.

This application relates to a rework condition learning device that learns setting values that are conditions for reworking to remove or attach a target part to a rework object 1, a rework condition determination device and a rework device that use the same (a rework condition learning method, a rework condition determination method and a rework method that use the same). The setting values that are conditions for reworking are a setting value that is a removal condition for removing a target part from a rework object 1, and a setting value that is an attachment condition for attaching a new target part to a rework object from which a target part has been removed. In detail, the former is a setting value that is a removal condition for removing a target part from a rework object 1 to which a target part has been attached by soldering to a processing object. The latter is a setting value that is an attachment condition for attaching a new target part to a rework object 1 from which a soldered target part has been removed.

The former, a rework condition learning device that learns the setting values that are the removal conditions for removing a target component from a rework target 1 in which the target component is attached to the processing target by soldering, a rework condition determination device and a rework device that use the same (a rework condition learning method, a rework condition determination method and a rework method that use the same) will be explained in the first to third embodiments.

The latter, a rework condition learning device that learns the setting values that are the attachment conditions for attaching a new target component to a rework target 1 from which a soldered target component has been removed, and a rework condition determination device and rework device using the same (a rework condition learning method, a rework condition determination method and a rework method using the same) will be described in embodiments 4 to 6. Note that embodiments 4 to 6 include not only cases in which only the setting values that are the attachment conditions are learned, but also cases in which the setting values that are the removal conditions, which corresponds to the former, are learned.

In other words, a rework device is a device that performs reprocessing (rework). Reprocessing (rework) refers to at least one of the following: reprocessing (rework) in which a target part is removed from a reprocessing target 1 in which a target part has been attached to a processing target by soldering, or reprocessing (rework) in which a new target part is attached to a reprocessing target 1 from which the soldered target part has been removed.

Embodiment 1.
Hereinafter, the rework condition learning device according to the first embodiment of the present invention, the rework condition determination device using the same, and the rework device using the same will be described with reference to FIGS. 1 to 6. In the drawings, the same reference numerals indicate the same or corresponding parts, and detailed description thereof will be omitted. In FIG. 3, the rework object 1 is, for example, a substrate 1 such as a printed circuit board, and includes a single-layer substrate and a multilayer substrate. A substrate on which wiring and components (target components) are soldered may be called a printed wiring board, a printed wiring substrate, a printed circuit board, or a printed circuit board. A substrate on which wiring is applied may simply be called a printed wiring board or a printed wiring substrate. In this case, it can be said that the rework object 1 (substrate 1) includes a printed wiring board and a printed wiring substrate. The substrate 1 may be called a processing object. Furthermore, it can be said that the rework object 1 (substrate 1) is composed of the first solder information (conditions for the first solder) of the first solder printed on the rework object 1, the target components to be soldered, and the member to which the target components are fixed by soldering.

The target components include electronic components and analog components, and their leads and terminals are fixed to the processing target by soldering through reflow. The materials to which the target components are fixed are the aforementioned printed circuit boards, printed wiring boards, and printed wiring boards, and solder is formed on the surface by printing or the like. The leads and terminals of the target components include those that are inserted into through holes formed in the board 1 and fixed by soldering. In summary, the reprocessing target 1 includes those in which the target components are attached to the processing target, which is the board 1, by soldering.

1 and 3, the rework condition learning device 2 has a set condition input unit 3, a good/bad judgment result input unit 4, and a learning unit 5. In FIG. 3, the rework unit 6 (removal unit 6a in the first embodiment) removes the target part from the rework target 1 in which the target part is attached to the processing target by soldering. The removal unit 6a has, for example, a heating unit (air blowing unit) and a removal unit (suction unit). The heating unit heats the joint part of the target part of the rework target 1 with hot air or the like. The removal unit removes the target part from the rework target 1 by sucking the target part with a suction nozzle or the like. The removal unit 6a may have a cooling unit that cools the rework target 1 heated by the heating unit. The rework condition learning device 2 learns the set value, which is the removal condition for removing the target part from the rework target 1 in which the target part is attached to the processing target by soldering, and constructs a learning model.

1 and 3, the good/bad judgment result input unit 4 inputs the good/bad judgment result of the reprocessing target 1 reprocessed according to the set value (set value as the removal condition) input to the set value input unit 3. The good/bad judgment result initially input to the good/bad judgment result input unit 4 is, for example, a digitalized result of an appearance inspection (including inspection by a person or a machine) of the reprocessing target 1, which is a printed circuit board (printed circuit board) after reprocessing. In other words, the result of the appearance inspection (including inspection by a person or a machine) can be digitalized and associated with the set value (set value input to the set value input unit 3) by which the reprocessing target 1 was processed. The appearance inspection here is an inspection (determination) of the good/bad appearance of the reprocessing target 1 after the target part is removed from the reprocessing target 1 in which the target part is attached to the processing target by soldering. The learning unit 5 learns the set value that will be a good product for each reprocessing target 1 based on the set value and the good/bad judgment result. The learning unit 5 learns the setting values that will result in a pass for each rework object 1 based on the setting values input to the setting condition input unit 3 and the pass/fail judgment result input to the pass/fail judgment result input unit 4.

Machine learning such as AI may be applied to the learning unit 5 (rework condition learning device 2). The learning unit 5 (rework condition learning device 2) constructs and accumulates a learning model. Preferably, the learning unit 5 may learn, for each rework device 10, different setting values that result in a good product even for the same rework object 1, by associating them as errors for each rework device 10. In addition, the setting value input unit 3 inputs, as removal conditions, setting values associated with the rework object 1 defined by a first combination of the first solder information of the first solder formed on the rework object 1, the soldered object part, the member to which the object part is fixed by soldering, the heating conditions, and the cooling conditions.

Here, the heating conditions are conditions for heating the joints of the target components of the rework target 1 by the heating section (air blowing section) of the removal section 6a. For example, the heating temperature, heating time, and hot air speed. In order to prevent damage to electronic components due to thermal shock, a multi-stage step heating profile is often used as the heating temperature profile. By gradually increasing the temperature in several steps up to the maximum heating temperature, the risk of component damage can be reduced. For example, 150°C for 20 seconds and 180°C for 20 seconds can be exemplified up to 230°C. The cooling conditions are conditions for cooling the rework target 1 heated by the heating section of the removal section 6a by the cooling section of the removal section 6a. For example, the cooling temperature, cooling time, and cold air speed. It can be said that the cooling conditions are mainly conditions for cooling the joints of the target components of the rework target 1. Rapid cooling is often performed to stabilize the soldering quality. For example, the temperature drop rate is about 2°C to 4°C/sec. The heating and cooling conditions may also include the temperature and humidity around the rework device 10, or the temperature and humidity of the space (room) in which the rework device 10 is installed.

In this case, the set value input unit 3 may further input a set value associated with the rework target 1 in the first combination, in which the first solder information is distinguished by at least one of the composition of the first solder, the type of flux added to the first solder, the content of flux added to the first solder, and the area of the first solder. Similarly, the set value input unit 3 may input a set value associated with the rework target 1 in the first combination, in which the target parts are distinguished by the external shape or the type of electrode surface treatment. Similarly, the set value input unit 3 may input a set value associated with the rework target 1 in the first combination, in which the members are distinguished by at least one of the thickness, material, number of layers, specific gravity, and type of surface treatment.

In other words, the first set value input unit 3 refers to a first combination of the first solder information, the soldered target component, the material to which the target component is fixed by soldering, the heating conditions, and the cooling conditions, in which the set values are input when the first solder information and the target component are the same and the material conditions, the heating conditions, and the cooling conditions have changed. The second set value input unit 3 refers to a first combination of the first solder information, the soldered target component, the material to which the target component is fixed by soldering, the heating conditions, and the cooling conditions, in which the set values are input when the first solder information and the material are the same and the target component conditions, the heating conditions, and the cooling conditions have changed. Furthermore, the third setting value input section 3 refers to the first combination of the first solder information, the soldered target component, the material to which the target component is fixed by soldering, the heating conditions, and the cooling conditions, in which the setting values are input when the first solder information (first solder conditions), the heating conditions, and the cooling conditions are changed while the target component and material are the same.

Here, we will explain the cases where the conditions of the component (reprocessing target 1 from which the target component is removed) change, and the cases where the conditions of the target component change. As mentioned above, a change in the conditions of the component means that the thickness, material, number of layers, specific gravity, and type of surface treatment of the component change. When the component is a board 1, the thickness means the plate thickness specifications (conductor foil thickness, conductor foil area, and number of layers). When the board 1 is a board, the material means the substrate material of the board and the conductor type of the conductor foil. Furthermore, a change in the conditions of the target component means that the volume, electrode shape, and electrode surface treatment change. The shape and size of the target component can be determined from the volume and electrode shape, so the volume and electrode shape of the target component can be said to be information on the external shape of the target component.

Next, a case where the first solder information (conditions of the first solder) has changed will be described. As described above, the first solder information means a case where the composition of the first solder, the type of flux added to the first solder, the content of the flux added to the first solder, for example, the area of the printed first solder is different. The composition of the first solder may include the type of flux and the content of the flux. In addition, the area of the printed first solder refers to the area of the first solder printed on the surface of the rework object 1. The area of the printed first solder also means the amount of the printed first solder, and considering the thickness of the first solder, it can also be said to be the volume of the printed first solder. Furthermore, the information (conditions) on the area of the printed first solder (printed amount of the printed first solder) may include information on the position of the printed first solder on the surface of the rework target 1, or the information (conditions) on the area of the printed first solder (printed amount of the printed first solder) may be information on the position of the printed first solder on the surface of the rework target 1 itself.

So far, the case where two of the first combinations of the first solder information, the target part to be soldered, the member to which the target part is fixed by soldering, the heating conditions, and the cooling conditions have been described, but all combinations may be covered, or one of the first combinations of the first solder information, the target part to be soldered, the member to which the target part is fixed by soldering, the heating conditions, and the cooling conditions may be fixed. In other words, the set value input unit 3 can be said to input set values (set values as removal conditions) associated with the rework target 1 defined by the first combination of at least one of the composition of the first solder printed on the rework target 1, the type of flux added to the first solder, the content of flux added to the first solder, the area of the first solder, either the outer shape of the target part to be soldered or the electrode surface treatment, at least one of the thickness, material, number of layers, specific gravity, and type of surface treatment of the member to which the target part is fixed by soldering, and at least one of the heating conditions or cooling conditions.

This includes inputting the above-mentioned setting values associated with the rework object 1 defined by a first combination of at least one of the composition of the solder printed on the rework object 1, the type of flux added to the solder, the content of the flux added to the solder, and the area of the solder, either the outer shape of the part to be soldered or the electrode surface treatment (at least one of the volume of the part to be soldered, the electrode shape, and the electrode surface treatment), at least one of the thickness, material, number of layers, specific gravity, and type of surface treatment of the member to which the part is fixed by soldering, and at least one of the heating conditions or cooling conditions. This makes it possible to greatly increase the number of setting values that can be handled. The setting values are used to remove the part from the rework object 1, which has the part attached to the processing object by soldering, using the removal unit 6a (rework unit 6).

Next, the operation of the rework condition learning device according to the first embodiment (the rework condition learning method according to the first embodiment) will be described with reference to FIG. 2. In FIG. 2, step 1 is a processing step in which a set value (a set value as a removal condition) associated with the rework object 1 is input to the set value input unit 3 for each device that performs rework (rework) (more specifically, the rework device 10 described later). Step 2 is a processing step in which a good/bad judgment result of the rework object 1 reworked according to the set value (a set value as a removal condition) input to the set value input unit 3 is input to the good/bad judgment result input unit 4. Steps 1 and 2 can be performed in any order. They may be performed simultaneously. Step 3 is a processing step in which the learning unit 5 learns a set value that becomes a good product for each rework object 1 based on the set value input to the set condition input unit 3 and the good/bad judgment result input to the good/bad judgment result input unit 4. Preferably, step 3 may incorporate a processing step in which the learning unit 5 learns that different setting values that result in a non-defective product for the same rework object 1 are associated with errors for each rework device 10.

In the operation of the rework condition learning device according to embodiment 1 (the rework condition learning method according to embodiment 1), similarly to the rework condition learning device according to embodiment 1, step 1 is preferably a step in which a set value associated with rework object 1 defined by a first combination of first solder information, the soldered target component, the material to which the target component is fixed by soldering, the heating conditions, and the cooling conditions is input to the set value input unit 3 as a removal condition. The conditions of the target component, the material conditions, and the combination thereof are also similar to those of the rework condition learning device according to embodiment 1.

In FIG. 3, the rework condition determination device 7 uses the learning results (learning model) of the rework condition learning device 2 shown in FIG. 1 and FIG. 3 (the rework condition determination device according to the first embodiment). The rework condition determination device 7 has a rework object input unit 8 and a rework condition determination unit 9. In FIG. 3, the rework object input unit 8 is used to input information on the rework object 1, which is the object to be newly processed (removed from solder). The rework condition determination unit 9 uses the learning results (learning model) of the learning unit 5 to determine the setting values (setting values as removal conditions) that will make the rework object 1, whose information has been input to the rework object input unit 8, a non-defective product.

Next, the operation of the rework condition determination device according to the first embodiment (the rework condition determination method according to the first embodiment) will be described with reference to FIG. 4. In FIG. 4, step 11 is a processing step in which information on the rework object 1, which is the object to be newly reworked (the object part is to be removed), is input to the rework object input unit 8. Step 12 is a processing step in which the rework condition determination unit 9 uses the learning result (learning model) of the learning unit 5. Step 13 is a processing step in which the rework condition determination unit 9 uses the learning result (learning model) of the learning unit 5 to determine the setting value (removal condition setting value) that makes the rework object 1, whose information has been input to the rework object input unit 8, a non-defective product.

In FIG. 3, the rework device 10 uses the rework condition determination device 7 shown in FIG. 3 (the rework device according to the first embodiment). The rework device 10 has a removal unit 6a (rework unit 6), a condition setting unit 11, and a removal control unit 12a (rework control unit 12). The condition setting unit 11 receives the setting values determined by the rework condition determination unit 7. The removal control unit 12a controls the removal unit 6a so that the target part is removed from the rework object 1 under the conditions input to the condition setting unit 11. The removal control unit 12a controls the removal unit 6a so that the new rework object 1 can be reworked (the target part can be removed) under the removal conditions determined by the setting values.

Next, the operation of the rework device according to the first embodiment (the rework method according to the first embodiment) will be described with reference to FIG. 5. In FIG. 5, step 21 is a processing step in which the set values determined by the rework condition determination unit 7 are input to the condition setting unit 11. Step 24 is a processing step in which removal is performed. More specifically, step 24 is a processing step in which the removal control unit 12a controls the removal unit 6a to remove the target part from the rework object 1 under the conditions input to the condition setting unit 11. In other words, step 24 is a processing step in which the removal unit 6a reworks the new rework object 1 (removes the target part) under the removal conditions determined by the set values.

In the rework condition learning device according to the first embodiment, the rework condition determination device and rework device using the same (the rework condition learning method according to the first embodiment, and the rework condition determination method and rework method using the same), as mentioned in the explanation of the rework condition learning device (rework condition learning method), the learning unit 5 (in step 3) may learn that for each rework device 10, even if the rework object 1 is the same, different setting values that result in a non-defective product are associated as an error for each rework device 10. This error for each rework device 10 can be used in the rework condition determination device and rework device (the rework condition determination method and rework method according to the first embodiment) as follows.

First, in the rework condition determination device (step 12), the learning result of the learning unit 5 is used to determine a set value that will make the rework target 1, whose information has been input to the rework target input unit 8, a non-defective product, and the rework condition determination unit 9 uses the learning result of the learning unit 5 to identify the presence or absence of an error associated with the determined set value. Then, as shown in FIG. 6, in the rework device 10, when there is an error associated with the conditions input to the condition setting unit 11 (processing of step 21) and the condition setting unit 11 has already corrected the error (processing of step 22 is YES), the removal control unit 12a controls the removal unit 6a to remove the soldering (processing of step 23) using the set value that is the reference value for which the error has been corrected (processing of step 24). On the other hand, even if there is an error associated with the conditions input to the condition setting unit 11 (processing of step 21) (processing of step 22 is NO), the removal control unit 12a controls the removal unit 6a to remove the target part under the conditions input to the condition setting unit 11 (processing of step 24).

When proceeding directly from step 22 to step 24, if there is no error associated with the conditions input to the condition setting unit 11 (the processing of step 21), the processing is as shown in the flowchart in FIG. 5. In other words, if the error has not been corrected by the condition setting unit 11, the processing will proceed directly from step 22 (step 21) to step 24 regardless of the presence or absence of an error associated with the conditions input to the condition setting unit 11 (the processing of step 21).

Furthermore, the set value (set value as the removal condition) that becomes the reference value with the error corrected may be prepared individually, or the learning result (learning model) of the learning unit 5 may be used. That is, the learning unit 5 of the rework condition learning device learns the different set values that become good products for the same rework object 1 for each rework device 10 by associating them as errors for each device, and determines the set value that becomes the reference value with the error corrected. Then, when there is an error associated with the conditions input to the condition setting unit 11 and has been corrected by the condition setting unit 11, the removal control unit 12a of the rework device 10 controls the removal unit 6a to remove the target part using the set value that becomes the reference value determined by the learning unit 5 (the set value that becomes the reference value may be obtained via the rework condition determination device 7).

As described above, the rework condition learning device according to the first embodiment, the rework condition determination device and rework device using the same (the rework condition learning method according to the first embodiment, the rework condition determination method and rework method using the same) are capable of learning and using the setting value that will result in a non-defective product for each rework object 1. The setting value is the removal condition for removing the target component from the rework object 1 that has been soldered to the processing object.

Embodiment 2.
A rework condition learning device according to a second embodiment of the present invention, and a rework condition determination device and a rework device using the same will be described below with reference to Figures 7 and 8. The description will focus on the parts that are different from the first embodiment. In the figures, the same reference numerals indicate the same or corresponding parts, and detailed descriptions thereof will be omitted. The major difference between the first and second embodiments is that the learning unit 5 learns the setting values (setting values as removal conditions) that result in a non-defective product for each rework object 1 and the rework object image data (such as a photograph of the area where solder has been removed) in association with each other. There is no major difference between the rework device 10 in the first and second embodiments.

In FIG. 7, the good/bad judgment result input unit 4 inputs the good/bad judgment result of the reprocessing target 1 reprocessed according to the setting value (setting value as the removal condition) input to the setting value input unit 3. Furthermore, the good/bad judgment result input unit 4 inputs the reprocessing target image data (such as a photo of the area where the soldering has been removed) after the target part in the reprocessing target 1 associated with the good/bad judgment result is removed. The learning unit 5 learns the setting value that will result in a good product for each reprocessing target 1 based on the setting value input to the setting condition input unit 3 and the good/bad judgment result input to the good/bad judgment result input unit 4. Furthermore, the learning unit 5 learns the setting value that will result in a good product for each reprocessing target 1 by associating it with the reprocessing target image data.

In this way, the rework condition learning device according to the second embodiment associates the setting value that results in a pass product with the rework target image data for each rework target 1, and therefore can easily identify the rework target image data corresponding to the determined setting value that results in the rework target 1 being a pass product, and can easily obtain an image of the completed product when the rework target 1 is reworked (the target part is removed) with the determined setting value that results in the rework target 1 being a pass product. The rework target image data may be determined by matching the learning result (learning model) of the learning unit 5 with the rework condition determination unit 9 when determining the setting value. The rework condition determination device 7 may be provided with a display unit, including an external unit, that displays the rework target image data obtained (determined) from the learning result (learning model) of the learning unit 5.

Next, the operation of the rework condition learning device according to the second embodiment (the rework condition learning method according to the second embodiment) will be described with reference to FIG. 8. In FIG. 8, step 1 is a processing step in which a setting value (a setting value as a removal condition) associated with the rework object 1 for each device that removes solder (more specifically, a rework device 10 described later) is input to the setting value input unit 3. The order of processing of step 1 and the next step 2A does not matter. They may be performed simultaneously.

8, step 2A is a processing step in which the result of the judgment of the quality of the reprocessing target 1, which has been reprocessed according to the setting value (the setting value as the removal condition) input to the setting value input unit 3, is input to the quality judgment result input unit 4, and further, the image data of the reprocessing target 1 after the target part in the reprocessing target 1 associated with the quality judgment result is removed (such as a photograph of the part where the soldering has been removed) is input to the quality judgment result input unit 4. Step 3A is a processing step in which the learning unit 5 learns the setting value that will be a quality product for each reprocessing target 1 based on the setting value input to the setting condition input unit 3 and the quality judgment result input to the quality judgment result input unit 4, and further, the learning unit 5 learns the setting value that will be a quality product for each reprocessing target 1 in association with the reprocessing target image data. As in step 3 of embodiment 1, step 3A may incorporate a processing step in which the learning unit 5 learns that different setting values for the same rework object 1 that become acceptable products are associated as errors for each rework device 10.

As described above, the rework condition learning device according to the second embodiment, the rework condition determination device and rework device using the same (the rework condition learning method according to the second embodiment, the rework condition determination method and rework method using the same) are capable of learning and using the setting values that result in a non-defective product for each rework object 1. Furthermore, it is also possible to obtain image data of rework objects that have actually been reworked in the past (with the target parts removed).

Embodiment 3.
Hereinafter, a rework condition learning device according to the third embodiment of the present invention, a rework condition determination device using the same, and a rework device using the same will be described with reference to FIG. 9, FIG. 10, and FIG. 11. The description will focus on the parts different from the first and second embodiments. In the drawings, the same reference numerals indicate the same or corresponding parts, and detailed description of them will be omitted. The major difference between the second and third embodiments is that in the second embodiment, the learning unit 5 learns the set value (set value as the removal condition) that becomes a non-defective product for each rework object 1 and the rework object image data (such as a photo of the soldered part) in association with each other, while in the third embodiment, a new rework object image data input unit 13 is further provided. There is no major difference between the first, second, and third embodiments with respect to the rework device 10. There is no major difference between the second and third embodiments with respect to the rework condition determination device 7.

9 and 10, the new reprocessing target image data input unit 13 is used to input new reprocessing target image data of the reprocessing target 1 from which the soldering has been removed, i.e., the target component has been removed. The setting value input unit 3 is used to input setting values associated with the reprocessing target 1 of the new reprocessing target image data. The learning unit 5 uses the learning result (learning model) to determine similar reprocessing target image data (input to the good/bad judgment result input unit 4) from the feature amount of the new reprocessing target image data input to the new reprocessing target image data input unit 13, and judges the good/bad judgment result of the reprocessing target 1 of the new reprocessing target image data from the good/bad judgment result associated with the determined reprocessing target image data. The good/bad judgment result input unit 4 is used to input the good/bad judgment result of the reprocessing target 1 from which the soldering has been removed, determined by the learning unit 5 using the learning result (learning model).

The learning unit 5 can improve classification accuracy by extracting features from the new rework target image data and learning in association with the features of the rework target image data of the rework target 1 from which the soldering has been removed, that is, from which the target parts have been removed. In other words, although not shown, the learning unit 5 has a feature extraction unit (feature extraction function) that extracts features from the new rework target image data and the rework target image data. This feature extraction unit (feature extraction function) may be formed in the setting condition input unit 3. The feature extraction unit (feature extraction function) may be formed in the pass/fail judgment result input unit 4. In this way, machine learning such as AI in the learning unit 5 (rework condition learning device 2) can be said to capture features by learning using a large amount of sample data and its correct answer labels, and make predictions and decisions on new data.

Next, the operation of the rework condition learning device according to the third embodiment (the rework condition learning method according to the third embodiment) will be described with reference to FIG. 10. In FIG. 10, step 31 is a processing step in which new rework target image data of rework target 1 from which soldering has been newly removed is input to new rework target image data input unit 13. In addition, a setting value associated with rework target 1 of the new rework target image data is input to setting value input unit 3. This may be collectively referred to as step 13.

In FIG. 10, step 32 is a processing step in which the learning unit 5 uses the learning results (learning model) to determine similar reprocessing target image data (input into the quality judgment result input unit 4) from the feature quantities of the new reprocessing target image data of the reprocessing target 1, which has been newly removed from the soldering, i.e., the target part, and which has been input into the new reprocessing target image data input unit 13, and judges the quality judgment result of the reprocessing target 1 of the new reprocessing target image data from the quality judgment result associated with the determined reprocessing target image data. Step 33 is a processing step in which the quality judgment result of the newly removed reprocessing target 1, whether it is a good product or a defective product, determined by the learning unit 5 using the learning results (learning model), is input into the quality judgment result input unit 4.

In the rework condition learning device according to the third embodiment (the rework condition learning method according to the third embodiment), the rework target image data of the rework target 1 associated with the good/bad judgment result is input in the good/bad judgment result input unit 4, and the more the learning progresses by associating the setting value that becomes a good product for each rework target 1 with the rework target image data in the learning unit 5, the higher the accuracy with which the good/bad judgment result can be determined even for "image data (in the third embodiment, image data referred to as "new rework target image data" of the rework target 1 from which the solder has been newly removed, i.e., the target component has been removed)" that is not associated with the good/bad judgment result.

The rework condition determination device and rework device shown in FIG. 11 can determine set values and remove soldering using the set values using the pass/fail judgment results determined by the learning unit 5 for "image data (in the third embodiment, image data referred to as "new rework target image data" of rework target 1 from which soldering has been newly removed, i.e., from which the target component has been removed)" that are not associated with a pass/fail judgment result. Of course, it is also possible to determine set values and remove the target component from rework target 1 using the set values, which are the operations described in the first and second embodiments.

As described above, the rework condition learning device according to the third embodiment, the rework condition determination device and rework device using the same (the rework condition learning method according to the third embodiment, the rework condition determination method and rework method using the same) are capable of learning and using the setting values (setting values as removal conditions) that result in a pass for each rework object 1. Furthermore, it is also possible to obtain image data of rework objects that have actually been processed in the past (solder removed), and to perform new pass/fail judgments using the image data of the rework objects.

The rework condition learning device according to the first to third embodiments, the rework condition determination device and the rework device using the same (the rework condition learning method according to the first to third embodiments, the rework condition determination method and the rework method using the same) may be read as the removal condition learning device according to the first to third embodiments, the removal condition determination device and the removal device using the same (the removal condition learning method according to the first to third embodiments, the removal condition determination method and the removal method using the same).

In the rework condition learning device according to the fourth to sixth embodiments, the rework condition determination device and rework device using the same (the rework condition learning method according to the sixth embodiment, and the rework condition determination method and rework method using the same), the setting value input unit 3 (step 1) in the first to third embodiments inputs setting values including the attachment conditions for attaching a new target part to the rework target 1 from which the target part has been removed (i.e., the setting values as the attachment conditions in addition to the setting values as the removal conditions), as shown in FIG. 14 (FIG. 23) described later.

Of course, the rework condition learning device according to the fourth to sixth embodiments, the rework condition determination device and rework device using the same (the rework condition learning method according to the sixth embodiment, the rework condition determination method and rework method using the same) may be such that only the setting values (setting values as attachment conditions) including the attachment conditions for attaching a new target part to the rework target 1 from which the target part has been removed are input in the setting value input unit 3 (step 1) as shown in FIG. 15 (FIG. 25) described later.

Therefore, for example, the rework condition learning device according to the fourth to sixth embodiments learns the setting values that are the attachment conditions for attaching a new target part to a rework target 1 from which a target part has been removed, and may be configured as follows: For each device (rework device 10) that performs reworking, it is provided with a setting value input unit 3 to which a setting value associated with the rework target 1 is input, a quality judgment result input unit 4 to which a quality judgment result of a quality or defective product of the rework target 1 reworked according to the setting value (setting value as an attachment condition) input to the setting value input unit 3 is input, and a learning unit 5 that learns the setting value that will result in a quality product for each rework target 1 based on the setting value and the quality judgment result.

In the explanation of the fourth to sixth embodiments, in order to simplify the explanation, the configuration of FIG. 15 (FIG. 24) will be mainly explained among FIG. 14 and FIG. 15 (FIG. 23 and FIG. 24) described later. The configuration of FIG. 14 (FIG. 23) minus FIG. 15 (FIG. 24) is the same as that of the first to third embodiments. In any case, in the fourth to sixth embodiments, it is desirable that the good/bad judgment result input unit 4 inputs the good/bad judgment result including the judgment result of whether the rework object 1 to which the new object part is attached is good or bad according to the setting value (setting value as the attachment condition) input to the setting value input unit 3. In addition, detailed explanations of the parts common to the first to third embodiments (mostly the parts related to the removal conditions) may be omitted.

Embodiment 4.
Hereinafter, a rework condition learning device according to the fourth embodiment of the present invention, a rework condition determination device using the same, and a rework device using the same will be described with reference to FIGS. 12 to 18. In the drawings, the same reference numerals indicate the same or corresponding parts, and detailed description thereof will be omitted. In FIGS. 14 and 15, the rework object 1 is, for example, a substrate 1 such as a printed circuit board, and includes a single-layer substrate and a multilayer substrate. A substrate on which wiring and components (target components) are soldered may be called a printed wiring board, a printed wiring substrate, a printed circuit board, or a printed circuit board. A substrate on which wiring is applied may simply be called a printed wiring board or a printed wiring substrate. In this case, it can be said that the rework object 1 (substrate 1) includes a printed wiring board and a printed wiring substrate. The substrate 1 may be called a processing object. Furthermore, it can be said that the rework object 1 (substrate 1) is composed of the second solder information (conditions for the second solder) of the second solder printed on the rework object 1, the target components to be soldered, and the member to which the target components are fixed by soldering.

The new target components include electronic components and analog components, whose leads and terminals are fixed to the reprocessing target by soldering through reflow. The members to which the target components are fixed are the aforementioned printed circuit boards, printed wiring boards, and printed wiring boards, with solder formed on the surface by printing or the like. The leads and terminals of the target components also include those that are inserted into through holes formed in the board 1 and fixed by soldering.

12, 14, and 15, the rework condition learning device 2 has a set condition input unit 3, a good/bad judgment result input unit 4, and a learning unit 5. In FIG. 14 and FIG. 15, the rework unit 6 (soldering unit 6b in the fourth embodiment) attaches a new target component to the rework target 1 from which the target component has been removed. Note that FIG. 14 illustrates a rework device 10 having not only the soldering unit 6b but also the removal unit 6a (described in the first to third embodiments) as the rework unit 6. The soldering unit 6b can perform soldering on the new target component, such as a soldering iron unit, a local soldering unit, or a reflow furnace. The rework condition learning device 2 learns the set value, which is the removal condition for removing the target component from the rework target 1 to which the target component has been attached by soldering to the processing target, and constructs a learning model.

The local soldering section may be configured to include at least one or more of a molten solder bath, a molten solder supply path, and a solder jet nozzle (straightening nozzle). Soldering equipment that uses one jet nozzle (straightening nozzle) is sometimes called a point-dip type. Soldering equipment that uses multiple jet nozzles (straightening nozzles) is sometimes called a multi-dip type.

The reflow furnace is equipped with a tunnel-shaped furnace, a reflow blower that sends gas such as oxygen into the furnace, a reflow heating unit that blows gas from the reflow blower or a reflow heating unit that heats the inside of the furnace, and a reflow transport unit (e.g., a conveyor) that transports the rework object 1 with the second solder printed at the soldering location. The reflow blower and reflow heating unit adjust (control) the gas concentration (e.g., oxygen concentration) in the furnace, the temperature in the furnace, and the amount of air blown into the furnace. In addition, the reflow transport unit adjusts (controls) the transport speed of the rework object 1 in the furnace.

In such a case (when the soldering unit 6b is a reflow furnace), the soldering unit 6b may have a solder printing function for printing the second solder. Then, the soldering control unit 12b described below controls the soldering unit 6b to print the second solder on the rework target 1 using the solder printing function before the soldering unit 6b solders the new target component.

Continuing with FIG. 12, FIG. 14, and FIG. 15, the good/bad judgment result input unit 4 inputs the good/bad judgment result of the rework object 1 reworked by the set value (set value as the mounting condition) input to the set value input unit 3. The good/bad judgment result initially input to the good/bad judgment result input unit 4 is, for example, a digitalized result of an appearance inspection (including inspection by a person or a machine) of the printed circuit board (printed circuit board) which is the rework object 1 after reworking. In other words, the result of the appearance inspection (including inspection by a person or a machine) can be digitalized and associated with the set value (set value input to the set value input unit 3) by which the rework object 1 was processed. The appearance inspection here is an inspection (determination) of the good/bad appearance of the rework object 1 after a new target part is attached to the rework object 1 from which the target part has been removed. The learning unit 5 learns the set value that will be a good product for each rework object 1 based on the set value and the good/bad judgment result. The learning unit 5 learns the setting values that will result in a pass for each rework object 1 based on the setting values input to the setting condition input unit 3 and the pass/fail judgment result input to the pass/fail judgment result input unit 4.

Machine learning such as AI may be applied to the learning unit 5 (rework condition learning device 2). The learning unit 5 (rework condition learning device 2) constructs and accumulates a learning model. Preferably, the learning unit 5 may learn, for each rework device 10, different setting values that result in a good product even for the same rework target 1, by associating them as errors for each rework device 10. In addition, the setting value input unit 3 inputs, as attachment conditions, setting values associated with the rework target 1 defined by a second combination of a new target part to the rework target 1 and second solder information for the second solder for attaching the new target part.

In this case, the set value input unit 3 may further input a set value associated with the rework target 1 in the second combination, in which the second solder information is distinguished by at least one of the composition of the second solder, the type of flux added to the second solder, the content of the flux added to the second solder, and the area of the second solder. Similarly, the set value input unit 3 may input a set value associated with the rework target 1 in the second combination, in which the target parts are distinguished by the external shape or the type of electrode surface treatment. Similarly, the set value input unit 3 may input a set value associated with the rework target 1 in the second combination, in which the members are distinguished by at least one of the thickness, material, number of layers, specific gravity, and type of surface treatment.

In other words, the first set value input unit 3 refers to a second combination of the second solder information, the target part to be soldered, and the part to which the new target part is fixed by soldering, in which the second solder information and the target part are the same and the set value is input when the conditions of the part change. The second set value input unit 3 refers to a second combination of the second solder information, the target part to be soldered, and the part to which the target part is fixed by soldering, in which the second solder information and the part are the same and the set value is input when the conditions of the target part change. The third set value input unit 3 refers to a second combination of the second solder information, the target part to be soldered, and the part to which the target part is fixed by soldering, in which the set value is input when the target part and the part are the same and the second solder information (second solder conditions) change.

Here, we will explain the cases where the conditions of the component (rework target 1 on which a new target component is attached) change, and the cases where the conditions of the target component change. As mentioned above, a change in the conditions of the component means that the thickness, material, number of layers, specific gravity, and type of surface treatment of the component change. When the component is a board 1, the thickness means the plate thickness specifications (conductor foil thickness, conductor foil area, and number of layers). When the board 1 is a board, the material means the substrate material of the board and the conductor type of the conductor foil. Furthermore, a change in the conditions of the target component means that the volume, electrode shape, and electrode surface treatment change. The shape and size of the target component can be determined from the volume and electrode shape, so the volume and electrode shape of the target component can be said to be information on the external shape of the target component.

Next, a case where the second solder information (conditions of the second solder) has changed will be described. As described above, the second solder information means a case where the composition of the second solder, the type of flux (type of flux) added to the second solder, the content of the flux added to the second solder, for example, the area of the printed second solder is different. The composition of the second solder may include the type of flux (type of flux) and the content of the flux. In addition, the area of the printed second solder refers to the area of the second solder printed on the surface of the rework object 1. The area of the printed second solder also means the amount of the printed second solder, and considering the thickness of the second solder, it can also be said to be the volume of the printed second solder. Furthermore, the information (conditions) on the area of the printed second solder (printed amount of the printed second solder) may include information on the position of the printed second solder on the surface of the rework target 1, or the information (conditions) on the area of the printed second solder (printed amount of the printed second solder) may be information on the position of the printed second solder on the surface of the rework target 1 itself.

So far, the case where two of the second combinations of the second solder information, the target part to be soldered, and the member to which the target part is fixed by soldering has been described, but all combinations may be covered, or one of the second combinations of the second solder information, the target part to be soldered, and the member to which the target part is fixed by soldering may be fixed. In other words, the set value input unit 3 can be said to input set values (set values as mounting conditions) associated with the rework target 1 defined by the second combination of at least one of the composition of the second solder printed on the rework target 1, the type of flux added to the second solder, the content of flux added to the second solder, the area of the second solder, either one of the outer shape of the target part to be soldered or the electrode surface treatment, and at least one of the thickness, material, number of layers, specific gravity, and type of surface treatment of the member to which the target part is fixed by soldering.

This includes inputting the above-mentioned setting values associated with the rework target 1 defined by a second combination of at least one of the composition of the solder printed on the rework target 1, the type of flux added to the solder, the amount of flux added to the solder, and the area of the solder, the external shape of the part to be soldered, or the electrode surface treatment (at least one of the volume of the part to be soldered, the electrode shape, and the electrode surface treatment), and at least one of the thickness, material, number of layers, specific gravity, and type of surface treatment of the member to which the part to be soldered is fixed. This makes it possible to greatly increase the number of setting values that can be handled. The setting values are used to attach a new part to the rework target 1 from which the part to be soldered has been removed, using the soldering unit 6b (rework unit 6).

The second combination of set values (set values as mounting conditions) may include conditions for the soldering iron section, the local soldering section, and the reflow furnace. For example, in the case of the soldering iron section, at least one of the soldering temperature and the tip shape of the soldering iron including the diameter of the soldering iron is considered for each rework object 1 to be soldered (reworked). In the case of the local soldering section, at least one of the moving speed of the nozzle forming the jet of solder (second solder), the gap dimension between the rework object 1 and the tip of the nozzle, and the jet height of solder (second solder) is considered for each rework object 1 to be soldered (reworked). In the case of the reflow furnace, at least one of the gas concentration in the furnace, the temperature in the furnace, the amount of air blown into the furnace, and the transport speed of the rework object 1 in the furnace is considered for each rework object 1 to be soldered (reworked).

Next, the operation of the rework condition learning device according to the fourth embodiment (the rework condition learning method according to the fourth embodiment) will be described with reference to FIG. 13. In FIG. 13, step 21 is a processing step in which a set value (a set value as an attachment condition) associated with the rework object 1 is input to the set value input unit 3 for each device (specifically, the rework device 10) that performs rework (rework). Step 22 is a processing step in which a good/bad judgment result of the rework object 1 reworked by the set value (a set value as an attachment condition) input to the set value input unit 3 is input to the good/bad judgment result input unit 4. Steps 21 and 22 can be performed in any order. They may be performed simultaneously. Step 23 is a processing step in which the learning unit 5 learns a set value that becomes a good product for each rework object 1 based on the set value input to the set condition input unit 3 and the good/bad judgment result input to the good/bad judgment result input unit 4. Preferably, step 23 may incorporate a processing step in which the learning unit 5 learns that different setting values that result in a non-defective product for the same rework object 1 are associated with errors for each rework device 10.

In the operation of the rework condition learning device according to the fourth embodiment (the rework condition learning method according to the fourth embodiment), similarly to the rework condition learning device according to the fourth embodiment, step 11 is preferably a step in which a setting value associated with the rework target 1 defined by the second combination of the second solder information, the soldered target component, and the member to which the target component is fixed by soldering is input to the setting value input unit 3 as an attachment condition. The conditions of the target component, the conditions of the member, and the combination thereof are also similar to those of the rework condition learning device according to the fourth embodiment.

In addition, in the fourth to sixth embodiments, the second solder and the second solder information may be the same as the first solder and the first solder information in the first to third embodiments, respectively. Similarly, the target part and the new target part may be of the same type or with the same specifications. In other words, the new target part may be any part that corresponds to a replacement for the target part. Therefore, the new target part may be referred to as a replacement target part.

In Fig. 14 and Fig. 15, the rework condition determination device 7 uses the learning results (learning model) of the rework condition learning device 2 shown in Fig. 12, Fig. 14, and Fig. 15 (rework condition determination device according to embodiment 4). The rework condition determination device 7 has a rework object input unit 8 and a rework condition determination unit 9. In Fig. 14 and Fig. 15, the rework object input unit 8 is used to input information on the rework object 1, which is the object to be newly processed (soldered). The rework condition determination unit 9 uses the learning results (learning model) of the learning unit 5 to determine the setting values (setting values as mounting conditions) that will make the rework object 1, whose information has been input to the rework object input unit 8, a non-defective product.

Next, the operation of the rework condition determination device according to the fourth embodiment (the rework condition determination method according to the fourth embodiment) will be described with reference to FIG. 16. In FIG. 16, step 211 is a processing step in which information on the rework object 1, which is the object to be newly reworked (to which the target part is to be attached), is input to the rework object input unit 8. Step 212 is a processing step in which the rework condition determination unit 9 uses the learning result (learning model) of the learning unit 5. Step 213 is a processing step in which the rework condition determination unit 9 uses the learning result (learning model) of the learning unit 5 to determine the setting value (setting value as the attachment condition) that will make the rework object 1, whose information has been input to the rework object input unit 8, a non-defective product.

In FIG. 14 and FIG. 15, the rework device 10 uses the rework condition determination device 7 shown in FIG. 14 and FIG. 15 (rework device according to embodiment 1). The rework device 10 has a soldering section 6b (rework section 6), a condition setting section 11, and a soldering control section 12b (rework control section 12). Note that FIG. 14 illustrates a rework device 10 having not only the soldering control section 12b but also the removal control section 12a (described in embodiments 1 to 3) as the rework control section 12. The condition setting section 11 is input with the set values determined by the rework condition determination section 7. The soldering control section 12b controls the soldering section 6b so that a new target component is attached to the rework target 1 under the conditions input to the condition setting section 11. The soldering control section 12b controls the soldering section 6b so that the new rework target 1 can be reworked (the target component is attached) under the attachment conditions determined by the set values.

Next, the operation of the rework device according to the fourth embodiment (the rework method according to the fourth embodiment) will be described with reference to FIG. 17. In FIG. 17, step 221 is a processing step in which the set values determined by the rework condition determination unit 7 are input to the condition setting unit 11. Step 224 is a processing step in which the mounting is performed. More specifically, step 224 is a processing step in which the soldering control unit 12b controls the soldering unit 6b so that the soldering is performed under the conditions input to the condition setting unit 11. In other words, step 224 is a processing step in which the soldering unit 6b is made to rework a new rework object 1 (attach a new target component) under the mounting conditions determined by the set values.

In the rework condition learning device according to the fourth embodiment, the rework condition determination device and rework device using the same (the rework condition learning method according to the fourth embodiment, and the rework condition determination method and rework method using the same), as mentioned in the explanation of the rework condition learning device (rework condition learning method), the learning unit 5 (in step 23) may learn that for each rework device 10, even if the rework object 1 is the same, different setting values that result in a non-defective product are associated as an error for each rework device 10. This error for each rework device 10 can be used in the rework condition determination device and rework device (the rework condition determination method and rework method according to the fourth embodiment) as follows.

First, in the rework condition determination device (step 212), the learning result of the learning unit 5 is used to determine a set value that will make the rework target 1, whose information has been input to the rework target input unit 8, a non-defective product, and the rework condition determination unit 9 uses the learning result of the learning unit 5 to identify the presence or absence of an error associated with the determined set value. Then, as shown in FIG. 18, in the rework device 10, when there is an error associated with the condition input to the condition setting unit 11 (processing of step 221) and the condition setting unit 11 has already corrected the error (processing of step 222 is YES), the soldering control unit 12b controls the soldering unit 6b to perform soldering using the set value that is the reference value for which the error has been corrected (processing of step 223) (processing of step 224). On the other hand, even if there is an error associated with the condition input to the condition setting unit 11 (processing of step 221) (processing of step 222 is NO), the soldering control unit 12b controls the soldering unit 6b to attach the target component under the condition input to the condition setting unit 11 (processing of step 224).

When proceeding directly from step 222 to step 224, if there is no error associated with the condition input to the condition setting unit 11 (the processing of step 221), the processing is as shown in the flowchart in FIG. 17. In other words, if the error has not been corrected by the condition setting unit 11, the processing will proceed directly from step 222 (step 221) to step 224, regardless of the presence or absence of an error associated with the condition input to the condition setting unit 11 (the processing of step 221).

Furthermore, the set value (set value as the mounting condition) that becomes the reference value after the error is corrected may be prepared individually, or the learning result (learning model) of the learning unit 5 may be used. That is, the learning unit 5 of the rework condition learning device learns the different set values that become good products for the same rework object 1 for each rework device 10 by associating them as errors for each device, and determines the set value that becomes the reference value after the error is corrected. Then, when there is an error associated with the conditions input to the condition setting unit 11 and has been corrected by the condition setting unit 11, the soldering control unit 12b of the rework device 10 controls the soldering unit 6b to mount a new target component using the set value that becomes the reference value determined by the learning unit 5 (the set value that becomes the reference value may be obtained via the rework condition determination device 7).

As described above, the rework condition learning device according to the fourth embodiment, the rework condition determination device and rework device using the same (the rework condition learning method according to the fourth embodiment, the rework condition determination method and rework method using the same) are capable of learning and using the setting value that results in a non-defective product for each rework object 1. The setting value is the attachment condition for attaching a new target part to the rework object 1 from which the target part has been removed.

Embodiment 5.
Hereinafter, a rework condition learning device according to a fifth embodiment of the present invention, a rework condition determination device using the same, and a rework device using the same will be described with reference to Figs. 19 and 20. The following description will focus on the differences from the fourth embodiment. In the drawings, the same reference numerals indicate the same or corresponding parts, and detailed descriptions thereof will be omitted. The major difference between the fourth and fifth embodiments is that the learning unit 5 learns the setting values (setting values as mounting conditions) that result in a non-defective product for each rework object 1 and the rework object image data (such as a photograph of the soldered portion). There are no major differences between the fourth and fifth embodiments with respect to the rework device 10.

In FIG. 19, the good/bad judgment result input unit 4 inputs the good/bad judgment result of the reprocessing target 1 reprocessed according to the setting value (setting value as the mounting condition) input to the setting value input unit 3. Furthermore, the good/bad judgment result input unit 4 inputs the reprocessing target image data (such as a photo of the soldered part) after a new target part is mounted on the reprocessing target 1 associated with the good/bad judgment result. The learning unit 5 learns the setting value that will result in a good product for each reprocessing target 1 based on the setting value input to the setting condition input unit 3 and the good/bad judgment result input to the good/bad judgment result input unit 4. Furthermore, the learning unit 5 learns the setting value that will result in a good product for each reprocessing target 1 by associating it with the reprocessing target image data.

In this way, the rework condition learning device according to the fifth embodiment associates the setting value that results in a pass product with the rework target image data for each rework target 1, and therefore can easily identify the rework target image data corresponding to the determined setting value that results in the pass product for the rework target 1, and can easily obtain an image of the completed product when the rework target 1 is reworked (a new target part is attached) with the determined setting value that results in the pass product for the rework target 1. The rework target image data may be determined by matching the learning result (learning model) of the learning unit 5 with the rework condition determination unit 9 when determining the setting value. The rework condition determination device 7 may be provided with a display unit, including an external unit, that displays the rework target image data obtained (determined) from the learning result (learning model) of the learning unit 5.

Next, the operation of the rework condition learning device according to the fifth embodiment (the rework condition learning method according to the fifth embodiment) will be described with reference to FIG. 20. In FIG. 20, step 21 is a processing step in which the setting values (setting values as attachment conditions) associated with the rework object 1 for each device performing soldering (specifically, rework device 10) are input to the setting value input unit 3. The order of processing of step 21 and the next step 22A does not matter. They may be performed simultaneously.

20, step 22A is a processing step in which the result of the judgment of the quality of the reprocessing target 1, which has been reprocessed according to the setting value (setting value as the mounting condition) input to the setting value input unit 3, is input to the quality judgment result input unit 4, and further, the image data of the reprocessing target 1 after the target part in the reprocessing target 1 associated with the quality judgment result is input to the quality judgment result input unit 4 (such as a photograph of the soldered part). Step 23A is a processing step in which the learning unit 5 learns the setting value that will be a quality product for each reprocessing target 1 based on the setting value input to the setting condition input unit 3 and the quality judgment result input to the quality judgment result input unit 4, and further, the learning unit 5 learns the setting value that will be a quality product for each reprocessing target 1 in association with the reprocessing target image data. As in step 23 in embodiment 4, step 23A may incorporate a processing step in which the learning unit 5 learns that different setting values for the same rework object 1 that become acceptable products are associated as errors for each rework device 10.

As described above, the rework condition learning device according to the fifth embodiment, the rework condition determination device and rework device using the same (the rework condition learning method according to the fifth embodiment, the rework condition determination method and rework method using the same) are capable of learning and using the setting values that result in a non-defective product for each rework object 1. Furthermore, it is also possible to obtain image data of rework objects that have actually been reworked in the past (with the target parts removed).

Embodiment 6.
Hereinafter, a rework condition learning device according to the sixth embodiment of the present invention, a rework condition determination device using the same, and a rework device using the same will be described with reference to FIG. 21 to FIG. 24. The description will focus on the parts different from the fourth and fifth embodiments. In the drawings, the same reference numerals indicate the same or corresponding parts, and detailed description of them will be omitted. The major difference between the fifth embodiment and the sixth embodiment is that, in the fifth embodiment, the learning unit 5 learns the set value (set value as the mounting condition) that becomes a good product for each rework object 1 and the rework object image data (such as a photo of the soldered part) in association with each other, but in the sixth embodiment, a new rework object image data input unit 13 is further provided. There is no major difference between the fourth embodiment, the fifth embodiment, and the sixth embodiment with respect to the rework device 10. There is no major difference between the fifth embodiment and the sixth embodiment with respect to the rework condition determination device 7.

21 and 22, the new reprocessing target image data input unit 13 is used to input new reprocessing target image data of the reprocessing target 1 that has been newly soldered, i.e., the target parts have been attached. The setting value input unit 3 is used to input setting values associated with the reprocessing target 1 of the new reprocessing target image data. The learning unit 5 uses the learning result (learning model) to determine similar reprocessing target image data (input to the quality judgment result input unit 4) from the feature amount of the new reprocessing target image data input to the new reprocessing target image data input unit 13, and judges the quality judgment result of the reprocessing target 1 of the new reprocessing target image data from the quality judgment result associated with the determined reprocessing target image data. The quality judgment result input unit 4 is used to input the quality judgment result of the reprocessing target 1 that has been newly soldered, whether it is a good product or a defective product, as determined by the learning unit 5 using the learning result (learning model).

The learning unit 5 can improve classification accuracy by extracting features from the new rework target image data and learning in association with the features of the rework target image data of the rework target 1 that has been newly soldered, i.e., the target parts attached. In other words, although not shown, the learning unit 5 has a feature extraction unit (feature extraction function) that extracts features from the new rework target image data and the rework target image data. This feature extraction unit (feature extraction function) may be formed in the setting condition input unit 3. The feature extraction unit (feature extraction function) may be formed in the good/bad judgment result input unit 4. In this way, machine learning such as AI in the learning unit 5 (rework condition learning device 2) can be said to capture features by learning using a large amount of sample data and its correct answer labels, and make predictions and decisions for new data.

Next, the operation of the rework condition learning device according to the sixth embodiment (the rework condition learning method according to the sixth embodiment) will be described with reference to FIG. 22. In FIG. 22, step 231 is a processing step in which new rework target image data of rework target 1 that has been newly soldered is input to new rework target image data input unit 13. In addition, a setting value associated with rework target 1 of the new rework target image data is input to setting value input unit 3. This may be collectively referred to as step 213.

In FIG. 22, step 232 is a processing step in which the learning unit 5 uses the learning results (learning model) to determine similar reprocessing target image data (input into the quality judgment result input unit 4) from the feature quantities of the new reprocessing target image data of the reprocessing target 1 that has been newly soldered, i.e., has a target part attached, input into the new reprocessing target image data input unit 13, and judges the quality judgment result of the reprocessing target 1 of the new reprocessing target image data from the quality judgment result associated with the determined reprocessing target image data. Step 233 is a processing step in which the quality judgment result of the newly soldered reprocessing target 1, whether it is a good product or a defective product, determined by the learning unit 5 using the learning results (learning model), is input into the quality judgment result input unit 4.

In the rework condition learning device according to the sixth embodiment (the rework condition learning method according to the sixth embodiment), the rework target image data of the rework target 1 associated with the good/bad judgment result is input in the good/bad judgment result input unit 4, and the more the learning progresses by associating the rework target image data with the setting value that becomes a good product for each rework target 1 in the learning unit 5, the higher the accuracy with which the good/bad judgment result can be determined even for "image data (in the third embodiment, image data referred to as "new rework target image data" of the rework target 1 that has been newly soldered, i.e., has the target component attached)" that is not associated with the good/bad judgment result.

The rework condition determination device and rework device shown in Figs. 23 and 24 can determine set values and perform soldering using the set values using the pass/fail judgment results determined by the learning unit 5 for "image data (in the sixth embodiment, image data referred to as "new rework target image data" of the rework target 1 on which new soldering has been performed, i.e., on which a target component has been attached)" that are not associated with a pass/fail judgment result. Of course, it is also possible to determine set values and attach a new target component to the rework target 1 using the set values, which are the operations described in the fourth and fifth embodiments.

The difference (relationship) between the rework condition determination device and rework device shown in FIG. 23 and the rework condition determination device and rework device shown in FIG. 24 is the same as the difference (relationship) between FIG. 14 and FIG. 15. That is, FIG. 23 illustrates a rework device 10 having not only a soldering section 6b but also a removal section 6a (as described in the first to third embodiments) as the rework section 6. FIG. 23 illustrates a rework device 10 having not only a soldering control section 12b but also a removal control section 12a (as described in the first to third embodiments) as the rework control section 12.

As described above, the rework condition learning device according to the sixth embodiment, the rework condition determination device and rework device using the same (the rework condition learning method according to the sixth embodiment, the rework condition determination method and rework method using the same) are capable of learning and using the setting values (setting values as mounting conditions) that result in a pass for each rework object 1. Furthermore, it is also possible to obtain image data of rework objects that have actually been reworked (soldered) in the past, and to perform new pass/fail judgments using the image data of the rework objects.

Furthermore, in the case of the configuration shown in FIG. 15 (FIG. 25), the rework condition learning device according to embodiment 4 to embodiment 6, the rework condition determination device and rework device using the same (the rework condition learning method according to embodiment 6, the rework condition determination method and rework method using the same) may be read as the soldering condition learning device according to embodiment 4 to embodiment 6, the soldering condition determination device and soldering device using the same (the soldering condition learning method according to embodiment 6, the soldering condition determination method and soldering method using the same).

The rework condition learning device according to the first to sixth embodiments is a rework condition learning device 2 that learns the setting values that are the conditions for reprocessing the rework object 1, and is characterized by including a setting value input unit 3 to which a setting value associated with the rework object 1 is input for each device that performs reprocessing (rework device 10), a pass/fail judgment result input unit 4 to which a pass/fail judgment result of the rework object 1 reprocessed according to the setting value input to the setting value input unit 3 is input, and a learning unit 5 that learns the setting value that will result in a pass/fail product for each rework object 1 based on the setting value and the pass/fail judgment result.

The rework condition determination device and the rework device according to the first to sixth embodiments use the rework condition learning device according to the first to sixth embodiments. The conditions for reprocessing are at least one of the conditions for removing the target component and the conditions for attaching the target component. The reprocessing is at least one of removing the target component soldered with the first solder and attaching the target component by soldering with the second solder. The rework device 10 is at least one of removing the target component by releasing the solder and attaching a new target component by soldering.

1 Rework object, 2 Rework condition learning device, 3 Setting condition input unit,
4: Good/bad judgment result input unit, 5: Learning unit, 6: Rework unit, 6a: Removal unit,
6b soldering unit, 7 rework condition determination device, 8 rework target input unit,
9 Rework condition determination unit, 10 Rework device, 11 Condition setting unit,
12 Rework control section, 12a Removal control section, 12b Soldering control section,
13. New reprocessing target image data input unit.

Claims (35)

A rework condition learning device that learns a setting value that is a removal condition for removing a target component from a rework object in which a target component is attached to a processing object by soldering,
A rework condition learning device comprising: a setting value input unit to which the setting value associated with the rework object is input for each device performing reworking; a pass/fail judgment result input unit to which a pass/fail judgment result of the rework object reworked using the setting value input to the setting value input unit is input; and a learning unit that learns the setting value that will result in a pass/fail judgment result for each rework object based on the setting value and the pass/fail judgment result. The rework condition learning device according to claim 1, characterized in that the setting value input unit inputs, as the removal condition, the setting value associated with the rework object defined by a first combination of first solder information of a first solder formed on the rework object, the soldered object component, a member to which the object component is fixed by soldering, heating conditions, and cooling conditions. The rework condition learning device according to claim 2, characterized in that the set value input unit inputs the set value associated with the rework target, in which the first solder information of the first combination is distinguished by at least one of the composition of the first solder, the type of flux added to the first solder, the content of the flux added to the first solder, and the area of the first solder. The rework condition learning device according to claim 2 or 3, characterized in that the set value input unit inputs the set value associated with the rework target in which the target part in the first combination is distinguished by its external shape or the type of electrode surface treatment. The rework condition learning device according to any one of claims 2 to 4, characterized in that the set value input unit inputs the set value associated with the rework target in which the member in the first combination is differentiated by at least one of thickness, material, number of layers, specific gravity, and type of surface treatment. The rework condition learning device according to claim 1, characterized in that the set value input unit inputs, as the removal conditions, the set values associated with the rework object defined by a first combination of at least one of the composition of the first solder formed on the rework object, the type of flux added to the first solder, the content of flux added to the first solder, the area of the first solder, either the outer shape of the soldered target component or the electrode surface treatment, at least one of the thickness, material, number of layers, specific gravity, and type of surface treatment of the member to which the target component is fixed by soldering, and at least one of the heating conditions and cooling conditions. The rework condition learning device according to any one of claims 1 to 6, characterized in that the learning unit learns, for each device, the setting values that result in a non-defective product even for the same rework target, by associating them as errors for each device. The quality/defective judgment result input unit further receives reprocessing target image data after removing the target part from the reprocessing target associated with the quality/defective judgment result,
The rework condition learning device according to claim 1 , wherein the learning unit further learns the setting value that becomes a non-defective product for each rework object and the rework object image data in association with each other. A new reprocessing target image data input unit is further provided,
The new re-processing target image data input unit receives the new re-processing target image data of the re-processing target from which the soldering has been newly removed,
the setting value input unit receives the setting value associated with the reprocessing target of the new reprocessing target image data,
The learning unit uses the learning result to determine similar reprocessing target image data from the feature amount of the new reprocessing target image data input to the new reprocessing target image data input unit, and judges the quality judgment result of the reprocessing target of the new reprocessing target image data from the quality judgment result associated with the determined reprocessing target image data,
9. The rework condition learning device according to claim 8, wherein the quality judgment result input unit receives a quality judgment result of the rework target, the quality judgment result being determined by the learning unit using a learning result, whether the rework target is a good or defective product. A rework condition determination device using a learning result of the rework condition learning device according to any one of claims 1 to 9,
a rework condition determination unit that determines the setting value that makes the rework object, whose information has been input to the rework object input unit, a non-defective product by using a learning result of the learning unit. A rework condition determination device using a learning result of the rework condition learning device according to claim 7,
a rework condition determination unit that uses a learning result of the learning unit to determine the setting value that makes the rework object, whose information has been input to the rework object input unit, a non-defective product, and identifies the presence or absence of the error associated with the determined setting value. A rework apparatus using the rework condition determination apparatus according to claim 10 or 11,
A rework apparatus comprising: a removal unit of the apparatus; a condition setting unit to which the setting values determined by the rework condition determination unit are input; and a removal control unit that controls the removal unit so as to remove the target part under the conditions input to the condition setting unit. A rework device using the rework condition determination device according to claim 10, which uses the learning result of the rework condition learning device according to claim 7,
a removal unit of the device, a condition setting unit to which the setting value determined by the rework condition determination unit is input, and a removal control unit that controls the removal unit to remove the target part under the condition input to the condition setting unit,
The rework apparatus is characterized in that the removal control unit controls the removal control unit so as to remove the target part under the conditions inputted into the condition setting unit even when there is an error associated with the conditions inputted into the condition setting unit. A soldering apparatus using the rework condition determination apparatus according to claim 11,
a removal unit of the device, a condition setting unit to which the setting value determined by the rework condition determination unit is input, and a removal control unit that controls the removal unit to remove the target part under the condition input to the condition setting unit,
The rework apparatus is characterized in that, when there is an error associated with the conditions input to the condition setting unit and the error has been corrected by the condition setting unit, the removal unit controls the removal unit to remove the target part using the setting value that becomes the reference value obtained by correcting the error. The rework condition learning device according to any one of claims 1 to 6, characterized in that the set value input unit inputs the set value including the attachment condition for attaching a new target part to the rework object from which the target part was removed. A rework condition learning device that learns a setting value that is an attachment condition for attaching a new target part to a rework object from which a target part has been removed,
A rework condition learning device comprising: a setting value input unit to which the setting value associated with the rework object is input for each device performing reworking; a pass/fail judgment result input unit to which a pass/fail judgment result of the rework object reworked using the setting value input to the setting value input unit is input; and a learning unit that learns the setting value that will result in a pass/fail judgment result for each rework object based on the setting value and the pass/fail judgment result. The rework condition learning device according to claim 15 or 16, characterized in that the good/bad judgment result input unit inputs the good/bad judgment result including the judgment result of whether the rework target to which the new target part is attached is a good or bad product based on the setting value input to the setting value input unit. The rework condition learning device according to any one of claims 15 to 17, characterized in that the setting value input unit inputs, as the attachment condition, the setting value associated with the rework object defined by a second combination of a new target component to the rework object and second solder information of a second solder for attaching the new target component. The rework condition learning device according to claim 18, characterized in that the set value input unit inputs the set value associated with the rework target in which the second solder information is distinguished by at least one of the composition of the second solder, the type of flux added to the second solder, the content of the flux added to the second solder, and the area of the second solder, among the second combinations. The rework condition learning device according to claim 18 or 19, characterized in that the set value input unit inputs the set value associated with the rework target in the second combination, the target part being distinguished by its external shape or the type of electrode surface treatment. The rework condition learning device according to any one of claims 18 to 20, characterized in that the set value input unit inputs the set value associated with the rework target, in which the member to which the new target component is fixed by soldering in the second combination is differentiated by at least one of thickness, material, number of layers, specific gravity, and type of surface treatment. The rework condition learning device according to any one of claims 15 to 17, characterized in that the set value input unit inputs, as the attachment conditions, the set values associated with the rework object defined by a combination of at least one of the composition of the second solder for attaching the new target component to the rework object, the type of flux added to the second solder, the content of flux added to the second solder, the area of the second solder, the external shape of the new target component to be soldered, or one of the electrode surface treatment, and at least one of the thickness, material, number of layers, specific gravity, and type of surface treatment of the member to which the new target component is fixed by soldering. The rework condition learning device according to any one of claims 15 to 22, characterized in that the learning unit learns, for each device, the setting values that result in a non-defective product even for the same rework target, by associating them as errors for each device. The rework condition learning device according to any one of claims 17 to 23, characterized in that the pass/fail judgment result input unit further inputs rework object image data after the target part is newly attached to the rework object associated with the pass/fail judgment result, and the learning unit further learns by associating the setting value that becomes a pass product with the rework object image data for each rework object. A new reprocessing target image data input unit is further provided,
The new re-processing target image data input unit receives new re-processing target image data of the re-processing target that has been newly soldered,
the setting value input unit receives the setting value associated with the reprocessing target of the new reprocessing target image data,
The learning unit uses the learning result to determine similar reprocessing target image data from the feature amount of the new reprocessing target image data input to the new reprocessing target image data input unit, and judges the quality judgment result of the reprocessing target of the new reprocessing target image data from the quality judgment result associated with the determined reprocessing target image data,
25. The rework condition learning device according to claim 24, wherein the quality judgment result input unit receives a quality judgment result of the rework target, the quality judgment result being determined by the learning unit using a learning result, whether the rework target is a good or defective product. A rework condition determination device using a learning result of the rework condition learning device according to any one of claims 15 to 25,
a rework condition determination unit that determines the setting value that makes the rework object, whose information has been input to the rework object input unit, a non-defective product by using a learning result of the learning unit. A rework condition determination device using a learning result of the rework condition learning device according to claim 23,
a rework condition determination unit that uses a learning result of the learning unit to determine the setting value that makes the rework object, whose information has been input to the rework object input unit, a non-defective product, and identifies the presence or absence of the error associated with the determined setting value. A rework apparatus using the rework condition determination apparatus according to claim 26 or 27,
A rework apparatus comprising: a soldering unit of the apparatus; a condition setting unit to which the setting values determined by the rework condition determination unit are input; and a soldering control unit that controls the soldering unit so as to perform soldering on a new target component under the conditions input to the condition setting unit. A rework apparatus using the rework condition determination apparatus according to claim 26, which uses a learning result of the rework condition learning apparatus according to claim 23 ,
a soldering unit of the device, a condition setting unit to which the set values determined by the rework condition determination unit are input, and a soldering control unit that controls the soldering unit so as to perform soldering on the new target component under the conditions input to the condition setting unit,
The rework apparatus is characterized in that the soldering control unit controls the soldering control unit to perform soldering on a new target component under the conditions inputted into the condition setting unit, even if there is an error associated with the conditions inputted into the condition setting unit. A rework apparatus using the rework condition determination apparatus according to claim 27,
a soldering unit of the device, a condition setting unit to which the set values determined by the rework condition determination unit are input, and a soldering control unit that controls the soldering unit so as to perform soldering on the new target component under the conditions input to the condition setting unit,
The rework apparatus is characterized in that, when there is an error associated with the conditions input to the condition setting unit and the error has been corrected by the condition setting unit, the soldering control unit controls the soldering unit to perform soldering on a new target component using the setting value that is a reference value obtained by correcting the error. A rework condition learning device that learns a setting value that is a removal condition for removing a target component from a rework object in which a target component is attached to a processing object by soldering,
A rework condition determination device using a learning result of a reflow condition learning device, comprising: a setting value input unit to which the setting value associated with the rework object is input for each device performing rework; a pass/fail judgment result input unit to which a pass/fail judgment result of the rework object reworked by the setting value input to the setting value input unit is input; and a learning unit that learns the setting value that results in a pass for each rework object based on the setting value and the pass/fail judgment result, wherein the learning unit learns the setting value that results in a pass for each device even for the same rework object, associating it with an error for each device, and determines the setting value that is a reference value that corrects the error,
A rework apparatus using a rework condition determination device, comprising: a rework object input unit to which information of the rework object is input; and a rework condition determination unit that uses a learning result of the learning unit to determine the setting value that makes the rework object, the information of which is input to the rework object input unit, a non-defective product,
a removal unit of the device, a condition setting unit to which the setting value determined by the rework condition determination unit is input, and a removal control unit that controls the removal unit so as to remove the target part under the condition input to the condition setting unit,
The rework apparatus is characterized in that, when there is an error associated with the conditions input to the condition setting unit and the error has been corrected by the condition setting unit, the removal unit controls the removal unit to remove the target part using the setting value that becomes the reference value obtained by correcting the error. A rework condition learning device that learns a setting value that is an attachment condition for attaching a new target part to a rework object from which a target part has been removed,
A rework condition determination device using a learning result of a rework condition learning device, comprising: a setting value input unit to which the setting value associated with the rework object is input for each device performing rework; a quality judgment result input unit to which a quality judgment result of a quality or defective product of the rework object reworked by the setting value input to the setting value input unit is input; and a learning unit that learns the setting value that becomes a quality product for each rework object based on the setting value and the quality judgment result, wherein the learning unit learns the setting value that becomes a quality product for each device even if the rework object is the same, by associating it with an error for each device, and determines the setting value that becomes a reference value by correcting the error,
A rework apparatus using a rework condition determination device, comprising: a rework object input unit to which information of the rework object is input; and a rework condition determination unit that uses a learning result of the learning unit to determine the setting value that makes the rework object, the information of which is input to the rework object input unit, a non-defective product,
a soldering unit of the device, a condition setting unit to which the set values determined by the rework condition determination unit are input, and a soldering control unit that controls the soldering unit so as to perform soldering on the new target component under the conditions input to the condition setting unit,
The rework apparatus is characterized in that, when there is an error associated with the conditions input to the condition setting unit and the error has been corrected by the condition setting unit, the soldering control unit controls the soldering unit to perform soldering on a new target component using the setting value that is a reference value obtained by correcting the error. The rework device according to any one of claims 28 to 30 and 32, characterized in that the soldering unit has a solder printing function, and the soldering control unit controls the soldering unit to perform printing on the rework target by the solder printing function before the soldering unit performs soldering on the new target component. A rework condition learning device that learns setting values that are conditions for reworking a rework object by removing or attaching a target part, the rework condition learning device being characterized in that it is equipped with a setting value input unit to which the setting value associated with the rework object is input for each device that performs the reworking, a pass/fail judgment result input unit to which a pass/fail judgment result of the rework object reworked using the setting value input to the setting value input unit is input, and a learning unit that learns the setting value that will result in a pass/fail product for each rework object based on the setting value and the pass/fail judgment result. The rework condition learning device according to claim 34, characterized in that the device is at least one of a device that removes the target component by releasing the soldering and a device that attaches a new target component by soldering.

Images