WO2023223240A1 - Method for determining a disposal operation of an electronic product, computer program, control unit and system - Google Patents
Method for determining a disposal operation of an electronic product, computer program, control unit and system Download PDFInfo
- Publication number
- WO2023223240A1 WO2023223240A1 PCT/IB2023/055087 IB2023055087W WO2023223240A1 WO 2023223240 A1 WO2023223240 A1 WO 2023223240A1 IB 2023055087 W IB2023055087 W IB 2023055087W WO 2023223240 A1 WO2023223240 A1 WO 2023223240A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- electronic
- electronic component
- training
- electronic product
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims description 30
- 238000004590 computer program Methods 0.000 title claims description 5
- 238000012549 training Methods 0.000 claims description 52
- 238000004422 calculation algorithm Methods 0.000 claims description 24
- 238000010801 machine learning Methods 0.000 claims description 13
- 239000000463 material Substances 0.000 claims description 10
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 239000003990 capacitor Substances 0.000 description 15
- 238000012545 processing Methods 0.000 description 15
- 238000011084 recovery Methods 0.000 description 6
- 238000013459 approach Methods 0.000 description 4
- 238000002372 labelling Methods 0.000 description 3
- KDLHZDBZIXYQEI-UHFFFAOYSA-N Palladium Chemical compound [Pd] KDLHZDBZIXYQEI-UHFFFAOYSA-N 0.000 description 2
- 238000007635 classification algorithm Methods 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 2
- 229910052737 gold Inorganic materials 0.000 description 2
- 239000010931 gold Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- BASFCYQUMIYNBI-UHFFFAOYSA-N platinum Chemical compound [Pt] BASFCYQUMIYNBI-UHFFFAOYSA-N 0.000 description 2
- 238000004064 recycling Methods 0.000 description 2
- 229910052709 silver Inorganic materials 0.000 description 2
- 239000004332 silver Substances 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 description 1
- BQCADISMDOOEFD-UHFFFAOYSA-N Silver Chemical compound [Ag] BQCADISMDOOEFD-UHFFFAOYSA-N 0.000 description 1
- 238000012896 Statistical algorithm Methods 0.000 description 1
- ATJFFYVFTNAWJD-UHFFFAOYSA-N Tin Chemical compound [Sn] ATJFFYVFTNAWJD-UHFFFAOYSA-N 0.000 description 1
- 229910052802 copper Inorganic materials 0.000 description 1
- 239000010949 copper Substances 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- -1 e.g. Substances 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 239000011133 lead Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229910052763 palladium Inorganic materials 0.000 description 1
- 229910052697 platinum Inorganic materials 0.000 description 1
- 229910052718 tin Inorganic materials 0.000 description 1
- 239000011135 tin Substances 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/34—Sorting according to other particular properties
- B07C5/342—Sorting according to other particular properties according to optical properties, e.g. colour
Definitions
- the present invention relates to a method for determining a disposal operation of an electronic product .
- the present invention relates to a respective computer program, a control unit implementing the method, and a system comprising the control unit .
- an electronic product when it reaches the end of its li fe , it is disposed of , for example, it can be permanently disposed of ( destroyed) or it can be recycled, at least in part , for example to recover and reuse speci fic materials of the electronic components making up the electronic product itsel f .
- an electronic board as an electronic product , such as a motherboard, a CPU board or a video card
- the electronic board is formed by a plurality of electronic components , for example a resistor, a capacitor or an integrated circuit , mounted on a printed circuit board
- the disposal of a discarded electronic board involves either destroying the entire electronic board or disassembling all the respective electronic components for later recovery .
- the obj ect of the present invention is to overcome the disadvantages of the prior art , in an optimi zed and cost- ef fective manner .
- a method for determining a disposal operation of an electronic product a respective computer program, a control unit implementing the method, and a system comprising the control unit are thus provided, as claimed in the attached claims .
- Figure 1 is a block diagram of a system for implementing a method for determining a disposal operation of an electronic product , according to one embodiment of the present invention
- Figure 2 is a flowchart of the present method, according to one embodiment ;
- Figure 3 is a flowchart of a training step of a machine learning algorithm, according to one embodiment .
- Figures 4A-4D show examples of images of electronic components used for the training step of Figure 3 .
- Figure 1 shows a system 1 configured for determining a disposal operation of an electronic product .
- the electronic product is an electronic board 3 , for example a motherboard, a daughterboard, a CPU board, a video , audio, memory, network card, etc . , of a generic type , such as for consumer applications , or of a speci fic type , such as for industrial or automotive applications .
- the electronic board 3 comprises a support 5 , for example a printed circuit board (PCB ) , and one or more electronic components 7 carried by the support 5 .
- PCB printed circuit board
- the electronic components 7 may be active circuit elements such as, for example, microprocessor integrated circuits, microcontrollers, amplifiers, etc., or passive circuit elements such as, for example, resistors, capacitors, inductors, or other types of circuit elements, such as batteries, antennas, etc.
- the support 5 has a surface 6 and the electronic components 7 are mounted on the surface 6, for example they can be welded, glued or fixed to the surface 6 in another way known per se.
- the electronic board 3 in this embodiment is a printed circuit assembly (PCA) or a printed circuit board assembly (PCBA) .
- PCA printed circuit assembly
- PCBA printed circuit board assembly
- the system 1 comprises an image acquisition apparatus 10 and a control unit 11, which are operatively coupled together .
- the image acquisition apparatus 10 acquires an image IMG, in particular an optical image, of the electronic board 3.
- the image acquisition apparatus 10 may comprise an optical microscope, a camera, etc.
- the image acquisition apparatus 10 can be operated by a user or be an automated image acquisition system, of a type known per se.
- the image acquisition apparatus 10 may also comprise handling means, not shown here, for example a conveyor belt, configured to move the electronic board 3, and a microscope to acquire the image of the electronic board 3.
- handling means for example a conveyor belt, configured to move the electronic board 3, and a microscope to acquire the image of the electronic board 3.
- the image IMG of the electronic board 3 is an image of the surface 6, including the electronic components 7.
- the control unit 11 comprises a classification module or classifier 13, a processing unit 14, and a memory 15, operatively coupled together.
- the classification module 13 is configured to run a machine learning algorithm trained to identify at least one electronic component from an image.
- the machine learning algorithm is trained to identify one or more classes of electronic components present in the image.
- a class may indicate whether the identified electronic component is a capacitor, resistor, inductor, integrated circuit, or any other type of circuit element .
- the classifier 13 can be trained to identify a plurality of classes for each circuit element, wherein each class comprises all circuit elements having one or more properties in common. For example, one class may be associated with a group of capacitors produced by a specific manufacturer, and a different class may be associated with a group of capacitors produced by a different specific manufacturer.
- the classification module 13 is configured to implement an artificial neural network, for example a convolutional artificial neural network, of a type known per se, for the recognition of objects in an image.
- the artificial neural network may be based on the well-known YOLO ("You Look Only Once") algorithm, provided by Ultralytics.
- the control unit 11 is configured, in use, to perform a method, described hereafter with reference to Figure 2 and indicated by reference number 20, for determining a disposal operation of the electronic board 3.
- control unit 11 acquires the image IMG of the electronic board 3.
- control unit 11 in particular the classifier 13, receives the image IMG of the electronic board 3 directly from the image acquisition apparatus 10.
- control unit 11 can be configured to control the image acquisition apparatus 10 so as to command the acquisition of the image IMG by the image acquisition apparatus 10, and the reception of the image IMG.
- the classifier 14 identifies at least one electronic component in the image IMG, by running the machine learning algorithm starting from the image IMG.
- the classifier 14 receives, as input, the image IMG and runs the machine learning algorithm on the image IMG.
- the classifier 13 provides, as output, identification data ID_COMP, indicative of the electronic components 7 of the electronic board 3 that have been identified in the image IMG.
- the identification data ID_COMP may include a list of the identified electronic components, with the respective class they belong to.
- each identified electronic component corresponds to a respective electronic component 7 of the electronic board 3.
- the machine learning algorithm for example the YOLO algorithm mentioned above, may be configured to provide, as output, for each identified electronic component, both the respective class it belongs to and the respective position and/or size in the image IMG.
- the table below shows an example of the identification data ID_COMP output by the classifier 13, purely by way of example :
- the classifier 13 identified a first electronic component, belonging to a class Rl, and a second electronic component, belonging to a class Cl.
- the class Rl is associated with a particular type of resistors
- the class Cl is associated with a particular type of capacitors.
- the classifier 13 also provides a respective centre (X_c and Y_c) and respective dimensions (X_dim and Y_dim) , in an XY reference system of the image IMG.
- the position of each identified electronic component can be used to trace the position of the respective electronic component 7 on the electronic board 3, for example in a subsequent step of disassembling the electronic component itself .
- the main processing unit 14 determines a disposal operation of the electronic board 3, based on the identified electronic components.
- the main processing unit 14 may determine that the electronic board 3 shall be subjected to specific recovery and/or recycling processes, based on the identified electronic components.
- the main processing unit 14 may determine the recovery and/or recycling of one or more of the identified electronic components.
- step 24 the main processing unit 14 processes the identification data ID_COMP and, based on the identification data ID_COMP, determines, for each identified electronic component, whether to command a recovery, step 25, or a final disposal (destruction) , step 26, of the respective identified electronic component.
- step 24 the main processing unit 14 determines, for each identified electronic component, a disposal parameter, which is indicative of a factory data of the electronic component itself.
- the factory data of an electronic component is indicative of composition data of the electronic component, i.e., it is indicative of one or more materials forming the electronic component.
- composition data may indicate whether the electronic component includes precious materials, e.g., gold, silver, platinum, palladium, copper, tin, lead, etc.
- composition data may indicate the type and quantity of precious material.
- Factory data associated with a specific class of electronic components can be stored in the memory 15.
- the main processing unit 14 can be configured to retrieve the factory data via a connection with a specific external apparatus, not shown here.
- the main processing unit 14 may for example compare the factory data associated with the identified electronic components with specific lookup tables, for example stored in the memory 15.
- these lookup tables can associate the factory data of each class with a respective disposal parameter .
- the processing unit 14 selects a specific disposal operation of the identified electronic component.
- the main processing unit 14 based on the disposal parameter, determines either a recovery operation, step 25, or the final disposal, for example the destruction, step 26, of the identified electronic component.
- the main processing unit can compare the disposal parameter with specific threshold values or tables stored in the memory 15.
- the main processing unit 14 can provide an output signal DISP_OP, indicating the respective determined disposal operation.
- the output signal DISP_OP can be used by a user or be provided to a specific machine, not shown here, configured to proceed with the final disposal or with disassembly and recovery of the electronic board 3 and the identified electronic components.
- Step 24 can be repeated for each electronic component identified in step 23.
- the training step 30 may be performed prior to the execution of the method 20, or as the first step of the method 20.
- the training step 30 may be repeated one or more times when using the method 20, depending on the specific application; for example, the training step 30 may be repeated if the number of classes of electronic components to be identified is wished to be expanded or changed.
- one or more training images of the electronic components intended to be identified in step 23 of Figure 2 are used; these electronic components are referred to as electronic components to be identified.
- the training 30 comprises receiving (step 32) , by the control unit 11, at least one training image for each electronic component to be identified.
- the training image may be acquired by the image acquisition apparatus 10 or by other means, for example by a different automated or user-operated imaging acquisition device .
- step 33 the training image is labelled; that is, the training image is associated with a class indicative of the respective electronic component to be identified. This allows for implementation of a supervised learning of the classifier 13.
- the labelling of the training image can be performed by means of a dedicated software, known per se, compatible with the specific supervised learning algorithm used.
- the labelling can be performed via the LabellMG software.
- the training image is processed (step 34) by the control unit 11, e.g., by the main processing unit 14.
- control unit 11 can rotate the training image .
- the training image has a magnification level, for example depending on the specific image acquisition apparatus used.
- the control unit may change the level of magnification of the training image, e.g., the training image may be magnified or reduced.
- the training image processing step 34 may then be used, for example, to form, from the training image, a plurality of training images of the respective electronic component to be identified.
- Steps 32-34 are repeated for each electronic component to be identified.
- steps 32-34 form a training database comprising the training images of the electronic components to be identified.
- the training database may be stored, during training 30, in the memory 15.
- the training database may be stored on a device external to the control unit 11, which is not shown here.
- Figures 4A, 4B, 4C and 4D show examples of training images that can be used for training 30.
- Figures 4A-4D each show a respective capacitor 40A, 40B, 40C and 40D, which can be mounted on a printed circuit board to form the electronic board 3.
- the capacitors 40A-40D differ in type, manufacturer, capacitance value, etc.
- the training images of the capacitors 40A-40D are divided into one or more groups, and a respective class is associated to each group.
- the training images of the capacitors 40A-40D can be labelled so that each capacitor of the plurality of capacitors 40 forms a respective class.
- the classifier 13 can be trained to separately identify each capacitor of the plurality of capacitors 40A-40D.
- the training images and the respective labels form supervised training data.
- the control unit 11, in particular the classifier 13, executes (step 35) a training algorithm using the supervised training data.
- the training algorithm may be an error back-propagation algorithm, of a type known per se.
- the use of the classifier 13 makes it possible to automate and speed up the identification of the electronic components mounted on an electronic board, as opposed, for example, to a manual identification carried out by an operator .
- the Applicant verified that the classifier 13 allows for high accuracy in the identification of the electronic components.
- This identification of the electronic components therefore allows a selective approach, component by component, to be implemented for determining the specific disposal operation to be carried out.
- this selective approach is particularly advantageous when recovering precious materials from the electronic components of an electronic board.
- the classi bomb 13 can be configured to implement a machine learning algorithm other than a neural network .
- control unit 11 can be configured to identi fy the electronic components in the image IMG by using a di f ferent algorithm, for example using statistical algorithms that compare the image IMG with images of the electronic components to be identi fied .
- the control unit 11 can receive the image IMG from an external electronic device or a user, for example from a server or other external devices .
- the disposal parameter may be indicative of a factory data other than a material of which the identi fied electronic component is made ; for example , it may be indicative of identi fication data of the identi fied electronic component , for example an identi fication code , of a speci fic manufacturer of the identi fied electronic component , or the like .
- the main processing unit 14 can determine , for each identi fied electronic component , a speci fic disposal parameter, process the speci fic disposal parameters for several identi fied electronic components , for example through mathematical operations and/or comparison with dedicated tables , and then determine an additional disposal parameter based on which the speci fic disposal operation can be selected ( step 25 or 26 ) .
- the training 30 of the classi bomb 13 may also be performed in a control unit other than the control unit used to perform the method 20 of Figure 2 .
- the training 30 may also comprise a step of testing the classi fication algorithm, following step 35 , which is configured to determine a level of error of the classi fication algorithm when identi fying the electronic components to be identi fied .
- the training database can be divided into a first and a second portion . Step 35 can be performed starting from the first portion of the training database , and the test step can be performed starting from the second portion of the training database .
- the electronic board 3 can have more than one surface on which the electronic components 7 are mounted, such as for example in the case of double-sided printed circuit boards ; in this case , the method 20 of Figure 2 may be repeated for each surface on which the electronic components 7 are mounted .
- the electronic product can be a product other than an electronic board; for example, the electronic product can be a single circuit component, active or passive, such as a capacitor, resistor, integrated circuit, etc.
- the image acquired in step 22 is an image of the single circuit component.
- the electronic component identified by the classifier 13 corresponds to the single circuit component.
Landscapes
- Image Analysis (AREA)
Abstract
In order to determine a disposal operation of an electronic product (3), a control unit (11) performs the steps of acquiring (22) an image (IMG) of the electronic product, wherein the image includes at least one electronic component of the electronic product; identifying (23) the at least one electronic component in the acquired image (IMG); and determining (24, 25, 26) a disposal operation of the electronic product based on the at least one identified electronic component.
Description
METHOD FOR DETERMINING A DISPOSAL OPERATION OF AN ELECTRONIC PRODUCT, COMPUTER PROGRAM, CONTROL UNIT AND SYSTEM
CROSS-REFERENCE TO RELATED APPLICATIONS
This Patent Application claims priority from Italian Patent Application No . 102022000010550 filed on May 20 , 2022 , the entire disclosure of which is incorporated herein by reference .
TECHNICAL FIELD
The present invention relates to a method for determining a disposal operation of an electronic product . In addition, the present invention relates to a respective computer program, a control unit implementing the method, and a system comprising the control unit .
STATE OF THE PRIOR ART
As is well known, when an electronic product reaches the end of its li fe , it is disposed of , for example, it can be permanently disposed of ( destroyed) or it can be recycled, at least in part , for example to recover and reuse speci fic materials of the electronic components making up the electronic product itsel f .
For example , considering an electronic board as an electronic product , such as a motherboard, a CPU board or a video card, the electronic board is formed by a plurality of electronic components , for example a resistor, a capacitor
or an integrated circuit , mounted on a printed circuit board
( PCB ) .
It is also known that the processes for recovering materials , particularly precious materials such as gold, silver, etc . , from an electronic component are highly expensive and polluting .
At present , the disposal of a discarded electronic board involves either destroying the entire electronic board or disassembling all the respective electronic components for later recovery .
However, this approach for recovering materials of the electronic components has a low ef ficiency from a costbenefit point of view .
The obj ect of the present invention is to overcome the disadvantages of the prior art , in an optimi zed and cost- ef fective manner .
SUMMARY OF THE INVENTION
According to the invention, a method for determining a disposal operation of an electronic product , a respective computer program, a control unit implementing the method, and a system comprising the control unit are thus provided, as claimed in the attached claims .
BRIEF DESCRIPTION OF THE DRAWINGS
For a better understanding of the present invention, a preferred embodiment is described below by way of non-
limiting example and with reference to the accompanying drawings , wherein :
• Figure 1 is a block diagram of a system for implementing a method for determining a disposal operation of an electronic product , according to one embodiment of the present invention;
• Figure 2 is a flowchart of the present method, according to one embodiment ;
• Figure 3 is a flowchart of a training step of a machine learning algorithm, according to one embodiment ; and
• Figures 4A-4D show examples of images of electronic components used for the training step of Figure 3 .
DETAILED DESCRIPTION OF THE INVENTION
Figure 1 shows a system 1 configured for determining a disposal operation of an electronic product .
In the embodiment shown here , the electronic product is an electronic board 3 , for example a motherboard, a daughterboard, a CPU board, a video , audio, memory, network card, etc . , of a generic type , such as for consumer applications , or of a speci fic type , such as for industrial or automotive applications .
In detail , the electronic board 3 comprises a support 5 , for example a printed circuit board ( PCB ) , and one or more electronic components 7 carried by the support 5 .
The electronic components 7 may be active circuit
elements such as, for example, microprocessor integrated circuits, microcontrollers, amplifiers, etc., or passive circuit elements such as, for example, resistors, capacitors, inductors, or other types of circuit elements, such as batteries, antennas, etc.
The support 5 has a surface 6 and the electronic components 7 are mounted on the surface 6, for example they can be welded, glued or fixed to the surface 6 in another way known per se.
In practice, the electronic board 3, in this embodiment, is a printed circuit assembly (PCA) or a printed circuit board assembly (PCBA) .
The system 1 comprises an image acquisition apparatus 10 and a control unit 11, which are operatively coupled together .
The image acquisition apparatus 10 acquires an image IMG, in particular an optical image, of the electronic board 3. For example, the image acquisition apparatus 10 may comprise an optical microscope, a camera, etc.
The image acquisition apparatus 10 can be operated by a user or be an automated image acquisition system, of a type known per se.
For example, the image acquisition apparatus 10 may also comprise handling means, not shown here, for example a conveyor belt, configured to move the electronic board 3,
and a microscope to acquire the image of the electronic board 3.
In detail, the image IMG of the electronic board 3 is an image of the surface 6, including the electronic components 7.
The control unit 11 comprises a classification module or classifier 13, a processing unit 14, and a memory 15, operatively coupled together.
The classification module 13 is configured to run a machine learning algorithm trained to identify at least one electronic component from an image.
In detail, the machine learning algorithm is trained to identify one or more classes of electronic components present in the image. For example, a class may indicate whether the identified electronic component is a capacitor, resistor, inductor, integrated circuit, or any other type of circuit element .
Additionally, or alternatively, the classifier 13 can be trained to identify a plurality of classes for each circuit element, wherein each class comprises all circuit elements having one or more properties in common. For example, one class may be associated with a group of capacitors produced by a specific manufacturer, and a different class may be associated with a group of capacitors produced by a different specific manufacturer.
In this embodiment, the classification module 13 is configured to implement an artificial neural network, for example a convolutional artificial neural network, of a type known per se, for the recognition of objects in an image.
For example, the artificial neural network may be based on the well-known YOLO ("You Look Only Once") algorithm, provided by Ultralytics.
The control unit 11 is configured, in use, to perform a method, described hereafter with reference to Figure 2 and indicated by reference number 20, for determining a disposal operation of the electronic board 3.
In a step 22, the control unit 11 acquires the image IMG of the electronic board 3.
In this embodiment, the control unit 11, in particular the classifier 13, receives the image IMG of the electronic board 3 directly from the image acquisition apparatus 10.
According to one embodiment, the control unit 11 can be configured to control the image acquisition apparatus 10 so as to command the acquisition of the image IMG by the image acquisition apparatus 10, and the reception of the image IMG.
In a subsequent step 23, the classifier 14 identifies at least one electronic component in the image IMG, by running the machine learning algorithm starting from the image IMG.
In practice, the classifier 14 receives, as input, the
image IMG and runs the machine learning algorithm on the image IMG.
In detail, the classifier 13 provides, as output, identification data ID_COMP, indicative of the electronic components 7 of the electronic board 3 that have been identified in the image IMG.
For example, the identification data ID_COMP may include a list of the identified electronic components, with the respective class they belong to.
In practice, each identified electronic component corresponds to a respective electronic component 7 of the electronic board 3.
Furthermore, the machine learning algorithm, for example the YOLO algorithm mentioned above, may be configured to provide, as output, for each identified electronic component, both the respective class it belongs to and the respective position and/or size in the image IMG.
The table below shows an example of the identification data ID_COMP output by the classifier 13, purely by way of example :
In the example of the table, the classifier 13 identified
a first electronic component, belonging to a class Rl, and a second electronic component, belonging to a class Cl. For example, the class Rl is associated with a particular type of resistors, and the class Cl is associated with a particular type of capacitors.
For the first and the second electronic components, the classifier 13 also provides a respective centre (X_c and Y_c) and respective dimensions (X_dim and Y_dim) , in an XY reference system of the image IMG.
The position of each identified electronic component can be used to trace the position of the respective electronic component 7 on the electronic board 3, for example in a subsequent step of disassembling the electronic component itself .
Subsequently, the main processing unit 14 determines a disposal operation of the electronic board 3, based on the identified electronic components.
For example, the main processing unit 14 may determine that the electronic board 3 shall be subjected to specific recovery and/or recycling processes, based on the identified electronic components.
Alternatively, the main processing unit 14 may determine the recovery and/or recycling of one or more of the identified electronic components.
In detail, step 24, the main processing unit 14 processes
the identification data ID_COMP and, based on the identification data ID_COMP, determines, for each identified electronic component, whether to command a recovery, step 25, or a final disposal (destruction) , step 26, of the respective identified electronic component.
In step 24, the main processing unit 14 determines, for each identified electronic component, a disposal parameter, which is indicative of a factory data of the electronic component itself.
According to one embodiment, the factory data of an electronic component is indicative of composition data of the electronic component, i.e., it is indicative of one or more materials forming the electronic component.
In particular, the composition data may indicate whether the electronic component includes precious materials, e.g., gold, silver, platinum, palladium, copper, tin, lead, etc.
Moreover, the composition data may indicate the type and quantity of precious material.
Factory data associated with a specific class of electronic components can be stored in the memory 15.
Alternatively, the main processing unit 14 can be configured to retrieve the factory data via a connection with a specific external apparatus, not shown here.
In order to determine the disposal parameters, the main processing unit 14 may for example compare the factory data
associated with the identified electronic components with specific lookup tables, for example stored in the memory 15.
For example, these lookup tables can associate the factory data of each class with a respective disposal parameter .
Based on the disposal parameter, the processing unit 14 selects a specific disposal operation of the identified electronic component.
In the embodiment shown here, the main processing unit 14, based on the disposal parameter, determines either a recovery operation, step 25, or the final disposal, for example the destruction, step 26, of the identified electronic component. For example, the main processing unit can compare the disposal parameter with specific threshold values or tables stored in the memory 15.
For example, in steps 25 and 26, the main processing unit 14 can provide an output signal DISP_OP, indicating the respective determined disposal operation. The output signal DISP_OP can be used by a user or be provided to a specific machine, not shown here, configured to proceed with the final disposal or with disassembly and recovery of the electronic board 3 and the identified electronic components.
Step 24 can be repeated for each electronic component identified in step 23.
Hereinafter, with reference to Figure 3, a training step
30 of the machine learning algorithm implemented by the classifier 13 is described. The training step 30 may be performed prior to the execution of the method 20, or as the first step of the method 20. In addition, the training step 30 may be repeated one or more times when using the method 20, depending on the specific application; for example, the training step 30 may be repeated if the number of classes of electronic components to be identified is wished to be expanded or changed.
For the training of the classifier 13, one or more training images of the electronic components intended to be identified in step 23 of Figure 2 are used; these electronic components are referred to as electronic components to be identified.
In detail, the training 30 comprises receiving (step 32) , by the control unit 11, at least one training image for each electronic component to be identified.
The training image may be acquired by the image acquisition apparatus 10 or by other means, for example by a different automated or user-operated imaging acquisition device .
Next, step 33, the training image is labelled; that is, the training image is associated with a class indicative of the respective electronic component to be identified. This allows for implementation of a supervised learning of the
classifier 13.
The labelling of the training image can be performed by means of a dedicated software, known per se, compatible with the specific supervised learning algorithm used.
For example, the labelling can be performed via the LabellMG software.
In the embodiment shown here, the training image is processed (step 34) by the control unit 11, e.g., by the main processing unit 14.
For example, the control unit 11 can rotate the training image .
Furthermore, the training image has a magnification level, for example depending on the specific image acquisition apparatus used. In addition to or as an alternative to the rotation, the control unit may change the level of magnification of the training image, e.g., the training image may be magnified or reduced.
The training image processing step 34 may then be used, for example, to form, from the training image, a plurality of training images of the respective electronic component to be identified.
Steps 32-34 are repeated for each electronic component to be identified.
In practice, steps 32-34 form a training database comprising the training images of the electronic components
to be identified.
The training database may be stored, during training 30, in the memory 15. Alternatively, the training database may be stored on a device external to the control unit 11, which is not shown here.
By way of example only, Figures 4A, 4B, 4C and 4D show examples of training images that can be used for training 30.
In detail, Figures 4A-4D each show a respective capacitor 40A, 40B, 40C and 40D, which can be mounted on a printed circuit board to form the electronic board 3.
The capacitors 40A-40D differ in type, manufacturer, capacitance value, etc.
During labelling 33, the training images of the capacitors 40A-40D are divided into one or more groups, and a respective class is associated to each group.
For example, according to one embodiment, the training images of the capacitors 40A-40D can be labelled so that each capacitor of the plurality of capacitors 40 forms a respective class. In this way, the classifier 13 can be trained to separately identify each capacitor of the plurality of capacitors 40A-40D.
In practice, the training images and the respective labels form supervised training data.
The control unit 11, in particular the classifier 13,
executes (step 35) a training algorithm using the supervised training data.
For example, in the case where the classifier 13 implements an artificial neural network, the training algorithm may be an error back-propagation algorithm, of a type known per se.
The advantages of the present invention are apparent from what is described and illustrated herein.
The use of the classifier 13 makes it possible to automate and speed up the identification of the electronic components mounted on an electronic board, as opposed, for example, to a manual identification carried out by an operator .
In addition, the Applicant verified that the classifier 13 allows for high accuracy in the identification of the electronic components.
This identification of the electronic components therefore allows a selective approach, component by component, to be implemented for determining the specific disposal operation to be carried out.
In particular, this selective approach is particularly advantageous when recovering precious materials from the electronic components of an electronic board.
In practice, this selective approach optimizes and maximizes the efficiency of the process for the disposal of
an electronic device .
Lastly, it is clear that modifications and variations can be made to what is described and illustrated herein without departing from the scope of the present invention, as defined by the appended claims .
For example , the classi fier 13 can be configured to implement a machine learning algorithm other than a neural network .
Alternatively, the control unit 11 can be configured to identi fy the electronic components in the image IMG by using a di f ferent algorithm, for example using statistical algorithms that compare the image IMG with images of the electronic components to be identi fied .
The control unit 11 can receive the image IMG from an external electronic device or a user, for example from a server or other external devices .
The disposal parameter may be indicative of a factory data other than a material of which the identi fied electronic component is made ; for example , it may be indicative of identi fication data of the identi fied electronic component , for example an identi fication code , of a speci fic manufacturer of the identi fied electronic component , or the like .
For example , in step 24 of Figure 2 , the main processing unit 14 can determine , for each identi fied electronic
component , a speci fic disposal parameter, process the speci fic disposal parameters for several identi fied electronic components , for example through mathematical operations and/or comparison with dedicated tables , and then determine an additional disposal parameter based on which the speci fic disposal operation can be selected ( step 25 or 26 ) .
The training 30 of the classi fier 13 may also be performed in a control unit other than the control unit used to perform the method 20 of Figure 2 .
For example , the training 30 may also comprise a step of testing the classi fication algorithm, following step 35 , which is configured to determine a level of error of the classi fication algorithm when identi fying the electronic components to be identi fied . For example , the training database can be divided into a first and a second portion . Step 35 can be performed starting from the first portion of the training database , and the test step can be performed starting from the second portion of the training database .
For example , the electronic board 3 can have more than one surface on which the electronic components 7 are mounted, such as for example in the case of double-sided printed circuit boards ; in this case , the method 20 of Figure 2 may be repeated for each surface on which the electronic components 7 are mounted .
The electronic product can be a product other than an electronic board; for example, the electronic product can be a single circuit component, active or passive, such as a capacitor, resistor, integrated circuit, etc. In this case, the image acquired in step 22 is an image of the single circuit component. In practice, the electronic component identified by the classifier 13 corresponds to the single circuit component.
Claims
1. A method for determining a disposal operation of an electronic product (3) , the method comprising performing, by a control unit (11) , the steps of: acquiring (22) an image (IMG) of the electronic product, said image (IMG) including at least one electronic component of the electronic product; identifying (23) the at least one electronic component in the acquired image (IMG) ; and determining (24, 25, 26) a disposal operation of the electronic product based on the at least one identified electronic component.
2. The method according to the preceding claim, wherein the at least one electronic component is identified by performing (23) a machine learning algorithm on the image (IMG) .
3. The method according to claim 1 or 2, wherein the disposal operation is determined based on a disposal parameter associated with the identified electronic component .
4. The method according to claim 3, wherein the disposal parameter is one of: a factory data of the identified electronic component; a material of the identified electronic component.
5. The method according to any one of the preceding
claims, wherein the electronic product is an electronic board
(3) having a surface (6) and said electronic component (7) is mounted on the surface, the image (IMG) of the electronic product being an image of the surface of the electronic board .
6. The method according to any one of claims 2 to 5, wherein the machine learning algorithm is an artificial neural network.
7. The method according to any one of claims 2 to 6, further comprising a training step (20) of the machine learning algorithm, including performing, by the control unit, the steps of: acquiring (32, 34) at least one training image relative to an electronic component to be identified; and performing (33, 35) a supervised training algorithm, using said training image as input of the supervised training algorithm.
8. The method according to the preceding claim, wherein the step of training the machine learning algorithm further comprises performing a rotation of the training image, prior to performing the supervised training algorithm.
9. The method according to claim 7 or 8, wherein the step of training the machine learning algorithm further comprising modifying a magnification level of the training image, prior to performing the supervised training
algorithm.
10. A computer program product comprising instructions which, when run by a control unit, cause the control unit to perform the method according to any one of the preceding claims .
11. A control unit (11) for determining a disposal operation of an electronic product (3) , the control unit being configured to: acquire (22) an image (IMG) of the electronic product, said image including at least one electronic component of the electronic product; identify the at least one electronic component in the acquired image (IMG) ; and determine (24, 25, 26) a disposal operation of the electronic product based on the at least one identified electronic component.
12. A system (1) for determining a disposal operation of an electronic product (3) , comprising an image acquisition apparatus (10) configured to acquire an image of the electronic product including at least one electronic component of the electronic product and a control unit (11) configured to: receive, from the image acquisition apparatus, the image of the electronic product; identify the at least one electronic component in the
acquired image ( IMG) ; and determine a disposal operation of the electronic product based on the at least one identi fied electronic component .
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
IT102022000010550 | 2022-05-20 | ||
IT102022000010550A IT202200010550A1 (en) | 2022-05-20 | 2022-05-20 | METHOD FOR DETERMINING A DISPOSAL OPERATION FOR AN ELECTRONIC PRODUCT |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2023223240A1 true WO2023223240A1 (en) | 2023-11-23 |
Family
ID=82850317
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/IB2023/055087 WO2023223240A1 (en) | 2022-05-20 | 2023-05-17 | Method for determining a disposal operation of an electronic product, computer program, control unit and system |
Country Status (2)
Country | Link |
---|---|
IT (1) | IT202200010550A1 (en) |
WO (1) | WO2023223240A1 (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3916637A1 (en) * | 2020-05-29 | 2021-12-01 | Vodafone Group Services Limited | Apparatus and method for detecting elements of an assembly |
EP3950153A1 (en) * | 2019-03-29 | 2022-02-09 | JX Nippon Mining & Metals Corporation | Method for processing electronic/electrical device component scraps |
EP3960312A1 (en) * | 2019-04-22 | 2022-03-02 | JX Nippon Mining & Metals Corporation | Processing method and processing device for electronic/electrical device component scrap |
-
2022
- 2022-05-20 IT IT102022000010550A patent/IT202200010550A1/en unknown
-
2023
- 2023-05-17 WO PCT/IB2023/055087 patent/WO2023223240A1/en unknown
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3950153A1 (en) * | 2019-03-29 | 2022-02-09 | JX Nippon Mining & Metals Corporation | Method for processing electronic/electrical device component scraps |
EP3960312A1 (en) * | 2019-04-22 | 2022-03-02 | JX Nippon Mining & Metals Corporation | Processing method and processing device for electronic/electrical device component scrap |
EP3916637A1 (en) * | 2020-05-29 | 2021-12-01 | Vodafone Group Services Limited | Apparatus and method for detecting elements of an assembly |
Also Published As
Publication number | Publication date |
---|---|
IT202200010550A1 (en) | 2023-11-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Delli et al. | Automated process monitoring in 3D printing using supervised machine learning | |
US20120070086A1 (en) | Information reading apparatus and storage medium | |
KR102657173B1 (en) | System and method of classifying manufactured products | |
JP2012115785A (en) | Sorting system of waste | |
CN104889520B (en) | A kind of intelligent AOI choosings weldering system and method | |
JP7328012B2 (en) | Electronic/electrical device parts scrap composition analysis method, electronic/electrical device parts scrap composition analysis device, and electronic/electrical equipment parts scrap processing method | |
CN1573800A (en) | Image recognition apparatus and image recognition method, and teaching apparatus and teaching method of the image recognition apparatus | |
WO2023223240A1 (en) | Method for determining a disposal operation of an electronic product, computer program, control unit and system | |
CN111126030B (en) | Label typesetting processing method, device and system | |
JP7052248B2 (en) | Identification system, identification method and program | |
JP6824398B2 (en) | Image processing equipment, multiplex communication system and image processing method | |
CN113647208B (en) | Production data creation device and production data creation method | |
JP7328011B2 (en) | Electronic/electrical device parts scrap composition analysis method, electronic/electrical device parts scrap composition analysis device, and electronic/electrical equipment parts scrap processing method | |
CN1936585A (en) | Classification system | |
JP2021159881A (en) | Composition analysis method for electronic/electrical equipment component scrap, disposal method for electronic/electrical equipment component scrap, analysis device for electronic/electrical equipment component scrap, and processing device for electronic/electrical equipment component scrap | |
Gokulnath et al. | Open Source Automated SMD Pick and Place Machine | |
JP2023051934A (en) | Adjustment method for obtaining uniformity of multiple control objects, adjustment system, conveyor device, and adjusted equipment | |
CN116258156A (en) | Method and system for acquiring bad board information by identifying two-dimensional code | |
JP2023024517A (en) | Component mounting equipment, and acquisition method of trace data | |
EP4036674A1 (en) | Individual identification system, individual identification method, and computer-readable recording medium | |
CN204857670U (en) | Detection and treatment facility of bad granule among chip resistor | |
CN111522902B (en) | Data entry method, device, electronic equipment and computer readable storage medium | |
CN113435908A (en) | Method for generating tracing information of PCBA (printed circuit board assembly) | |
KR101559813B1 (en) | Management system for bar-code omission process of surface mounting technology line and control method thereof | |
WO2021024235A1 (en) | Selection system for the parts constituting electrical and electronic equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 23729509 Country of ref document: EP Kind code of ref document: A1 |