Apparatus and method for assigning material value score to printed circuit board waste material or portion thereof and system for sorting printed circuit board waste material

Abstract

To provide an apparatus and a method for assigning a material value score to a printed circuit board waste material or a portion thereof, and a system for sorting the printed circuit board waste material.SOLUTION: An apparatus is configured to determine a material value score and to perform the determination on the basis of a dual energy or spectral X-ray image of a printed circuit board waste material 200 or a portion 210 thereof. Furthermore, the apparatus is configured to perform the determination by applying the dual energy or spectral X-ray image to a machine learning module.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] Embodiments according to the invention relate to an apparatus and method for assigning material value scores to PCB waste or portions thereof, and a system for sorting PCB waste, e.g., applying machine learning assisted part detection methods to x-ray images, for recycling based on material value estimation, for example. [Background technology]

[0002] Electronic equipment and devices play a major role in our everyday lives. However, to achieve the goal of a circular economy, recycling of waste electrical and electronic equipment (WEEE) has become important.

[0003] Currently, in the process of recycling waste printed circuit boards (WPCB) from waste electrical and electronic equipment (WEEE), only rough estimates of the value of the entire WPCB are made before the actual metallurgical recovery. The value of large batches in particular can only be estimated based on a few random samples and visual inspection by trained human personnel, which is time-consuming and prone to error. Neither seller nor buyer of WPCB knows the total value of a batch of WPCB, making it difficult for both to set a price. Currently, there is no known commercially available system that estimates the value of WPCBs quickly and accurately enough.

[0004] There are early scientific publications on the detection of components on PCBs [1] and on estimating the value of WPCBs [2]. However, both are based on visual imaging of the (W)PCB. This technique provides insufficient information when both sides of a (W)PCB are populated with components, since only one side can be analyzed at a time using visible light-based imaging. Also, electronic components that look similar, such as integrated circuits, may have different materials inside. These limitations apply to any other type of imaging that uses a type of radiation that does not penetrate the object but only produces a surface image (reflected light imaging), such as, but not limited to, infrared, ultraviolet and terahertz.

[0005] It is therefore desirable to provide a concept that provides a better compromise between speed and accuracy in determining the value of the WPCB.

[0006] The above mentioned object is achieved by the subject matter of the independent claims of the present application.

[0007] Further embodiments according to the invention are defined by the subject matter of the dependent claims of the present application. Summary of the Invention

[0008] In one aspect of the present invention, the inventors have realized that one problem when attempting to recycle printed circuit boards stems from the fact that the value of waste printed circuit boards (WPCBs) is often determined by humans. According to a first aspect of the present application, the above-mentioned problem is solved by using a device configured to automatically analyze WPCBs, thereby minimizing errors and increasing the speed of analysis. Furthermore, the inventors have found that it is advantageous to determine a material value score associated with each PCB. This material value score makes it possible to determine whether sufficient valuable materials, in particular valuable metals, can be recovered from the WPCBs. This allows for improved accuracy in the assessment of the value of WPCBs to be recycled.

[0009] Thus, according to this aspect of the present application, an embodiment relates to an apparatus for assigning a material value score to a waste printed circuit board or a portion thereof. The material value score may depend, for example, on the content of one or more predetermined materials in or on the WPCB. The material value score may indicate a monetary value of the WPCB. The one or more predetermined materials may be one or more materials to be recovered from the waste printed circuit board. The predetermined materials may be predefined, for example by a user of the apparatus. The material value score may thus indicate whether it is cost-effective to recover the one or more predetermined materials from the WPCB. The content of the one or more predetermined materials may be determined based on an elemental analysis of a representative part of the WPCB. The elemental analysis may be performed using inductively coupled plasma optical emission spectroscopy (ICP-OES).

[0010] In one embodiment, the device is configured to determine a material value score and is configured to perform the determination based on a dual energy or spectral X-ray image of the waste printed circuit board or a portion thereof, where the dual energy or spectral X-ray image may include derivative images thereof, such as elemental material decomposition images or images decomposed into effective atomic number and areal density. The device may be configured to perform the determination of the material value score, for example, by detecting components on the WPCB from the dual energy or spectral X-ray image and determining, for example, a number of components per component type. Optionally, the device may be configured to derive a size and / or weight of the components on the WPCB from the dual energy or spectral X-ray image. The device may, for example, calculate an amount of a predetermined material contained in or recoverable from the WPCB based on information obtained from the dual energy or spectral X-ray image and determine the material value score based on the calculated amount of the predetermined material. The device may comprise, for example, a dual energy or spectral X-ray unit configured to obtain a dual energy or spectral X-ray image of the waste printed circuit boards or a portion thereof, or the device may be configured to receive said dual energy or spectral X-ray image from an external dual energy or spectral X-ray unit. The use of such a dual energy or spectral X-ray image is based on the finding that the dual energy or spectral X-ray image shows the components on the WPCB regardless of the orientation of the WPCB. The advantage is that in the dual energy or spectral X-ray image, information is obtained about the components on both the front and back sides of the WPCB. As no hidden components are present, it is possible to improve the assessment of the WPCB to be recycled by analysing all the components on the WPCB. In this way, the value of the WPCB can be determined more simply and efficiently. Furthermore, the inventors have found that dual energy or spectral X-ray is a very robust imaging technique for applications in dusty, sandy or dirt-filled environments. Therefore, it is possible to distinguish between the components placed on the WPCB even if the components are covered with sand, dust or dirt.This provides a high degree of accuracy and allows for a meaningful material value score to be assigned to the WPCB.

[0011] According to one embodiment, the device is configured to perform the decision by subjecting the dual energy or spectral X-ray image to a machine learning module. This is based on the idea that a high component detection performance can be achieved by the machine learning module. To date, there is no freely accessible annotated X-ray image data of PCBs and their respective components. The inventors have found that an efficient and accurate machine learning module can be achieved by selecting and annotating representative components that are subject to recycling of WPCBs, for example according to the value of the components that are subject to the recycling process. In addition, the inventors have realized that the use of the machine learning module can achieve the processing speed required for the analysis of the value of WPCBs.

[0012] In one embodiment, the device is configured to detect components of printed circuit board waste from dual energy or spectral X-ray images, such as electronic components such as integrated circuits (ICs), tantalum capacitors, ball grid arrays (BGAs), pin grid arrays (PGAs), connectors, etc.

[0013] The apparatus may be configured to obtain information regarding an average material composition for each part type. For example, an external device may analyze a plurality of parts for each part type to determine, for one or more materials, an average content of each material in a plurality of parts of each part type. The average content may be expressed as an average molar concentration or average mass of each material per part or per mass of parts for each part type. Analysis of the materials of the parts may be performed using inductively coupled plasma optical emission spectroscopy (ICP-OES). The apparatus may receive information regarding the average content from an external device as the information regarding the average material composition. Alternatively, the apparatus may include a database indicating, for one or more part types, an average content of one or more materials in parts of each part type as the information regarding the average material composition.

[0014] According to one embodiment, the apparatus may be configured to perform a material value score determination for each part type and for each material (e.g., for at least one part type and one material) to be recovered from the WPCB by the following steps: multiplying the number of parts of each part type by the average mass of each material per part of each part type to obtain a quantity / mass of each material in each part type to determine a quantity / mass value of each material.

[0015] Additionally, the device may be configured to obtain a material value score by summing all values ​​or for each material, by summing all values ​​associated with each material. The materials to be collected may be predefined by the user of the device. Thus, the material value score depends on the material of interest. The material value score of the WPCB may depend on the material selected for the determination. In case of more than one material, the material value score may indicate either an individual value for each of the two or more materials, or a single value that is the sum or weighted sum of all values. The determination of the quantity / mass value of each material may depend on the rate / price of the material. The device may be configured to update the rate / price of the material, for example, daily.

[0016] According to one embodiment, the device is configured to estimate the weight and / or size of the entire WPCB and / or the type, weight and (optionally) size of its components, e.g. electronic components, from the dual energy or spectral X-ray image. Information on the weight and size of the entire WPCB allows the content of valuable material present in the WPCB to be determined. When determining the content of valuable material, the device may be configured to take into account the weight and size of the entire WPCB together with information on the type, weight and (optionally) size of its components. In particular, information on the weight and / or size of the components is advantageous in determining the content of valuable material.

[0017] According to one embodiment, the apparatus may be configured to perform a material value score determination for each part type and for each material to be recovered from the WPCB (e.g., for at least one part type and one material) by the following steps: per part, multiplying the weight of each part by the average mass of each material per mass of each part type to obtain a part-specific quantity-mass of each material for each part, and summing the part-specific quantities-masses of each material across all parts to obtain an overall quantity-mass of each material to determine the overall quantity-mass value of each material.

[0018] Additionally, the device may be configured to sum all values ​​to obtain a material value score, or for each material, sum all values ​​associated with each material to obtain a material value score. The materials to be collected may be predefined by the user of the device. Thus, the material value score depends on the material of interest. The material value score of the WPCB may depend on the material selected for the determination. In the case of more than one material, the material value score may indicate either individual values ​​for each of the two or more materials, or a single value that is the sum of all values. The determination of the quantity / mass value of each material may depend on the rate / price of the material. The device may be configured to update the rate / price of the material, for example, daily.

[0019] In one embodiment, the apparatus is configured to determine a component material value for each component, e.g., electronic component, based on the type, weight, and / or size associated with each component (e.g., electronic component). The component material value depends on the content of one or more predetermined materials in each component. The one or more predetermined materials may be materials targeted for recovery from the WPCB. The component material value may indicate a monetary value of each component. The material value may be negative if the component or one or more components of the component interfere with one of the subsequent steps in the recycling process.

[0020] In one embodiment, the apparatus is configured to perform a material value score determination based on the component material values, for example, the apparatus is configured to sum all component material values ​​to obtain a single value for WPCB.

[0021] In one embodiment, the device is configured to obtain a material value score by simultaneously analyzing both sides of the waste printed circuit boards, thus allowing for efficient analysis of double-sided WPCBs and avoiding overlooking face-down parts, i.e. parts facing the conveyor belt or support pad. This feature eliminates the need to evaluate each side of the WPCB separately and allows for all parts to be detected, whether they are facing the camera or not.

[0022] A further embodiment relates to a method for assigning a material value score to a PCB waste or a part thereof, said method being based on the same considerations as the device described above, otherwise said method being completed by all the features and functions described for said device.

[0023] A further embodiment relates to a system for sorting PCB waste, comprising the above-mentioned device for assigning a material value score to PCB waste or parts thereof, and additionally comprising a sorting device for classifying a plurality of PCB waste into two or more classes of PCB waste according to the material value scores assigned to each of the plurality of PCB waste by the above-mentioned device.

[0024] A further embodiment relates to a method for separating PCB waste, comprising the above-mentioned method for assigning a material value score to PCB waste or parts thereof, and further comprising a step of classifying a plurality of PCB waste into two or more classes of PCB waste according to the material value scores assigned to each of the plurality of PCB waste by the above-mentioned method. The method is based on the same considerations as the above-mentioned system. Otherwise, the method can be completed with all the features and functions described with respect to the system. [Brief description of the drawings]

[0025] The drawings are not necessarily to scale, with emphasis instead generally being placed upon illustrating the principles of the invention. In the following description, various embodiments of the invention are described with reference to the following drawings:

[0026] [Figure 1] FIG. 1 is a schematic diagram of an apparatus for assigning a material value score to a WPCB or portion thereof. [Diagram 2] FIG. 2 is a detailed schematic diagram of an apparatus for assigning a material value score to a WPCB or portion thereof. [Diagram 3] FIG. 3 is a block diagram of a method for assigning a material value score to a WPCB or portion thereof. [Figure 4] FIG. 4 is a schematic diagram of a system for sorting WPCBs. [Diagram 5] FIG. 5 is a block diagram of a method for sorting WPCBs. [Figure 6] FIG. 6 illustrates a method for determining a model for determining a material value score. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0027] Detailed Description of the Embodiments In the following description, equal or similar elements, or elements having equal or similar functions, are given equal or similar reference numbers even if they appear in different figures.

[0028] In the following description, numerous details are set forth to provide a more complete description of the embodiments of the present invention. However, it will be apparent to those skilled in the art that the embodiments of the present invention may be practiced without these specific details. On the other hand, well-known structures and devices are shown in block diagram form instead of detailed illustrations to avoid obscuring the embodiments of the present invention. In addition, the features of the different embodiments described herein can be combined with each other unless otherwise specified.

[0029] 1 illustrates an apparatus 100 for assigning a material value score 110, e.g., 110a or 110b, to a WPCB 200 or a portion 210 thereof. The material value score 110a may be indicative of the value of the entire WPCB when recycling the WPCB 200, and the material value score 110b may be indicative of the value of the portion 210 of the WPCB when recycling the WPCB 200. The material value score 110 may be indicative of whether salvaging material from the WPCB 200 or portion 210 is cost-effective.

[0030] In one embodiment, the device 100 may be configured to divide the WPCB 200 into a number of portions 210 and assign an individual material value score 110b to one or more of the portions. Optionally, the device 100 may be configured to assign an individual material value score 110b to only those portions 210 that have the highest material value scores 110b, effectively indicating which portions 210 of the WPCB 200 are most cost-effective to salvage.

[0031] In one embodiment, the device 100 can be configured to provide both a material value score 110a for the entire WPCB 200 and a material value score 110b for one or more portions 210 of the WPCB 200. The overall material value score 110a allows for a determination of whether the WPCB 200 is waste or worthy of recycling, while the individual material value scores 110b for each portion allow for differentiation of the most valuable portions for recovery of valuable materials.

[0032] In one embodiment, the material value score 110 depends on the content of one or more predetermined materials to be recovered from the waste printed circuit boards 200. In other words, the material value score 110 may depend on the amount of one or more predetermined materials present on the WPCB 200, such as the amount of one or more predetermined materials present on the components 220 of the WPCB 200.

[0033] Optionally, the device 100 may include any of the features described with respect to the device of FIG.

[0034] FIG. 2 shows an apparatus 100 with a dual energy or spectral X-ray image acquisition unit 120 and a material value score determination unit 130 .

[0035] The dual energy or spectral X-ray image acquisition unit 120 can be configured to acquire a dual energy or spectral X-ray image 122 of the WPCB 200 or portion 210 thereof. Optionally, the dual energy or spectral X-ray image acquisition unit 120 is not included with the device 100, but rather the device 100 is configured to receive the dual energy or spectral X-ray image 122 from an external dual energy or spectral X-ray image acquisition unit 120.

[0036] The device 100 is configured to obtain the material value score 110 by simultaneously analyzing both sides 230, 240 of the waste printed circuit board 200. In contrast to other approaches based on visual or NIR images, the X-ray image 122 shows the components 220 on the WPCB 200 regardless of the orientation of the WPCB 200, i.e., the components 220 on the front side 230 and the back side 240 can be detected simultaneously. Furthermore, for applications in sandy, dirty or dusty environments, the XRT 120 is a more robust imaging technique.

[0037] The material value score determination unit 130 is configured to determine a material value score 110 for the WPCB 200 or one or more portions 210 thereof. This determination is performed based on dual energy or spectral X-ray images 122 (XRT).

[0038] Optionally, the material value score determination unit 130 is, for example, a machine learning module.

[0039] Optionally, the material value score determination unit 130 may include an image analysis unit 131 for analyzing the dual energy or spectral X-ray images 122 for individual parts 220 of the WPCB 200. For example, the image analysis unit 131 may be configured to obtain information regarding part types of the parts 220 placed on the WPCB 200 and information regarding the number of parts 220 per part type. Optionally, the image analysis unit 131 may additionally obtain information regarding the size and / or weight of the parts 220.

[0040] The material value score determination unit 130 is configured, for example, for each part type, to determine a part material value score, e.g., using the part material value determination unit 136, to obtain a plurality of part material value scores 112. The part material value score may be determined by multiplying the number of parts 220 of each part type by an average part material value score associated with each part type. Each average part material value score may be a predefined value score assigned to each part type. The part material value determination unit 136 may comprise a database indicating the plurality of average part material value scores and, for each average part material value score, its associated part type. Alternatively, the part material value determination unit 136 may be configured to update, for each part type, the corresponding average part material value score based on a change in the market value of valuable materials contained in each part type. The apparatus 100 is configured to perform the determination of the material value score 110 based on the plurality of part material value scores 112, e.g., by summing the plurality of part material value scores 112.

[0041] The image analysis unit 131 is, for example, a machine learning module.

[0042] Optionally, the apparatus 100 may be configured to detect components 220 of the PCB waste 200 from the dual energy or spectral X-ray image 122, for example using the component detection unit 132. Additionally, the apparatus 100 may be configured to estimate characteristics of the components 220, such as type, weight and / or size, from the dual energy or spectral X-ray image 122, for example using the component characterization unit 133. The apparatus 100 may be configured to determine a component material value for each component 220 based on one or more characteristics associated with each component 220.

[0043] The apparatus 100 may be configured to obtain information regarding an average mass of a predefined material per size or weight of a part type. Optionally, the average mass may be obtainable for each different material and / or each different part type. Using the information regarding the average mass and the information regarding the size and / or weight of each individual part 220 on the WPCB 200, the apparatus 100 may be configured to determine an amount of the predefined material included on the WPCB 200. In addition, the apparatus 100 may be configured to obtain a rate / price of the predefined material included on the WPCB 200 and determine a material value score 110 based on the rate / price of the predefined material and the amount of the predefined material. This may be performed for one or more predefined materials, and the material value score may indicate the value of each individual material or the value of all predefined materials combined.

[0044] Similarly, the apparatus 100 may be configured to obtain information regarding an average mass of a predefined material per part of a part type. Optionally, the average mass may be obtainable for each different material and / or for each different part type. Using the information regarding the average mass and the information regarding the number of individual parts 220 on the WPCB 200, the apparatus 100 may be configured to determine the amount of the predefined material included on the WPCB 200. In addition, the apparatus 100 may be configured to obtain a rate / price of the predefined material included on the WPCB 200 and determine a material value score 110 based on the rate / price of the predefined material and the amount of the predefined material. This may be performed for one or more predefined materials, and the material value score may indicate the value of each individual material or the value of all predefined materials combined.

[0045] 3 illustrates a block diagram of a method 300 for assigning 310 a material value score 110 to a WPCB 200 or portion 210 thereof. Optionally, the method 300 comprises determining 320 the material value score 110. The determining 320 may be performed based on a dual energy or spectral X-ray image 122 of the WPCB 220 or portion 210 thereof. It may be advantageous to perform the determining 320 by subjecting the dual energy or spectral X-ray image 122 to a machine learning process.

[0046] The method 300 may include features and / or functionality described with respect to the apparatus 100 of FIG. 1 and / or FIG.

[0047] 4 shows a system 400 for sorting the waste printed circuit boards 200, e.g., 2001-2005. The system 400 comprises an apparatus 100 for assigning a material value score 110, e.g., 1101-1105, to the WPCB 200 or a portion 210 thereof. In addition, the system 400 comprises a sorting device 410 for classifying the plurality of waste printed circuit boards 200 into two or more classes 420, e.g., 4201-4202, for the waste printed circuit boards 200 according to the material value score 110 assigned by the apparatus 100 to each of the plurality of waste printed circuit boards 200.

[0048] For example, a first class 4201 may represent waste, i.e., WPCBs 200 that are not cost-effective to recycle, and a second class 4202 may represent valuable WPCBs 200, i.e., WPCBs 200 that are cost-effective to recycle.

[0049] The device 100 may include the features and / or functionality described with respect to the device 100 of FIG. 1 and / or FIG.

[0050] 5 shows a block diagram of a method 500 for separating the waste printed circuit boards 200. The method 500 includes the method 300 for assigning a material value score 110 to the WPCB 200 or a portion 210 thereof, and classifying 510 the plurality of waste printed circuit boards 200 into two or more classes 420 for the waste printed circuit boards 200 according to the material value score 110 assigned to each of the plurality of waste printed circuit boards 200 by the method 300.

[0051] The method 300 may include the features and / or functionality described with respect to the method 300 of FIG.

[0052] The method 500 may include features and / or functionality described with respect to the method 400 of FIG.

[0053] 4 and 5 illustrate a sorting system 400 and method 500 for automatically sorting WPCBs 200 based on the predicted value or material value score 110 of each individual WPCB 200. The system 400 may acquire dual energy or spectral X-ray images 122 (XRT) from the WPCBs 200 on a conveyor belt or chute. These images 122, after pre-processing, may be input to, for example, a deep neural network (or other suitable machine learning method, such as a method for detecting objects in an image), which detects components 220, such as ICs, BGA / PGAs, tantalum capacitors, connectors, etc. The system 400 calculates the value or component material value of each component 220 on the WPCB 200 using a model based on features such as type, size, and / or weight of the component 220. Both the size and weight of the detected components 220 may be estimated from the X-ray images 122.

[0054] The resulting monetary value or component material value 112 of the WPCBs 200 from the model may be adapted to the particular metallurgical process used to process these WPCBs 200 at a later stage. Furthermore, the model for predicting the value or material value score 110 of the WPCBs 200 may be adapted to the actual process of the user of the system 400, e.g., the yield of the actual metallurgical process, the current rate of material value, or the value-decreasing components that are impeding the process. The user may set a threshold value for the predicted value or value content 110 of the WPCBs 200 below which the system 400 will reject and classify the WPCB, e.g., classify the WPCB into grade 4201. Value content in this case refers to value or monetary value per unit mass or area, e.g., euro / kg or euro / cm. 2 means.

[0055] FIG. 6 illustrates a process diagram of a method 600 for determining a model that may be used by the apparatus 100 described herein to determine a material value score 110.

[0056] FIG. 6 illustrates, for example, the training of a model 132 for object detection, e.g., a deep learning model (DL model). In machine learning, a representative database is essential for training and evaluating the model. Thus, a first step may be to generate such a machine-readable database and label base from a sample of real-world WPCBs 200. In one experiment, the inventors performed annotation 610, e.g., of parts 220 (e.g., representative components) from, e.g., individual WPCBs 200 (e.g., the number of WPCBs may be 104) to obtain annotated components 612 (e.g., the number of annotated components may be 1514). For example, for training, the parts 220 of the WPCBs 200 are annotated 610 to obtain a dataset or database of annotations, i.e., annotated components 612. A further step may involve the acquisition 120 of dual energy or spectral X-ray images 122 of the WPCBs. A model 132 for object detection can be trained based on the x-ray image 122 of the WPCB 200 and based on the annotated parts 612 of the same WPCB 200. The trained model 132 and / or the model under training 132 are configured to identify, recognize, or detect parts of the WPCB 200 to obtain detected parts 614.

[0057] In one embodiment, the object detection model 132 may be configured to estimate characteristics such as type, weight and / or size of the part 220 from the dual energy or spectral X-ray image 122 .

[0058] Additionally, the method 600 may include the step of preparing and removing representative samples of different part classes from the WPCB 200. These samples, i.e., the WPCB 200, are then analyzed 620, for example by ICP-OES (Inductively Coupled Plasma Optical Emission Spectroscopy, or any other method that provides elemental content may be used), to obtain the content 622 of one or more predetermined materials, e.g., valuable metals. This information, i.e., the content 622 of one or more predetermined materials in the representative parts 220 and the detected parts 614, provides a truth value on which the model 130 is calibrated, which predicts the actual value of all applicable parts, e.g., the detected parts 614, and ultimately the value 110 of the WPCB 200, e.g., as a sum or weighted sum of the values ​​of all applicable parts. To obtain the elemental content 622 of one or more predetermined materials in the representative parts of the WPCB 200, the elemental analysis 620 of the representative parts may be performed only once per application. The obtained elemental content 622 can be used to calibrate the model 130 and to perform value estimation.

[0059] Representative components may include integrated circuits (ICs), tantalum capacitors, ball grid arrays (BGAs), pin grid arrays (PGAs), and / or connectors. The selection of representative components may depend on one or more given materials to be recovered from the WPCB. Representative components may also include components that impede subsequent process steps in reclamation resulting in negative value, such as large aluminum heat sinks or toxic-containing components.

[0060] In one embodiment, the apparatus 100 described herein for determining a material value score 110 may be configured to obtain a dual energy or spectral X-ray image 122 of the WPCB 200 and determine the detected parts 614 using a trained model 132 for object detection. Additionally, the apparatus 100 may be configured to determine the material value score 110 based on the detected parts 614 using a model 130 for value estimation.

[0061] From this, a machine learning model can be trained to detect the relevant parts 220. From the ICP-OES, the content of valuable materials in every part class is known. The value of each material in each part can be estimated using the weight estimated from the X-ray images 122 and the current rate of the material. The value 110 of the entire WPCB 200 can then be calculated by summing the values ​​of all the parts, i.e., summing the part material values ​​112. Depending on the subsequent metallurgical processes and other user specific parameters, a weighted sum of the material values ​​and the weighted sum of the total values ​​can be used to predict the material value (in terms of achievable profit). The actual weights need to be adjusted to reflect the respective process. These weights can also reflect the recovery rate of a given material in the process. Some parts may have a negative value assigned if they interfere with subsequent process steps.

[0062] In the described approach to detect components 220 on WPCB 200 by deep learning methods, the inventors found a suitable approach to integrate prior art object detection networks such as YOLO [3] and EfficientDet [4]. In training and evaluating the database, the inventors followed the established evaluation scheme in machine learning. The data is divided into three separate sets, namely training, validation and test. The validation set is used to adjust the model training process and verify its performance. After the final model is found, the test data is used as an independent final evaluation. The inventors obtained the harmonic mean scores of recall and precision for four component classes, namely IC, tantalum capacitor, connector and BGA / PGA, which are 87.83%, 82.54%, 79.26% and 88.89%. Based on the component detection performance and statistical findings from chemical analysis, the model performs well in predicting the material value score 110 from the WPCB 200 of individual WEEE.

[0063] The advantage is that the component 220 can be detected regardless of the visibility of the component 220. That is, the component 220 can appear in the X-ray image 122 whether the PCB 200 is located above (230) or below (240) the component 220, or whether it is populated on both. Thus, the X-ray based approach is independent of the position and orientation of the PCB during the imaging process, and thus can still detect all components 220 on the PCB 200 without the need for explicit positioning.

[0064] Assessment of value 110 based on detected components 614 is advantageous because it allows flexibility in prioritizing types of materials etc. that are important to the user and incorporating effects such as recovery rates of material components.

[0065] Although some aspects have been described in the context of an apparatus, it will be appreciated that these aspects also represent a description of a corresponding method, with the blocks or apparatus corresponding to method steps or features of a method step. Similarly, aspects described in the context of a method step also represent a description of a corresponding block or item or feature of a corresponding apparatus.

[0066] Depending on the requirements of a particular implementation, the embodiments of the present invention may be implemented in hardware or in software, the implementation being possible using a digital storage medium, such as a floppy disk, DVD, CD, ROM, PROM, EPROM, EEPROM or flash memory, storing electronically readable control signals and cooperating (or being capable of cooperating) with a programmable computer system to cause the corresponding methods to be carried out.

[0067] Some embodiments according to the invention include a data carrier having electronically readable control signals and operable with a programmable computer system to perform one of the methods described herein.

[0068] Generally, embodiments of the invention can be realized as a computer program product having program code operable to perform one of the above-mentioned methods when the computer program product is run on a computer, the program code can for example be stored on a machine readable carrier.

[0069] Other embodiments comprise the computer program for performing one of the methods described herein, stored on a machine readable carrier.

[0070] In other words, an embodiment of the inventive method is, therefore, a computer program having a program code for performing one of the methods described herein, when the computer program runs on a computer.

[0071] A further embodiment of the inventive method is, therefore, a data carrier (or digital storage medium, or computer readable medium) comprising, recorded thereon, the computer program for performing one of the methods described herein.

[0072] A further embodiment comprises a processing means, for example a computer, or a programmable logic device, configured to or adapted to perform one of the methods described herein.

[0073] A further embodiment comprises a computer having installed thereon the computer program for performing one of the methods described herein.

[0074] In some embodiments, a programmable logic device (e.g., a field programmable gate array) may be used to perform some or all of the functions of the methods described herein. In some embodiments, a field programmable gate array may cooperate with a microprocessor to perform one of the methods described herein. In general, the methods may be performed by any hardware device.

[0075] The above-described embodiments merely exemplify the principles of the present invention. It is understood that modifications and variations of the arrangements and details described herein will be apparent to those skilled in the art. It is therefore intended to be limited only by the scope of the appended claims and not by the specific details presented as descriptions and explanations of the embodiments herein.

[0076] [References] [1] Mallaiyan Sathiaseelan, MA, Paradis, OP, Taheri, S., & Asadizanjani, N. (2021). Why Is Deep Learning Challenging for Printed Circuit Board (PCB) Component Recognition and How Can We Address It?. Cryptography, 5(1), 9. [2] Silva, LHDS, Junior, AA, Azevedo, GO, Oliveira, SC, & Fernandes, BJ (2021). Estimating Recycling Return of Integrated Circuits Using Computer Vision on Printed Circuit Boards. Applied Sciences, 11(6), 2808. [3] Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You only look once: Unified, real-time object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 779-788). [4] Tan, M., Pang, R., & Le, Q. V. (2020). Efficientdet: Scalable and efficient object detection. In Proceedings of the IEEE / CVF conference on computer vision and pattern recognition (pp. 10781-10790).

Claims

1. An apparatus (100) for assigning a material value score (110) to a printed circuit board waste (200) or a portion thereof (210), comprising: An apparatus configured to determine the material value score (110) and configured to perform said determination based on a dual energy or spectral X-ray image (122) of the printed circuit board waste (200) or a portion thereof (210).

2. 2. The apparatus (100) of claim 1, The apparatus is configured to perform the determination by subjecting the dual energy or spectral X-ray image (122) to a machine learning module (132).

3. 3. The apparatus of claim 2, The machine learning module (132) is trained based on a dataset of annotated parts (612), the annotated parts (612) representing annotations of representative parts of a WPCB (200).

4. An apparatus (100) according to any one of claims 1 to 3, comprising: An apparatus configured to detect components (220) of the printed circuit board waste (200) from the dual energy or spectral X-ray image (122).

5. 5. The apparatus (100) of claim 4, An apparatus configured to estimate a type, weight and / or size of the part (220) from the dual energy or spectral X-ray image (122).

6. 6. The apparatus (100) of claim 5, The apparatus is configured to determine, for each part (220), a part material value based on a type, weight, and / or size associated with each part (220).

7. 7. The apparatus (100) of claim 6, An apparatus configured to perform the determination of the material value score (110) based on the part material value (112).

8. An apparatus (100) according to any one of claims 1 to 7, comprising: An apparatus configured to obtain said material value score (110) by simultaneously analyzing both sides (230, 240) of said printed circuit board waste (200).

9. An apparatus (100) according to any one of claims 1 to 8, comprising: The material value score (110) depends on the content of one or more predetermined materials to be recovered from the printed circuit board waste (200).

10. A method (300) for assigning (310) a material value score (110) to a printed circuit board waste (200) or a portion (210) thereof, comprising: The method comprises determining (320) the material value score (110), the determining (320) being performed based on a dual energy or spectral X-ray image (122) of the printed circuit board waste (200) or a portion thereof (210).

11. 11. The method (300) of claim 10, wherein the determining (320) is performed by subjecting the dual energy or spectral X-ray image (122) to a machine learning process.

12. A system (400) for separating printed circuit board waste (200), comprising: An apparatus (100) for allocating according to any one of claims 1 to 9, A sorting device (410) for classifying a plurality of printed circuit board waste (200) into two or more classes (420) of printed circuit board waste (200) according to a material value score (110) assigned to each of the plurality of printed circuit board waste (200) by the device (100) for assigning.

13. A method (500) for separating printed circuit board waste (200), comprising: A method (300) for allocating according to claim 10 or claim 11, and classifying (510) a plurality of printed circuit board waste (200) into two or more classes (420) of printed circuit board waste (200) according to a material value score (110) assigned to each of the plurality of printed circuit board waste (200) by the method for assigning (300).

Images