Method and system for sorting metal-containing objects in a metal-containing object stream

Abstract

The present disclosure relates to a method for sorting metal-containing objects in a stream of metal-containing objects, the method comprising: measuring a contamination level on a surface of a metal-containing object. The method further includes locating at least one point on the surface at which a contamination level is below a threshold level; and directing a laser beam at the at least one point to locally form a plasma and measure a spectrum of the plasma; or directing a laser beam at the at least one point to form an optical signal comprising a plurality of Raman wavelengths and measuring a spectrum of the Raman wavelengths. The method further comprises assigning a category to the metal-containing object based on the measured spectrum, and sorting the metal-containing object into one of the first output fraction (A) and the second output fraction (B) based on the assigned category.

Description

Technical Field

[0001] This invention generally relates to a method for sorting metal-containing objects in a flow containing metal. This invention also generally relates to a system for sorting metal-containing objects in a flow containing metal. Background Technology

[0002] With today's focus on sustainability and green technologies, material recycling is becoming increasingly important. Since the mining of primary metal materials is not only costly and environmentally damaging, but also often fails to meet sustainability standards, recycling metals contained in waste is both necessary and highly desirable from an economic perspective. Once discarded metal-containing objects have been largely separated from non-metallic waste, further sorting should be performed to classify and separate the metal-containing objects, thereby facilitating a more efficient recycling process. Therefore, there is a clear need for an improved method and system for sorting metal-containing objects in a flow of materials containing metals. Summary of the Invention

[0003] In view of the above, the object of this disclosure is to provide an improved method for sorting metal-containing objects in a flow of metal-containing objects. The method includes: measuring the contamination level on the surface of the metal-containing object; locating at least one point on the surface where the contamination level is below a threshold level; directing a laser beam to the located point to locally form a plasma and measuring the spectrum of the plasma; or directing the laser beam to the located point to form an optical signal comprising a plurality of Raman wavelengths and measuring the spectrum of these Raman wavelengths; the method further includes: assigning a category to the metal-containing object based on the measured spectrum, and sorting the metal-containing object into one of a first category and a second category based on the assigned category.

[0004] According to the implementation method, the pollution level is measured by a visible light sensor, a UV sensor, and / or a NIR sensor.

[0005] According to the implementation method, the measurement of the plasma spectrum is performed by LIBS, or the measurement of the spectrum of multiple Raman wavelengths is performed by Raman spectroscopy analysis, such as SERS.

[0006] According to the implementation method, locating at least one point further includes: object detection by laser triangulation and / or image analysis.

[0007] According to the implementation method, object detection also includes: height detection based on time-of-flight technology, structured light technology, optical parallax technology and / or the output of a height detector.

[0008] According to an implementation, the method further includes cleaning the object and optionally drying the object before measuring the contamination level.

[0009] According to an implementation, the method further includes: if a point where the contamination level is below a threshold level cannot be located, then the object is re-cleaned.

[0010] According to the implementation method, cleaning and / or re-cleaning includes polishing, spraying liquid, mechanical cleaning, chemical cleaning, high-pressure water cleaning, laser-based cleaning, or any combination thereof.

[0011] According to an embodiment, the method further includes: if a plurality of points on the surface have a contamination level below a threshold level, selecting one of the plurality of points and directing a laser beam to the selected point.

[0012] According to the implementation method, guiding the laser beam is performed by manipulating a movable reflector that reflects the laser beam.

[0013] According to the implementation method, the category assignment is performed by a machine learning algorithm.

[0014] According to the implementation method, sorting a metal-containing object into one of a first class and a second class is based on one of the metallic properties of the metal-containing object and the metallic properties of the metal-containing object being composed of an alloy.

[0015] Another object of this disclosure is to provide an improved system for sorting metal-containing objects in a flow of metal-containing objects. The system includes: a measurement unit for measuring the contamination level on the surface of the metal-containing object; a laser unit including a laser direction unit and a spectrometer; and a processing unit including circuitry configured to: receive the measured contamination level from at least one measurement unit; locate at least one point on the surface where the contamination level is below a threshold level; and transmit the located point to the laser direction unit, wherein the laser direction unit is configured to form a plasma at the located point, or to form an optical signal including a plurality of Raman wavelengths; wherein the spectrometer is configured to measure the spectrum of the plasma, or to measure the spectrum of the plurality of Raman wavelengths; wherein the circuitry is further configured to receive the measured spectrum from the spectrometer, and to assign a category to the metal-containing object based on the received spectrum; and wherein the system further includes: at least one sorting unit for sorting the metal-containing object into one of a first category and a second category based on the assigned category.

[0016] According to another embodiment, the measurement unit includes a sensor, such as a visible light sensor, a UV sensor, and / or a NIR sensor.

[0017] According to another implementation, the laser unit is a LIBS unit, or a Raman spectroscopy analysis unit, such as a SERS unit.

[0018] According to another embodiment, the system includes a cleaning unit and optionally a drying unit.

[0019] According to another embodiment, the cleaning unit is one of a polishing unit, a liquid spraying unit, a mechanical cleaning unit, a chemical cleaning unit, a high-pressure water cleaning unit, a laser-based cleaning unit, or any combination thereof.

[0020] According to another embodiment, the circuitry of the processing unit is further configured to: if multiple points on the surface have a contamination level below a threshold level, select one point from the multiple points and guide the laser beam to the selected point.

[0021] According to an embodiment, the laser direction unit includes at least one movably arranged reflector.

[0022] According to the implementation, the circuit is also configured to assign categories using machine learning algorithms.

[0023] According to another embodiment, the sorting unit is configured to sort objects containing metal based on either the metallic properties of the object containing metal or the metallic properties of the alloy composition of the object containing metal. Attached Figure Description

[0024] Various aspects of the inventive concept, including its specific features and advantages, will be readily understood from the following detailed description and accompanying drawings. The drawings are provided to illustrate the overall structure of the inventive concept. The same reference numerals refer to the same elements throughout the text.

[0025] Figure 1A This is a schematic diagram of the system according to the present invention.

[0026] Figure 1B This is a detailed schematic diagram of a part of the system according to the present invention.

[0027] Figure 2 This is a flowchart of contamination detection and sorting of metal-containing objects according to the present invention.

[0028] Figure 3 This is a flowchart of contamination detection and sorting of metal-containing objects according to another embodiment. Detailed Implementation

[0029] The inventive concept can be implemented in many different forms and should not be construed as being limited to the variations set forth herein; rather, these variations are provided for thoroughness and completeness and to fully communicate the scope of the inventive concept to those skilled in the art.

[0030] refer to Figure 1A The system 1 according to the invention includes an input side 1' and an output side 1'. The system also includes a conveyor 2 for transporting a flow of objects containing metal from the input side 1' to the output side 1'. The overall transport direction of the conveyor 2 is indicated by a solid arrow in Figure 1. The conveyor 2 may include at least one conveyor belt. The conveyor 2 may be driven by a drive unit 2A. The conveyor 2 includes a transport surface on which the objects containing metal are transported. The conveyor 2 may have a speed of 0.2 m / s to 20 m / s, preferably 0.4 m / s to 15 m / s, and most preferably 1 m / s to 10 m / s.

[0031] System 1 also includes a measuring unit 3, a laser unit 4B, and a sorting unit 5 arranged along the conveyor 2 from the input side 1' to the output side 1'". The system further includes a processing unit 6. The system may also include a positioning unit 4A arranged between the measuring unit 3 and the laser unit 4B. The measuring unit 3, positioning unit 4A, laser unit 4B, and sorting unit 5 are coupled to the processing unit 6. Preferably, the measuring unit 3, positioning unit 4A, laser unit 4B, and sorting unit 5 are placed on the transport surface of the conveyor 2. In operation, a flow of objects containing metal is transported by the conveyor 2 from the input side 1' to the output side 1', passing through the measuring unit 3, positioning unit 4A, laser unit 4B, and sorting unit 5 respectively. The measuring unit 3, positioning unit 4A, and laser unit 4B will be referred to below. Figure 1B To elaborate further.

[0032] Sorting unit 5 is arranged to sort metal-containing objects into one of two or more output stages based on instructions from processing unit 6. Additionally, sorting unit 5 includes sorting devices, such as one or more robotic arms, one or more push rods, or one or more nozzles, or any other means capable of moving objects. The sorting devices are arranged to move metal-containing objects from a stream of metal-containing objects into the desired output stage. For each output stage, sorting unit 5 may be coupled to a hopper, container, or chute to collect metal-containing objects from the output stage.

[0033] One or more robotic arms may include means for applying suction or mechanical gripping to an object containing metal. Each robotic arm can thus grip the object containing metal and move it to a desired output stage. One or more push rods may include mechanical push rods, pneumatic push rods, spring-loaded push rods, and / or hydraulic push rods. Each push rod can thus push the object containing metal and move it to the desired output stage. One or more nozzles may be arranged to inject pressurized gas to move the object containing metal into the desired stage. Each nozzle includes a valve for opening or closing the nozzle, and one or more nozzles may be activated simultaneously, for example, according to the mass of the object containing metal. In operation, the object containing metal is moved by the sorting device to one of two or more output stages and thus exits system 1.

[0034] Processing unit 6 includes circuitry 6A. Circuitry 6A includes at least one processor, such as a central processing unit (CPU), microcontroller, microprocessor, graphics card, or field-programmable gate array (FPGA). Circuitry 6A also includes working memory, I / O modules, network connectivity modules, and optionally at least one screen. The processor is arranged to execute program code stored in memory to perform measurement operations, positioning operations, object detection operations, dispensing operations, cleaning operations, drying operations, transport operations, processing operations, and / or sorting operations described herein.

[0035] The processing unit may optionally be coupled to the drive unit 2A. In operation, the speed of the conveyor 2 can be adjusted by the processing unit 6 sending a control signal to the drive unit 2A. The conveyor speed may be set based on the throughput of metal-containing objects or other operational criteria. The circuit 6A is configured to receive the measured contamination level of the metal-containing objects from at least one measuring unit 3. The circuit 6A is also configured to: locate at least one point on the surface of the metal-containing objects where the contamination level is below a threshold level; and transmit the at least one located point to the laser unit 4B, which is further detailed below. The processing unit 6 may be a local processing unit located near other units in the system. Alternatively, the processing unit 6 may be a remote processing unit. The processing unit 6 may be coupled to other units in the system by means of a cable connection and / or a wireless connection. Further aspects of the processing unit 6 and the circuit 6A will be described below.

[0036] refer to Figure 1BThe measurement unit 3, positioning unit 4A, and laser unit 4B will now be described. The measurement unit 3 includes at least one light source 3A and a sensor 3B. The light source 3A may include one or more LEDs, halogen lamps, or bulbs. The light source 3A is arranged to illuminate the area on the conveyor 2 scanned by the sensor 3B. Advantageously, the sensor can thereby obtain a clear and bright image of objects containing metal, thereby improving the accuracy of contamination level detection. The sensor 3B can be configured to detect light radiation within wavelength intervals of 100 nm to 1000 nm, 400 nm to 1100 nm, and / or 1100 nm to 1900 nm. The sensor 3B can be adapted to detect visible light, NIR light, IR light, UV light, or combinations thereof. Preferably, the sensor 3B may include a visible light sensor, such as an RGB (red, green, blue) sensor, a UV sensor, an IR sensor, and / or a NIR (near-infrared) sensor. According to one example, a first sensor may be adapted to detect UV light or primarily detect UV light. A second sensor may be adapted to detect visible light or primarily detect visible light. The third sensor is adapted to detect NIR light. Sensor 3B is configured to capture one or more digital images of an object containing metal. Light source 3A and sensor 3B are coupled to processing unit 6. Measurement unit 3 is configured to send one or more images of the object containing metal to processing unit 6.

[0037] Processing unit 6 is configured to process one or more received images and locate one or more points on an object containing metal where the contamination level is below a threshold. The threshold may be, for example, 99% contamination, 90% contamination, 75% contamination, or 50% contamination. In an image, a point may be represented by a single pixel or by multiple clustered pixels and / or neighboring pixels. Therefore, a point may refer to a point or region on an object containing metal. Optionally, processing unit 6 may be configured to perform foreground / background segmentation on one or more images before locating one or more points. This allows the location of pixels associated with the object containing metal but not with the image background. Further optionally, processing unit 6 may be arranged to perform connected component analysis. This determines which pixels are connected and depicts the object in one or more images.

[0038] Processing unit 6 can also be configured to locate one or more points on an object containing metal based on the processed image, using color, color gradient, texture, shape, and / or thermal features. Color may include, for example, hue, saturation, and / or brightness (HSL). Alternatively, color may include hue, saturation, and / or value (HVL). Processing unit 6 can be arranged to compare the color, color gradient, texture, shape, and / or thermal features of the object containing metal with one or more reference values ​​to determine whether the points have a contamination level below a threshold level. The reference values ​​may be stored by processing unit 6.

[0039] Alternatively, the circuitry 6A of processing unit 6 may include a machine learning protocol for locating points where the contamination level is below a threshold. The machine learning can be trained using sample data, such as images of objects containing metal, images of regions containing metal, and / or pixel-level data of objects containing metal. Training may include supervised or unsupervised learning.

[0040] The located point is transmitted from the processing unit 6 to the laser unit 4B and optionally to the positioning unit 4A.

[0041] Continue to refer to Figure 1B The positioning unit 4A is arranged to receive information from the processing unit 6 regarding one or more located points on an object containing metal. The positioning unit 4A is also arranged to perform object detection on the object containing metal. Object detection generates position information for one or more located points on the object containing metal. The position information may include two-dimensional position information. The two-dimensional position information is preferably provided in a two-dimensional coordinate system parallel to and inert relative to the conveyor surface. The position information is stored by the processing unit 6. Advantageously, by providing accurate position information, spectral analysis (described in detail below) can be performed at the desired and correct location on the object containing metal.

[0042] Object detection may include performing laser triangulation at one or more located points. Alternatively or additionally, object detection may include analyzing images of objects containing metal. Image analysis may include analyzing color images, UV images, NIR images, X-ray images, and / or spectral images. Image analysis may be performed by machine learning algorithms or by other automated software. For example, machine learning may include deep learning algorithms and / or supervised learning algorithms. Color image analysis and / or spectral image analysis may include image segmentation. In each case, image analysis produces two-dimensional positional information, as described above for laser triangulation. Processing unit 6A may be arranged to perform triangulation calculations and / or image analysis based on data received from positioning unit 4A.

[0043] Object detection may also include height detection. Height detection preferably includes detecting the height of one or more located points relative to the conveyor surface. Therefore, height detection is preferably performed in a direction normal to the conveyor surface. Height detection may be based on time-of-flight technology, structured light technology, optical parallax technology, and / or the output of one or more height detectors (detailed below). Height detection allows objects with substantially different heights or complex three-dimensional geometries to be processed by the system, which could lead to detection errors in planar object recognition techniques. Therefore, pre-filtering or pre-sorting of objects containing metal is not required before feeding them into the system, making the system less complex. Furthermore, height detection largely prevents problems associated with objects containing metal that are the same color as the conveyor surface. Two-dimensional object recognition techniques based on RGB images may have the problem of failing to distinguish between the conveyor surface and the object containing metal when both are the same color. By using height detection, it can be determined whether a point is located on the conveyor surface or on an object carried by the conveyor surface.

[0044] The detected height information is stored by processing unit 6. The height information can be combined with two-dimensional position information. This provides three-dimensional position information for the located point. Advantageously, providing three-dimensional position information reduces problems associated with partially or completely overlapping objects, which can hinder two-dimensional object recognition techniques such as edge detection or color distribution. Therefore, it is unnecessary to separate or filter the feed of objects containing metal to avoid object overlap. The three-dimensional position information is used for laser-induced breakdown spectroscopy (LIBS), detailed below. Alternatively, the three-dimensional position information can be used for Raman spectroscopy, detailed below. Advantageously, the spectral analysis can thus be precisely focused on the located point, thereby generating accurate spectral data.

[0045] In another embodiment, processing unit 6 can be configured to calculate geometric data using laser triangulation and / or image analysis. The geometric data may include the contours, contour lines, surface area, surface normals, surface curvature, centroid, roughness, and / or surface texture of the metal-containing object. Advantageously, providing geometric data can further improve the differentiation of overlapping or partially overlapping objects, the handling of objects with complex shapes, and the handling of objects with varying dimensions. Processing unit 6 can be configured to combine the geometric data with one or more processed images from measurement unit 3. Combining the geometric data with one or more processed images enables processing unit 6 to identify which located points and / or which metal-containing objects are the best candidates for performing spectral analysis (described in detail below). For example, if the surface normals of the metal-containing object point away from the field of view of the spectrometer (described in detail below), the acquired spectral data may be of lower quality. Advantageously, from a batch of metal-containing objects, only the most suitable objects can be selected for spectral analysis. Alternatively and / or if multiple points have been located, the processing unit 6 may be arranged to select only one point and send the selected point to the positioning unit 4A.

[0046] The positioning unit 4A may include a light source and a scanner for object detection. The light source may include a laser 4A', which is arranged to emit at least one laser beam 4A. Alternatively, the light source may include one or more LEDs, halogen lamps, UV lamps, etc. The scanner may include at least one camera 4A. Alternatively or additionally, the positioning unit 4A may include a structured light scanner, X-ray scanner, spectral imaging scanner, or ultrasonic scanner. Further alternatively, the positioning unit 4A may include at least two cameras to enable stereoscopic imaging and / or optical parallax measurement. Optionally, the positioning unit 4A may include one or more height detection units, such as a height sensor array.

[0047] The light source and scanner are coupled to the processing unit 6. The laser beam 4A emitted by the laser 4A'... It can be a line laser. Laser beam 4A It can cover the width of conveyor 2. The light source and scanner are preferably aimed at the same line or the same area on the conveyor surface. Both the light source and scanner can be arranged in a fixed orientation. Alternatively, the orientation of the light source and / or scanner can be adjustable. The positioning unit 4A may, for example, include components for directing light emitted by the light source (such as a laser beam 4A)... An adjustable reflector or array of adjustable reflectors guides the desired line or area onto the conveyor surface. The scanner can be mounted on a movable ball head or otherwise adjustable. For example, a laser beam 4A... The angle between the field of view of camera 4A and conveyor 2 can be recorded by processing unit 6. Thus, processing unit 6 can perform triangulation calculations.

[0048] Additionally, the positioning unit 4A can be arranged to perform altitude detection of one or more positioned points. Altitude detection can be based on time-of-flight technology, structured light technology, optical parallax technology, and / or the output of an altitude detector.

[0049] Processing unit 6 is arranged to calculate the position of the located point based on laser triangulation and / or image analysis, along with optional height detection. Advantageously, the precise three-dimensional position of the located point can be obtained, which facilitates spectral analysis at the correct location and thus enables accurate spectral analysis and correct object classification.

[0050] Continue to refer to Figure 1B The laser unit 4B is arranged to receive the two-dimensional and / or three-dimensional positions of one or more located points from the processing unit 6. The laser unit 4B is also arranged to perform spectral analysis on one or more located points. The laser unit 4B can be as follows: Figure 1B The LIBS unit is shown. Alternatively, the laser unit 4B can be a Raman spectroscopy analysis unit, such as a surface-enhanced Raman spectroscopy (SERS) unit. The laser unit 4B may include a scanner 4B', a laser focusing unit 4B"", a laser 4B'"", and a spectrometer 4C. The scanner 4B', the laser focusing unit 4B"", the laser 4B'"", and the spectrometer 4C are coupled to the processing unit 6.

[0051] The LIBS unit can be arranged to locally form plasma at the located point. Additionally, laser 4B'" is configured to deliver laser beam 4B The light is guided to a designated point, thereby locally forming plasma on the surface of an object containing metal. The Raman spectroscopy analysis unit can be arranged to generate a light signal comprising multiple Raman wavelengths at the designated point. Additionally, laser 4B'" is configured to direct the laser beam 4B onto the surface of the object containing metal. The laser beam is directed to the designated point to generate an optical signal. Laser 4B'" can be, for example, an Nd:YAG solid-state laser, or any other suitable solid-state laser, or a gas laser. Laser 4B'" can emit a laser beam 4B with wavelengths in the optical, near-infrared, or near-ultraviolet domains. Laser beam 4B The laser beam 4B can be guided to the desired point by the scanner 4B' and the laser focusing unit 4B" based on instructions from the processing unit 6. The scanner 4B' may include one or more adjustable mirrors and / or galvanometers. The laser focusing unit 4B" is arranged to focus the laser beam 4B The system focuses on the minimum radius, line, or region at the designated point. This allows for the localized formation of plasma or optical signals comprising multiple Raman wavelengths on objects containing metal.

[0052] Spectrometer 4C is arranged to measure spectral data of a plasma, or spectral data of an optical signal comprising multiple Raman wavelengths. Alternatively, spectrometer 4C can measure light emitted by locally formed plasma. Alternatively, spectrometer 4C can detect multiple Raman wavelengths included in the optical signal. Spectrometer 4C can be configured to analyze optical radiation with wavelength intervals of 100 nm to 1000 nm, 400 nm to 1100 nm, and / or 1100 nm to 1900 nm. Spectrometer 4C can be adapted to analyze visible light, NIR light, IR light, UV light, or combinations thereof. Spectrometer 4C can be arranged to scan multispectral data or hyperspectral data. Multispectral data may include non-adjacent spectral bands. Hyperspectral data may include adjacent spectral bands. Circuitry 6A of processing unit 6 is arranged to receive spectral data from spectrometer 4C and assign categories to objects containing metals based on the received spectral data.

[0053] Optionally, processing unit 6 can be arranged to perform normalization of the received spectral data, including temperature compensation and / or dark / white calibration. Further optionally, processing unit 6 can be arranged to preprocess the spectral data. Preprocessing may include spectral binning and / or smoothing. Advantageously, preprocessing can improve the quality of the spectral data at the spectral or pixel level. Processing unit 6 can also be arranged to perform data reduction on the spectral data. Data reduction can be performed by applying principal component analysis, by expert-driven data reduction, and / or by applying partial least squares. Expert-driven data reduction may, for example, involve selecting only relevant spectral bands.

[0054] Processing unit 6 can also be arranged to compare spectral data with previously measured spectral data of the test object, wherein the metallic or alloy composition of the test object is known in detail. Alternatively, processing unit 6 can be configured to compare spectral data with a database having known spectral properties of metallic or alloy materials. In each case, processing unit 6 is arranged to assign categories based on the comparison.

[0055] System 1 can be arranged to output objects containing metal, including a desired metal or alloy. Therefore, circuit 6A can be arranged to assign one of a first category A and a second category B to the objects containing metal. In operation, the first category A is assigned when the desired metal or alloy is present, and the second category B is assigned when the desired metal or alloy is not present. In another embodiment, system 1 can be arranged to output several different metals and / or alloys. The first category can then be divided into two or more subcategories. Circuit 6A can be arranged to assign one of the subcategories when the desired metal or alloy is present, or to assign the second category B when the desired metal or alloy is not present. Each subcategory is thus associated with a different output level.

[0056] refer to Figure 1A System 1 may also include an optional return conveyor 7 for returning metal-containing objects from sorting unit 5 to the system's input side 1'. The return conveyor 7 is schematically indicated by a dashed arrow in Figure 1. The return conveyor 7 may include one or more conveyor belts. Sorting unit 5 may be configured to divert metal-containing objects with contamination levels above a threshold level to the return conveyor 7. In operation, these contaminated metal-containing objects can then be transported back to the input side 1' by the return conveyor 7. The contaminated metal-containing objects can then be cleaned in an optional cleaning unit, detailed below, before re-entering measuring unit 3.

[0057] The system may also include an optional cleaning unit 8 for cleaning contaminated metal-containing objects transported by conveyor 2. Cleaning unit 8 is preferably arranged along conveyor 2 at the input side 1'. Cleaning unit 8 is coupled to processing unit 6. Cleaning unit 8 may include means for grinding, spraying liquids, mechanical cleaning, chemical cleaning, high-pressure water cleaning, laser-based cleaning, or any combination thereof. The means for mechanical cleaning may, for example, include one or more brushes for scrubbing the metal-containing objects. Cleaning unit 8 may also include a bypass, allowing uncontaminated metal-containing objects transported on conveyor 2 to bypass cleaning unit 8. In operation, cleaning unit 8 can receive contaminated metal-containing objects from the flow of metal-containing objects and / or from the return conveyor 7. Cleaning unit 8 may receive signals from processing unit 6 to clean or stop cleaning metal-containing objects in the flow of metal-containing objects.

[0058] The system may also include an optional drying unit 9 for drying metal-containing objects exiting the cleaning unit 8. The drying unit 9 is preferably positioned between the cleaning unit 8 and the measuring unit 3. The drying unit 9 is coupled to the processing unit 6. The drying unit 9 may include means for blowing, heat drying, or a combination thereof. The drying unit 9 may also include a bypass, allowing metal-containing objects transported by the conveyor 2 to bypass the drying unit 9. In operation, the drying unit 9 receives signals from the processing unit 6 to dry or stop drying the metal-containing objects in the flow.

[0059] The system may also include an optional detection unit 10, which is arranged at the return conveyor 7. The detection unit 10 is coupled to the processing unit 6. The detection unit 10 is arranged to send a signal to the processing unit 6 when the number of metal-containing objects on the return conveyor 7 reaches a predetermined threshold. The processing unit 6 is configured to, upon receiving the signal from the detection unit 10, instruct the sorting unit 5 to output the metal-containing objects to the second-stage sorting B instead of the return conveyor 7. This advantageously prevents the overflow of metal-containing objects from the conveyor 2.

[0060] According to another embodiment, the system may further include Figure 3 An auxiliary measuring unit 11 is schematically shown. The auxiliary measuring unit 11 is arranged along the conveyor 2 at the input side 1' of the conveyor. The auxiliary measuring unit 11 is coupled to the processing unit 6. In operation, the contamination level of the metal-containing object can be measured in the auxiliary measuring unit 11. Metal-containing objects with high contamination levels can optionally be cleaned in the cleaning unit 8 and further optionally dried in the drying unit 9 before entering the measuring unit 3. Conversely, metal-containing objects with low contamination levels can bypass the cleaning unit 8 and the drying unit 9 and move directly to the measuring unit 3. The auxiliary measuring unit 11 can be coupled to the cleaning unit 8 via the processing unit 6. In operation, contamination information measured by the auxiliary measuring unit 11 (such as the contamination level or location on the metal-containing object) can be transmitted to the cleaning unit 8. This allows for more efficient cleaning, where cleaning can be performed on the correct location on the metal-containing object or with the desired intensity based on the contamination level.

[0061] The return conveyor 7 can connect the sorting unit 5 to an optional cleaning unit 8. Alternatively, the return conveyor 7 can connect the sorting unit 5 to an auxiliary measuring unit 11. The auxiliary measuring unit 11 is configured and functions as described above for the measuring unit 3, and includes at least one light source and a camera.

[0062] Figure 2A schematic flowchart of the method of the present invention for sorting metal-containing objects in a stream of metal-containing objects is shown. According to the method, an input stream I of metal-containing objects is transported from input side 1' to output side 1'". An unsorted stream of metal-containing objects is provided at input side 1'. At output side 1', the stream is divided into two or more grades as detailed below. The stream of metal-containing objects passes through a measurement unit 3, a laser unit 4, and a sorting unit 5 as detailed above. The stream of metal-containing objects may also pass through an optional cleaning unit 8, an optional drying unit 9, and / or an optional auxiliary measurement unit 11, each detailed above.

[0063] The input stream I containing metallic objects may include a mixed waste recycling stream containing both metallic and non-metallic objects. Alternatively, the input stream I containing metallic objects may include a pre-sorted waste stream that primarily or solely comprises metallic objects. Further alternatively, the input stream I may include a waste stream originating directly from the metal industry.

[0064] Each metal-containing object may have a metal fraction of at least 5% by weight and / or by volume, preferably at least 10%, more preferably at least 20%, and most preferably at least 50%. Advantageously, a higher metal fraction makes the recycling process more efficient. Metal-containing objects may, for example, comprise composites, such as objects containing metal particles embedded in a matrix material.

[0065] The metal contained in each object may include one or more pure metals and / or one or more metal alloys. Pure metals may include aluminum, copper, chromium, iron, magnesium, molybdenum, nickel, silver, or titanium. Alloys may include aluminum alloys, brass, bronze, gunmetal, nickel alloys, solder, steel, stainless steel, or titanium alloys.

[0066] Objects containing metal in input stream I may be contaminated. Contamination includes surface contamination and may include foreign particles or foreign layers covering part or all of the surface of the object containing metal. Foreign particles may include dirt, debris, dust, rust, etc. Foreign layers may include one or more layers of paint, coating, grease, or other materials or compositions. Contamination may partially or completely cover the surface of the object containing metal. Therefore, contamination may adversely affect the sorting and sorting process.

[0067] For one or more metal-containing objects in flow I, the surface contamination level is measured. The contamination level measurement is performed by measurement unit 3 as detailed above and / or optional auxiliary measurement unit 11. The contamination level measurement is performed by capturing one or more images of the metal-containing objects. The one or more images may be, for example, visible light images, such as RGB images, UV images, or NIR images. The one or more images are processed to locate one or more points on the surface of each metal-containing object where the contamination level is below a threshold level. This processing is performed by processing unit 6 as detailed above.

[0068] Optionally, object detection can be performed by the localization unit 4A as described above. Object detection may include laser triangulation at one or more localized points. Alternatively or additionally, object detection may include analysis of an image of an object containing metal. Image analysis may include analysis of color images, UV images, NIR images, X-ray images, and / or spectral images. Image analysis may be performed by automated software including deep learning and / or supervised machine learning algorithms. Image analysis may include image segmentation, such as foreground / background segmentation.

[0069] Object detection generates position information. This position information is preferably provided in a two-dimensional coordinate system that is parallel to and inert relative to the conveyor surface.

[0070] Object detection can also include height detection. Height detection generates height information, as detailed above, in the direction of the surface normal above the conveyor surface. This allows for the processing of objects with substantially different heights or complex shapes. Such objects can lead to detection errors in purely 2D object recognition techniques. Therefore, this method does not require pre-filtering or pre-sorting of objects. Furthermore, height detection can detect objects with the same color as the conveyor surface. For 2D object recognition techniques, such as those based on RGB images, it may be impossible to distinguish the conveyor surface and the object with a sufficient level of confidence when they have the same color.

[0071] Altitude detection can be based on time-of-flight technology, structured light technology, optical parallax technology, and / or the output of one or more altitude detectors. Three-dimensional position information can be obtained by combining altitude information with two-dimensional position information. Three-dimensional position information is particularly advantageous when applying LIBS (Liquidity-Induced Broadband) lasers. This allows the LIBS laser to be precisely focused on the located point. Three-dimensional position information can also be used to apply Raman spectroscopy analysis, guiding the laser to the located point. Furthermore, three-dimensional position information avoids problems associated with partially or completely overlapping objects. Such overlapping objects can be difficult to distinguish using two-dimensional techniques (such as edge detection or color distribution). Therefore, with current methods, there is no need to separate or filter the feed of objects containing metal to avoid object overlap.

[0072] In another implementation, geometric data can be calculated based on data obtained from laser triangulation and / or image analysis. Geometric data may include the outline, contour lines, surface area, surface normals, surface curvature, centroid, roughness, and / or surface texture of the object containing metal. The object's geometric data can be combined with one or more processed images generated from contamination detection of the same object. The combination of geometric data and processed image data can determine the suitability of the located points and / or the object containing metal for spectral analysis. For example, from a batch of objects containing metal, only the most suitable objects can be selected for spectral analysis.

[0073] A laser beam is directed to a designated point to locally form a plasma. Alternatively, the laser beam is directed to the designated point to form an optical signal comprising multiple Raman wavelengths. The laser beam is emitted by a laser 4B'” of laser unit 4B as described above. The laser beam may have wavelengths in the optically visible, NIR, or near-UV domains. The light emitted by the plasma is detected by a spectrometer 4C. Alternatively, an optical signal comprising multiple Raman wavelengths is detected by a spectrometer 4C.

[0074] Next, the spectrum of the plasma is measured. Alternatively, the spectrum including the Raman wavelengths in the optical signal is measured. The spectrum is measured by spectrometer 4C. The measured spectrum may include multispectral data or hyperspectral data. Multispectral data may include non-adjacent spectral bands. Hyperspectral data may include adjacent spectral bands. The spectrum is transmitted to processing unit 6.

[0075] In another implementation, the spectrum can be measured at several located points. Therefore, one or more located points with the most suitable spectral data can be selected. The most suitable spectral data may, for example, include spectral data with minimal noise, spectral data with the clearest spectral characteristics of a metal or alloy, and so on. Thus, the method can be optimized and more reliably sorted.

[0076] Next, based on the measured spectra, objects containing metals are classified. Classification is performed by processing unit 6 as described above. The spectral data of the measured spectra can be normalized, preprocessed, and / or dimensionality reduced before classification. Normalization may include temperature compensation and / or dark / white calibration. Preprocessing may include spectral binning and / or smoothing. Advantageously, preprocessing can improve the quality of the spectral data at the spectral or pixel level. Dimensionality reduction of the spectral data may include applying principal component analysis, applying expert-driven data dimensionality reduction, and / or applying partial least squares. Expert-driven data dimensionality reduction may, for example, involve selecting only relevant spectral bands.

[0077] Spectral data can be compared with previously measured spectral data for test objects composed of known metals or alloys. Alternatively, spectral data can be compared with a database of known spectral properties including different metals and / or alloys. Based on this comparison, categories can be assigned to objects containing metals. A first category can be assigned when the desired metal or alloy is present. A second category can be assigned when the desired metal or alloy is not present. Classification can be performed by machine learning algorithms. Machine learning algorithms can be trained to classify objects containing metals using supervised learning, unsupervised learning, or hybrid learning methods.

[0078] Alternatively, the first category can be divided into two or more subcategories. Thus, objects can be classified into two or more categories with different metals and / or alloys. Then, when a first desired metal or alloy is present, one of the subcategories can be assigned, and when a second desired metal or alloy is present, a second category can be assigned, where the second desired metal or alloy differs from the first desired metal or alloy. Each subcategory is thus associated with a different output metal or alloy.

[0079] The method further includes sorting objects containing metal into one of two or more output levels based on the assigned category. Sorting is performed by sorting unit 5. Figure 2 As shown, two or more output stages may include, for example, a first output stage A and a second output stage B. The first output stage A may include recyclable metal-containing objects and, for example, include a sufficient alloy content. The second output stage B may include non-recyclable metal-containing objects and, for example, include an insufficient alloy content. For example, the metal-containing objects in output stage B may be discarded. For example, the metal-containing objects in output stage A may be used as raw materials for further industrial processing.

[0080] Alternatively, one or more metal-containing objects can be added to the return stage C. The metal-containing objects in the return stage C may be too contaminated to be sorted into the output stage. The return stage C can then return to the input stream I. Thus, the return stage C can undergo a new cycle of optional cleaning, drying, and contamination detection. Advantageously, this ensures that the output stage includes metal-containing objects that are clean enough for sorting and classification.

[0081] refer to Figure 3 According to another embodiment, the contamination level of metal-containing objects can be detected before entering the cleaning and / or drying cycle. Additionally, the flow of metal-containing objects can pass through an optional auxiliary measuring unit 11 as described above.

[0082] Advantages of this invention include the ability to selectively apply cleaning processes to contaminated areas, thereby achieving energy efficiency and cost-effectiveness. Drying is optional, further improving energy efficiency. The laser can be selectively routed, providing enhanced reliability.

[0083] Adaptive laser power control can be based on contamination sensor data, such as contamination or pre-classification data.

[0084] The inventive concept can also include automated transfer learning. Such learning can be performed by an AI model. As long as the sorted metal-containing objects look similar, the model can be shared across multiple locations.

[0085] It should be noted that the steps or actions in the method described above can be performed in any suitable order, and therefore not only in the order given above. Furthermore, one or more of these steps or actions can be performed in parallel. It should also be noted that these steps or actions can be performed by different equipment at different times and / or at different sites. In other words, as an example, the method can be executed in a distributed manner at multiple sites, with different steps or actions performed at different points in time. However, the method can advantageously be performed at a single site in the order described above.

[0086] Furthermore, those skilled in the art, when practicing the claimed invention, can understand and implement variations of the disclosed variants by studying the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements, and the indefinite articles "a" or "an" do not exclude multiple elements. The fact that certain measures are described in mutually different dependent claims does not indicate that combinations of these measures cannot be used to achieve an advantage.

[0087] Figure Labels

Claims

1. A method for sorting metal-containing objects in a stream of metal-containing objects, the method comprising: - measuring a contamination level on a surface of the metal-containing objects; - locating at least one point on the surface where the contamination level is below a threshold level; o directing a laser beam at the located point to locally form a plasma and measure a spectrum of the plasma; or o directing a laser beam at the located point to form a light signal comprising a plurality of Raman wavelengths and measure a spectrum of the Raman wavelengths; the method further comprising: - assigning a class to the metal-containing objects based on the measured spectrum; and - sorting the metal-containing objects into one of a first fraction (A) and a second fraction (B) based on the assigned class.

2. The method of any of claim 1, wherein, The contamination level is measured by a visible light sensor, a UV sensor and / or a NIR sensor.

3. The method of claim 1 or 2, wherein, Measuring the spectrum of the plasma is performed by LIBS, or wherein measuring the spectrum of the plurality of Raman wavelengths is performed by Raman spectroscopy, such as SERS.

4. The method of any one of claims 1 to 3, wherein, Locating at least one point further comprises object detection by laser triangulation and / or image analysis.

5. The method of claim 4, wherein, Object detection further comprises height detection based on time-of-flight techniques, structured light techniques, optical parallax techniques and / or output of a height detector.

6. The method of any one of claims 1 to 6, further comprising: The object is cleaned before measuring the contamination level and optionally dried.

7. The method of claim 6, further comprising: If no point with a contamination level below the threshold level can be located, the object is recleaned.

8. The method of claim 6 or 7, wherein, Cleaning and / or recleaning comprises sanding, spraying a liquid, mechanical cleaning, chemical cleaning, high-pressure water cleaning, laser-based cleaning or any combination thereof.

9. The method of any one of claims 1 to 8, further comprising: If multiple points on the surface have a contamination level below the threshold level, one point is selected from the multiple points and the laser beam is directed to the selected point.

10. The method of any one of claims 1 to 9, wherein, Directing the laser beam is performed by manipulating a mirror of a movable arrangement that reflects the laser beam.

11. The method of any one of claims 1 to 10, wherein, Assigning a class is performed by a machine learning algorithm.

12. The method of any one of claims 1 to 11, wherein, Sorting the metal-containing objects into one of a first fraction (A) and a second fraction (B) is based on one of a metal property of the metal-containing objects and a metal property of an alloy composition of the metal-containing objects.

13. A system (1) for sorting metal-containing objects in a stream of metal-containing objects, the system comprising: a measurement unit (3) for measuring a contamination level on a surface of the metal-containing objects; a laser unit (4B) comprising a laser direction unit (4B’) and a spectrometer (C); and a processing unit (6) comprising a circuit (6A) configured to: - receive the measured contamination level from at least one of the measurement units (3); - locate at least one point on the surface where the contamination level is below a threshold level; and - transmit the located point to the laser direction unit (4B’); wherein the laser direction unit (4A) is configured to form a plasma at the located point or to form a light signal comprising a plurality of Raman wavelengths; wherein the spectrometer (4C) is configured to measure a spectrum of the plasma, or to measure a spectrum of the plurality of Raman wavelengths; wherein the circuitry (6A) is further configured to: - receive the measured spectrum from the spectrometer (4C); and - assign a class to the metal-containing object based on the received spectrum; and wherein the system (1) further comprises: at least one sorting unit (5) for sorting the metal-containing object into one of a first fraction (A) and a second fraction (B) based on the assigned class.

14. The system of claim 13, wherein, The measurement unit (3) comprises a sensor (3B), such as a visible light sensor, a UV sensor, and / or a NIR sensor.

15. The system of claim 13 or 14, wherein, The laser unit (4B) is a LIBS unit, or is a Raman spectroscopy unit, such as a SERS unit.

16. The system according to any one of claims 13 to 15, further comprising a cleaning unit (8) and optionally a drying unit (9).

17. The system of claim 16, wherein, The cleaning unit (8) is one of an abrasive unit, a liquid spray unit, a mechanical cleaning unit, a chemical cleaning unit, a high-pressure water cleaning unit, a laser-based cleaning unit, or any combination thereof.

18. The system of any one of claims 13 to 17, wherein, The circuitry (6A) is further configured to select a point from the plurality of points and direct the laser beam to the selected point if the plurality of points on the surface have a contamination level below the threshold level.

19. The system of any one of claims 13 to 18, wherein, The laser direction unit (4A) comprises at least one moveably arranged mirror.

20. The system of any one of claims 13 to 19, wherein, The circuitry (6A) is further configured to assign a class by means of a machine learning algorithm.

21. The system of any one of claims 13 to 20, wherein, The sorting unit (5) is configured to sort the metal-containing object based on one of a metal property of the metal-containing object and a metal property of an alloy composition of the metal-containing object.

Images