Fractionation Based on Chemical Composition

Abstract

A system and method for sorting and separating materials to create a collection of materials that are generally composed of a particular chemical composition. The system may utilize a vision system and one or more sensor systems that may implement machine learning systems to identify or classify each material. Separation is then performed according to the classification.

Description

Technical Field

[0001] This application claims priority to U.S. Provisional Patent Application No. 63 / 249,069 and U.S. Provisional Patent Application No. 63 / 285,964. This application is a continuation-in-part of U.S. Patent Application No. 17 / 667,397, and claims priority to U.S. Provisional Patent Application No. 63 / 146,892 and U.S. Provisional Patent Application No. 63 / 173,301. This application is a continuation-in-part of U.S. Patent Application No. 17 / 495,291, which is a continuation of U.S. Patent Application No. 17 / 380,928, which is a continuation-in-part of U.S. Patent Application No. 17 / 227,245, which is a continuation-in-part of U.S. Patent Application No. 16 / 939,011, which is a continuation of U.S. Patent Application No. 16 / 375,675 (issued as U.S. Patent No. 10,722,922), which is a continuation of U.S. Patent Application No. 15 / 963,755 (issued as U.S. Patent No. 10,710,119), and claims priority to U.S. Provisional Patent Application No. 62 / 490,219, which is a continuation-in-part of U.S. Patent Application No. 15 / 213,129 (issued as U.S. Patent No. 10,207,296) and claims priority to U.S. Provisional Patent Application No. 62 / 193,332, all of which are hereby incorporated by reference herein. U.S. Patent Application No. 17 / 495,291 is also a continuation-in-part of U.S. Patent Application No. 17 / 491,415 (issued as U.S. Patent No. 11,278,937), which is a continuation-in-part of U.S. Patent Application No. 16 / 852,514 (issued as U.S. Patent No. 11,260,426), which is a divisional of U.S. Patent Application No. 16 / 358,374 (issued as U.S. Patent No. 10,625,304), which is a continuation-in-part of U.S. Patent Application No. 15 / 963,755 (issued as U.S. Patent No. 10,710,119), all of which are hereby incorporated by reference herein.

[0002] Government Licensing Rights This disclosure was made with government support under grant number DE-AR0000422 awarded by the U.S. Department of Energy. The U.S. government may have certain rights in this disclosure.

[0003] This disclosure generally relates to the separation of materials, and more particularly to the separation of materials to achieve a specific composition of chemical elements within the separated materials. BACKGROUND OF THE DISCLOSURE

[0004] Recycling is the process of collecting and processing materials that would otherwise be discarded as waste and turning them into new products. Recycling benefits local communities and the environment by reducing the amount of waste sent to landfills and incinerators, conserving natural resources, enhancing economic security by utilizing domestic material sources, preventing pollution by reducing the need to collect new raw materials, and saving energy. After collection, recyclable materials are typically sent to a materials recovery facility, where they are sorted, cleaned, and processed into materials that can be used in manufacturing. PRIOR ART DOCUMENTS PATENT DOCUMENTS

[0005] PATENT DOCUMENT 1 U.S. Published Patent Application No. 2022 / 0016675 NON-PATENT DOCUMENTS

[0006] NON-PATENT DOCUMENT 1 Krizhevsky et al., "ImageNet Classification with Deep Convolutional Networks", Proceedings of the 25th International Conference on Neural Information Processing Systems, December 3 - 6, 2012, Lake Tahoe, Nevada NON-PATENT DOCUMENT 2 LeCun et al., "Gradient-Based Learning Applied to Document Recognition", Proceedings of the IEEE, Institute of Electrical and Electronics Engineers (IEEE), November 1998

Summary of the Invention

Means for Solving the Problems

[0007] Aspects of the present disclosure include a step of determining an approximate mass of each of a plurality of pieces of material, wherein at least one of the plurality of pieces of material has a different material classification from other pieces of material; a step of classifying each of the plurality of pieces of material as belonging to one of a plurality of different material classifications; and a step of separating a particular one of the plurality of pieces of material from the plurality of pieces of material according to the determined approximate mass and classification of each of the plurality of pieces of material, wherein the separation generates a set of pieces of material having a chemical composition of a predetermined particular aggregate. The separation may include redirecting a particular one of the pieces of material into a container. The separation may include continuously determining the chemical composition of the aggregate of the redirected piece of material. The separation may include redirecting the next piece of material into the container to increase the weight percentage of a particular chemical element of the chemical composition of the aggregate of the redirected piece of material. The separation may include not redirecting the next piece of material into the container to reduce the weight percentage of a particular chemical element of the chemical composition of the aggregate of the redirected piece of material. The separation may include not redirecting the next piece of material into the container because the next piece of material contains undesirable contaminants within the chemical composition of the predetermined particular aggregate. The separation may continue until the chemical composition of the aggregate of a predetermined minimum number of redirected pieces of material equals a threshold level of the chemical composition of the predetermined particular aggregate. The set of pieces of material having the chemical composition of the predetermined particular aggregate may include at least one piece of material having a different material classification from other pieces of material within the set. The plurality of pieces of material may include pieces of material having different metal alloy compositions. The chemical composition of the predetermined particular aggregate may be different from the chemical composition of each of the plurality of pieces of material. The chemical composition of the predetermined particular aggregate may be different from the chemical composition of all of the aggregates of the plurality of pieces of material. The set of pieces of material may include pieces of material having different material classifications. The set of pieces of material may include at least one of the pieces of material having a different material classification from other pieces of material.The plurality of pieces of material may include brazing aluminum alloy pieces and cast aluminum alloy pieces, the assembly of pieces of material may include at least one brazing aluminum alloy piece and at least one cast aluminum alloy piece, the chemical composition of a given specific aggregate is different from the chemical composition of the brazing aluminum alloy pieces, and the chemical composition of a given specific aggregate is different from the chemical composition of the cast aluminum alloy pieces. Classification may include processing image data captured from each of the plurality of pieces of material through a machine learning system.

[0008] Aspects of the present disclosure provide a system comprising: a sensor configured to capture one or more characteristics of each of a mixture of pieces of material, the mixture of pieces of material being capable of including pieces of material having different material classifications; a data processing system configured to classify each piece of material of the mixture of pieces of material as belonging to one of a plurality of different material classifications; and a sorting device configured to sort specific pieces of material from the mixture of pieces of material according to the classification of each piece of material of the mixture of pieces of material, the sorting resulting in an assembly of pieces of material having the chemical composition of a given specific aggregate. The sensor may be a camera, the one or more characteristics captured are captured by a camera configured to capture an image of each of the mixture of pieces of material as the mixture of pieces of material is conveyed past the camera, the camera being configured to capture a visual image of each of the mixture of materials and generate image data, and the characteristics are visually observable characteristics. The data processing system may include a machine learning system implementing a neural network configured to classify each piece of material of the mixture of pieces of material as belonging to one of a plurality of different material classifications based on the captured visually observable characteristics. The system may further include a device configured to determine an approximate mass of each of the plurality of pieces of material, and the sorting is performed according to the determined approximate mass and classification of each piece of material. The device may include a line scanner configured to measure an approximate size of each piece of material.

[0009] Aspects of the present disclosure provide a computer program product stored on a computer-readable storage medium that, when executed by a data processing system, performs a process including: determining an approximate mass of each of a plurality of pieces of material, where at least one of the plurality of pieces of material has a different material classification from other pieces of material; classifying each of the plurality of pieces of material as belonging to one of a plurality of different material classifications; and instructing to sort out specific ones of the plurality of pieces of material to generate an aggregate of pieces of material having a chemical composition of a predetermined specific aggregate, where the sorting is performed according to the determined approximate mass and classification of each of the plurality of pieces of material, and the aggregate of pieces of material includes pieces of material having different material classifications. The classification may include processing image data captured from each of the plurality of pieces of material through a machine learning system. The chemical composition of the predetermined specific aggregate may be different from the chemical composition of each of the plurality of pieces of material.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

DETAILED DESCRIPTION OF THE INVENTION

[0011] Various detailed embodiments of the present disclosure are disclosed herein. However, it should be understood that the disclosed embodiments are merely examples of the present disclosure and can be embodied in various alternative forms. The numerical values are not necessarily to scale, and in order to show the details of specific components, some functions may be exaggerated or minimized. Therefore, the specific structural and functional details disclosed herein should not be construed as limitations, but should only be construed as a representative basis for teaching those skilled in the art how to use various embodiments of the present disclosure.

[0012] As used herein, the term "chemical element" means a chemical element in the periodic table of chemical elements, including chemical elements that may be discovered after the filing date of this application. The "materials" used herein may include solids composed of compounds or mixtures of one or more chemical elements, and the complexity of the compounds or mixtures may range from simple to complex (all of which may be referred to herein as materials having a specific "chemical composition").

[0013] As used herein, "chemical composition of the aggregate" means the composition of chemical elements and their relative weight percentages (wt%) within a collection or group of individual separate pieces of material. (Note that weight percentage (or weight percent) is also called mass fraction and is the percentage of the mass of a particular chemical element within a material or substance relative to the total mass of the material or substance). For example, when an aggregate of individual parts of a metal alloy is melted together, the resulting "melt" will have a chemical composition equivalent to that of the aggregate. As referred to herein, a "melt" is when selected pieces of material are melted together and a composition analysis is performed on the melted-together pieces of material to determine the percentages (e.g., weight percentages) of the various chemical elements present in the melt.

[0014] The classes of materials include metals (ferrous and non-ferrous), metal alloys, plastics (including but not limited to PCBs, HDPE, UHMWPE, and plastics of various colors), rubbers, foams, glasses (including but not limited to borosilicate glass, soda-lime glass, and various colored glasses), ceramics, papers, cardboard, Teflon®, PE, bundled wires, insulated coated wires, rare earth elements, leaves, woods, plants, parts of plants, fibers, bio-waste, packaging, electronic waste, batteries, accumulators, scrap pieces of used vehicles, mines, construction, demolition waste, crop waste, forest residues, purpose-grown grasses, woody energy crops, microalgae, urban food waste, food waste, hazardous chemical waste and biomedical waste, construction waste materials, farm waste, articles of biological origin, articles of non-biological origin, objects having a specific carbon content, any other objects that may be found within municipal solid waste, and any other objects, articles, or materials disclosed herein, including but not limited to one or more sensor systems, and including but not limited to any further types or classes of the foregoing that are distinguishable from one another. Within this disclosure, the terms "scrap", "scrap piece", "material", "material piece", and "piece" may be used interchangeably. As used herein, a material piece or scrap piece referred to as having a metal alloy composition is a metal alloy having a specific chemical composition that is distinguishable from other metal alloys.

[0015] As is well known in the industry, a "polymer" is a substance or material composed of very large molecules or macromolecules made up of many repeating subunits. Polymers can be natural polymers found in nature or synthetic polymers.

[0016] A "multi-layer polymer film" is composed of two or more different compositions and may have a thickness of up to about 7.5 -8 ×10 -4 m. The layers are at least partially continuous and preferably, but in some cases, have the same spread.

[0017] As used herein, the terms "plastic", "plastic piece", and "plastic material piece" (which may all be used interchangeably) refer to any object that comprises or consists of a polymer composition of one or more polymers and / or multilayer polymer films.

[0018] As used herein, the term "chemical signature" refers to a unique pattern (such as a fingerprint spectrum) generated by one or more analytical instruments that indicates the presence of one or more specific elements or molecules (including polymers) in a sample. The elements or molecules may be organic and / or inorganic. Such analytical instruments include any of the sensor systems disclosed herein. According to embodiments of the present disclosure, one or more of the sensor systems disclosed herein may be configured to generate the chemical signature of a piece of material (e.g., a plastic piece).

[0019] As used herein, "fraction" refers to a specific combination of organic and / or inorganic elements or molecules, polymer types, plastic types, polymer compositions, chemical characteristics of plastics, physical properties of plastic pieces (such as color, transparency, strength, melting point, density, shape, size, manufacturing type, uniformity, reaction to stimuli, etc.), including all of the various classifications and types of plastics disclosed herein. Non-limiting examples of fractions include those with a relatively high proportion of aluminum added to LDPE, those with a relatively low proportion of iron added to LDPE and PP, those with zinc added to PP, combinations of PE, PET, and HDPE, any type of red LDPE plastic pieces, any combination of plastic pieces excluding PVC, black plastic pieces, combinations of plastics of types #3 - #7 including a specified combination of organic and inorganic molecules, combinations of one or more different types of multilayer polymer films, combinations of specific plastics that do not contain certain contaminants or additives, any type of plastic with a melting point exceeding a specified threshold, multiple specific types of thermosetting plastics, specific plastics that do not contain chlorine, combinations of plastics with similar densities, combinations of plastics with similar polarities, plastic bottles without caps attached, or vice versa, and are one or more different types of plastic pieces.

[0020] "Contact pyrolysis" includes the decomposition of a polymer material by heating the polymer material in the presence of a catalyst in the absence of oxygen.

[0021] The term "predetermined" refers to something that has been established or determined in advance.

[0022] "Spectral imaging" refers to imaging that uses multiple bands across the entire electromagnetic spectrum. A normal camera captures light across the three wavelength bands of red, green, and blue ("RGB") in the visible spectrum, but spectral imaging includes a variety of techniques that include RGB. In spectral imaging, infrared, visible, ultraviolet, and / or X-ray spectra, or combinations thereof, may be used. Spectral data, or spectral image data, is the digital data representation of a spectral image. Spectral imaging may include the simultaneous acquisition of spectral data in the visible and non-visible bands, illumination from outside the visible range, or the use of optical filters to capture specific spectral ranges. It is also possible to capture hundreds of wavelength bands for each pixel in a spectral image.

[0023] As used herein, the term "image data packet" refers to a packet of digital data relating to the captured spectral image of an individual piece of material.

[0024] As used herein, the terms "classify," "identify," "select," "recognize," and the terms "classification," "identification," "selection," "recognition," and their derivatives may be used interchangeably. As used herein, to "classify" a piece of material is to determine (i.e., identify) the type or class of material to which the piece of material belongs (or should belong at least according to the perceived properties of the piece of material). For example, according to certain embodiments of the present disclosure, a sensor system (further described herein) can be configured to collect and analyze any type of information for classifying materials and utilize the classification within a sorting system to selectively sort pieces of material according to a set of one or more perceived physical and / or chemical properties (e.g., user definable) including, but not limited to, color, texture, hue, shape, brightness, weight, density, composition, size, uniformity, manufacturing type, chemical characteristics, a predetermined fraction, radioactive characteristics, transmission of light, sound, or other signals, emission and / or reflection of electromagnetic radiation ("EM") by the piece of material, and responses to stimuli in various fields. As used herein, "manufacturing type" refers to the type of manufacturing process by which a piece of material, such as a metal part formed by brazing, casting (including, but not limited to, expendable mold casting, permanent mold casting, powder metallurgy), forging, material removal processes, etc., is manufactured.

[0025] The type or class of material (i.e., classification) can be user-defined and is not limited to known material classifications. The granularity of the type or class can range from very coarse to very fine. For example, the type or class can include relatively coarse types or classes of plastics, ceramics, glass, metals, and other materials, such as various metals and metal alloys like zinc, copper, brass, chromium plate, aluminum, etc., of a finer type or class, or specific subclasses of metal alloys of a relatively fine type or class. Thus, the type or class can be configured to distinguish materials of significantly different compositions, such as plastics or metal alloys, or to distinguish materials of substantially similar or nearly identical chemical compositions, such as different subclasses of metal alloys. It should be understood that the methods and systems discussed herein can be applied to identify / classify materials whose chemical composition is completely unknown prior to classification.

[0026] As used herein, the term "conveyor system" can be any known mechanical handling device that moves materials from one location to another, including but not limited to aircraft conveyors, automotive conveyors, belt conveyors, belt-driven live roller conveyors, bucket conveyors, chain conveyors, chain-driven live roller conveyors, drag conveyors, dust-proof conveyors, electric rail vehicle systems, flexible conveyors, gravity conveyors, gravity skate wheel conveyors, line shaft roller conveyors, electric-driven roller conveyors, overhead I-beam conveyors, overland conveyors, pharmaceutical conveyors, plastic belt conveyors, pneumatic conveyors, screw or auger conveyors, spiral conveyors, tubular gallery conveyors, vertical conveyors, vibrating conveyors, and wire mesh conveyors.

[0027] The systems and methods described herein according to certain embodiments of the present disclosure receive a mixture of a plurality of pieces of material, where at least one piece of material within the mixture has a different chemical composition (e.g., metal alloy composition, chemical characteristics) from one or more other pieces of material, and / or at least one piece of material within the mixture is manufactured in a different manner from one or more other materials, and / or at least one piece of material within the mixture is distinguishable from other pieces of material (e.g., visually distinguishable properties or characteristics, different chemical characteristics, etc.), and the systems and methods are configured to accordingly identify / classify / sort the pieces of material. Embodiments of the present disclosure can be utilized to classify any type or class, or fraction, of material as defined herein.

[0028] Note that the size and shape of the pieces of material to be sorted can be irregular (see, for example, FIGS. 6 - 8). For example, a material (e.g., Zorba and / or Twitch) may have been shredded into pieces of such irregular shapes and sizes (generating scrap pieces) and passed through some sort of shredder mechanism before being fed or deposited onto a conveyor system.

[0029] Embodiments of the present disclosure are described herein as sorting material pieces into such distinct groups or aggregates by physically disposing (e.g., redirecting or discharging) the material pieces into separate containers or multiple containers, or onto another conveyor system, according to user-defined groupings or aggregates (e.g., a predefined chemical composition of certain aggregates, a classification or fraction of a particular material type). As an example, in certain embodiments of the present disclosure, material pieces may be sorted into separate containers or multiple containers to separate material pieces composed of a particular chemical composition from other material pieces composed of different particular chemical compositions, and to produce a predefined chemical composition of certain aggregates within the sorted aggregate or group of material pieces. By way of non-limiting example, an aggregate of Twitch containing various aluminum alloys (e.g., various different wrought and / or cast aluminum alloys) may be sorted according to embodiments of the present disclosure to produce an aluminum alloy having a desired chemical composition (which may include an aluminum alloy having a unique chemical composition different from known aluminum alloys).

[0030] FIG. 1 shows an example of a system 100 configured according to various embodiments of the present disclosure. The conveyor system 103 is implemented to convey one or more (organized or random) flows of individual material pieces 101 through the system 100 so as to track, classify, and be able to classify each of the individual material pieces 101 into a predetermined desired group or aggregate (e.g., the chemical composition of one or more predetermined specific aggregates). Such a conveyor system 103 can be implemented using one or more conveyor belts on which the material pieces 101 typically move at a predetermined constant speed. However, certain embodiments of the present disclosure include systems in which the material pieces free fall through selected components of the system 100 (or any other type of vertical classifier), or other types of conveyor systems (such as those disclosed herein) including vibrating conveyor systems. Hereinafter, when applicable, the conveyor system 103 may also be referred to as the conveyor belt 103. In one or more embodiments, some or all of the operations of transmitting, tracking, stimulating, detecting, classifying, and sorting may be performed automatically, i.e., without human intervention. For example, in the system 100, one or more stimulation sources, one or more radiation detectors, classification modules, sorting devices, and / or other system components may be configured to perform these and other operations automatically.

[0031] Furthermore, while the simplified diagram of FIG. 1 shows a single flow of material pieces 101 on conveyor belt 103, embodiments of the present disclosure may be implemented such that multiple such flows of material pieces pass through the various components of system 100 in parallel with each other. For example, as further described in U.S. Patent No. 10,207,296, the material pieces may be distributed into two or more parallel individualized flows moving on a single conveyor belt or a set of parallel conveyor belts. According to certain embodiments of the present disclosure, the incorporation or use of singulators is not necessary. Instead, the conveyor system (e.g., conveyor system 103) may simply convey a large number of material pieces deposited (or deposited in large quantities on conveyor system 103 and then separated by a vibrating mechanism or the like) on conveyor system 103 in a random manner. Thus, certain embodiments of the present disclosure can track, sort, and / or separate such multiple conveyed material pieces simultaneously.

[0032] According to certain embodiments of the present disclosure, a suitable feeder mechanism (e.g., another conveyor system or hopper 102) can be utilized to supply the material pieces 101 onto the conveyor system 103, whereby the conveyor system 103 can convey the material pieces 101 through various components within the system 100. After the material pieces 101 are received by the conveyor system 103, an optional tumbler / vibrator / singulator 106 can be utilized to separate the individual material pieces from the combined mass of material pieces. In certain embodiments of the present disclosure, the conveyor system 103 operates to move at a predetermined speed by the conveyor system motor 104. This predetermined speed may be programmable and / or adjustable by an operator in any well-known manner. Alternatively, monitoring of the predetermined speed of the conveyor system 103 may be performed using a position detector 105. In certain embodiments of the present disclosure, control of the conveyor system motor 104 and / or the position detector 105 may be performed by an automatic control system 108. Such an automated control system 108 can be operated under the control of a computer system 107 and / or the functions for performing the automated control can be implemented in software within the computer system 107.

[0033] Thus, as further described herein, through the utilization of control for the conveyor belt drive motor 104 and / or the automatic control system 108 (optionally including the position detector 105), when each of the material pieces 101 moving on the conveyor belt 103 is identified, they can be tracked by position and time (with respect to the various components of the system 100), so that the various components of the system 100 can be activated / deactivated as each material piece 101 passes in its vicinity. As a result, the automatic control system 108 can track the respective positions of the material pieces 101 while the material pieces 101 are moving along the conveyor belt 103.

[0034] According to certain embodiments of the present disclosure, after the piece of material 101 is received by the conveyor belt 103, a tumbler and / or vibrator can be utilized to separate individual pieces of material from a mass of pieces of material (e.g., a physical pile, etc.). According to alternative embodiments of the present disclosure, the pieces of material can be arranged in one or more singulated (i.e., single file) streams that can be performed by an active or passive singulator 106. Examples of passive singulators are further described in U.S. Patent No. 10,207,296. As previously described, the incorporation or use of a singulator is not necessary. Instead, a conveyor system (e.g., conveyor belt 103) may simply convey a collection of pieces of material deposited on the conveyor belt 103 in a random manner.

[0035] Referring again to FIG. 1, certain embodiments of the present disclosure can utilize a visual or optical recognition system 110 and / or a material tracking measurement device 111 to track each of the pieces of material 101 moving on the conveyor belt 103. The vision system 110 can utilize one or more still cameras or action cameras 109 to record the respective positions (i.e., location and timing) of the pieces of material 101 on the moving conveyor belt 103.

[0036] The vision system 110 may be further or alternatively configured to perform identification (e.g., classification) of all or some specific types of the pieces of material 101, as further described herein. For example, such a vision system 110 may be utilized to capture or obtain information about each of the pieces of material 101. For example, the vision system 110 may be configured (e.g., using a machine learning system) to capture or collect any type of information from the pieces of material that can be used within the system 100 to classify and / or selectively sort the pieces of material 101 according to a set of one or more characteristics (e.g., physical and / or chemical and / or radioactive substances, etc.). According to a particular embodiment of the present disclosure, the vision system 110 may capture a visual image (including one-dimensional, two-dimensional, three-dimensional, or holographic images) of each of the pieces of material 101, for example, by using an optical sensor used in a general digital camera or video device. Such a visual image captured by the optical sensor is stored in a memory device as image data (e.g., initialized as an image data packet). According to a particular embodiment of the present disclosure, such image data may represent an image captured within the optical wavelength of light (i.e., the wavelength of light observable by a typical human eye). However, alternative embodiments of the present disclosure may utilize a sensor system configured to capture an image of a material composed of a wavelength of light outside the visual wavelength of the human eye. All such images may also be referred to herein as spectral images.

[0037] According to certain embodiments of the present disclosure, system 100 can be implemented using one or more sensor systems 120 that can be utilized alone or in combination with vision system 110 to classify / identify piece of material 101. Sensor system 120 can be composed of any type of sensor technology that utilizes irradiated or reflected electromagnetic radiation (e.g., infrared (“IR”), Fourier transform IR (“FTIR”), forward-looking IR (“FLIR”), very-near infrared (“VNIR”), near infrared (“NIR”), short-wavelength infrared (“SWIR”), long-wavelength infrared (“LWIR”), mid-wavelength infrared (“MWIR” or “MIR”), X-ray transmission (“XRT”), gamma rays, ultraviolet (“UV”), X-ray fluorescence (“XRF”), laser-induced breakdown spectroscopy (“LIBS”), Raman spectroscopy, anti-Stokes Raman spectroscopy, gamma spectroscopy, hyperspectral spectroscopy (e.g., beyond visible wavelengths), acoustic spectroscopy, NMR spectroscopy, microwave spectroscopy, terahertz spectroscopy, and one-dimensional, two-dimensional, three-dimensional, or holographic imaging using any of the foregoing), or by any other type of sensor technology including, but not limited to, chemical or radioactive substances. Implementation of an exemplary XRF system (e.g., used as sensor system 120 herein) is further described in U.S. Patent No. 10,207,296.

[0038] FIG. 1 is shown as a combination of vision systems 110 and one or more sensor systems 120, but it should be noted that embodiments of the present disclosure can be implemented with any combination of the sensor technologies disclosed herein, or other sensor technologies that are currently available or will be developed in the future. Although FIG. 1 is shown as including one or more sensor systems 120, the implementation of such sensor systems is optional within particular embodiments of the present disclosure. In particular embodiments of the present disclosure, a combination of both the vision system 110 and one or more sensor systems 120 can be used to classify the pieces of material 101. In particular embodiments of the present disclosure, any combination of one or more different sensor technologies disclosed herein can be used to classify the pieces of material 101 without using the vision system 110. Further, embodiments of the present disclosure may include any combination of one or more sensor systems and / or vision systems, and the outputs of such sensors and / or vision systems can be processed within a machine learning system (as further disclosed herein) to classify / identify materials from a mixture of materials and then separated from each other. If a sorting system (e.g., system 100) is configured to operate with only such a vision system 110, the sensor system 120 may be omitted from the system 100 (or simply turned off).

[0039] According to particular embodiments of the present disclosure, as further described herein with respect to FIG. 4, the vision system 110 and / or the sensor system may be configured to identify that a piece of material 101 is not of the type to be sorted by the system 100 for inclusion in an aggregate to produce a particular chemical composition of an aggregate (e.g., a piece of material containing a particular contaminant or chemical element), and to send a signal so that such a piece of material is not diverted together with other sorted pieces of material.

[0040] In certain embodiments of the present disclosure, the material tracking and measurement device 111 and the accompanying control system 112 can be utilized and configured to measure the respective size and / or shape of each material piece 101 passing within the vicinity of the material tracking and measurement device 111 that can be utilized by the system 100 to determine the approximate mass of each material piece, along with the position (i.e., position and timing) of each material piece 101 on the moving conveyor system 103. Alternatively, the vision system 110 may be utilized to track the respective position (i.e., position and timing) of each material piece 101 transported by the conveyor system 103.

[0041] Non-limiting, exemplary operations of such a material tracking and measuring device 111 and control system 112 are described herein with respect to FIG. 5. Such a material tracking and measuring device 111 can be implemented using a well-known laser light system that continuously measures the distance the laser light travels before being reflected by a detector of the laser light system. Thus, as each piece of material 101 passes near the device 111, the device 111 outputs a signal indicative of such a distance measurement to the control system 112. Thus, such a signal can represent a substantially discontinuous series of pulses, and the baseline of the signal is generated as a result of measuring the distance between the device 111 and the conveyor belt 103 at those instants when the piece of material is not near the device 111, and each pulse provides a measurement of the distance between the device 111 and the piece of material 101 passing over the conveyor belt 103. Because the shape of the piece of material 101 can be irregular, the height of such a pulse signal can also be irregular. Nevertheless, each pulse signal generated by the device 111 can provide the height of each portion as each piece of material 101 passes over the conveyor belt 103. The length of each such pulse also provides a measurement of the length of each piece of material 101 measured along a line substantially parallel to the direction of movement of the conveyor belt 103. What can be utilized within embodiments of the present disclosure to determine, or at least approximate, the mass of each piece of material 101 is this length measurement (corresponding to the time stamp of processing block 506 in FIG. 5) (and alternatively the height measurement), and the length measurement can be utilized to assist in sorting the pieces of material, as further described herein.

[0042] Next, referring to FIG. 5, a flowchart diagram of an exemplary system and process 500 for determining the approximate size, shape, and / or mass of each piece of material is shown. Such a system and process 500 can be implemented within any of the visual / optical recognition systems and / or material tracking and measurement devices described herein, such as the material tracking and measurement device 111 and the control system 112 shown in FIG. 1. In processing block 501, the material tracking and measurement device can be initialized with n = 0. Here, n represents the condition that the first piece of material being conveyed along the conveyor system has not yet been measured. As described above, such a material tracking and measurement device can establish a baseline signal representing the distance between the material tracking and measurement device and the conveyor belt when there is no object (i.e., piece of material) being conveyed thereon. In processing block 502, the material tracking and measurement device generates a continuous or substantially continuous measurement of distance. Processing block 503 represents a determination within the material tracking and measurement device as to whether the detected distance has changed from a predetermined threshold amount. It should be recalled that when the system 100 is started, at some point, the piece of material 101 moves along the conveyor system close enough to the material tracking and measurement device to be detected by the employed mechanism whose distance is being measured. In embodiments of the present disclosure, this can occur when the moving piece of material 101 passes within the line of the laser light utilized to measure the distance. When an object such as the piece of material 101 begins to be detected by the material tracking and measurement device (e.g., laser light), the distance measured by the material tracking and measurement device changes from its baseline value. The material tracking and measurement device can be pre - determined to detect only the presence of the piece of material 101 passing in its vicinity if the height of any part of the piece of material 101 is greater than a predetermined threshold distance value. FIG. 5 shows an example where such a threshold is 0.15 (e.g., representing 0.15 mm), but embodiments of the present disclosure should not be limited to any particular value.

[0043] System and process 500 continues to measure the current distance (i.e., repeat processing blocks 502 - 503) until this threshold distance value is reached. When a measured height exceeding the threshold is detected, the process proceeds to processing block 504, which records that a material piece 101 passing within the vicinity of the material tracking and measuring device has been detected on the conveyor system. Then, in processing block 505, the variable n is incremented to indicate to system 100 that another material piece 101 has been detected on the conveyor system. This variable n can be used to assist in tracking each of the material pieces 101. In processing block 506, the time stamp of the detected material piece 101 is recorded, and this time stamp can be used by system 100 to track the specific position and timing of the detected material piece 101 moving on the conveyor system, and at the same time, also represents the length of the detected material piece 101. In optional processing block 507, this recorded time stamp can be used to determine when to activate (start) and deactivate (stop) the acquisition of a measurement signal (e.g., X-ray fluorescence spectrum from the material piece 101) associated with the time stamp initiated by the sensor. The start time and stop time of the time stamp may correspond to the aforementioned pulse signal generated by the material tracking and measuring device. In processing block 508, this time stamp, together with the recorded height of the material piece 101, can be recorded in a table used by system 100 to track each of the material pieces 101 and the resulting classification.

[0044] Subsequently, in an optional processing block 509, next, the signal can be transmitted to a sensor system indicating a period for activating / deactivating the acquisition of a sensor start measurement signal from the piece of material 101, which may include a start time and a stop time corresponding to the length of the piece of material 101 determined by the material tracking and measurement device. Embodiments of the present disclosure can achieve such a task with a time stamp of the conveyor system received from the material tracking and measurement device indicating when the leading edge of the piece of material 101 passes through the irradiation source and then when the trailing edge of the piece of material 101 passes through the irradiation source thereafter, and a known predetermined speed.

[0045] Next, the system and process 500 for measuring each distance of the piece of material 101 moving along the conveyor system may be repeated each time the piece of material 101 passes through.

[0046] In certain embodiments of the present disclosure implementing one or more sensor systems 120, the one or more sensor systems 120 may be configured to assist the vision system 110 in identifying the respective chemical composition, relative chemical composition, and / or manufacturing type of the pieces of material 101 as the pieces of material 101 pass within the vicinity of the one or more sensor systems 120. The one or more sensor systems 120 may include an energy emission source 121 that can be powered by a power supply 122, for example, to stimulate a response from each of the pieces of material 101.

[0047] According to certain embodiments of the present disclosure implementing an XRF system as the sensor system 120, the line source 121 may include an in-line X-ray fluorescence ("IL-XRF") tube as further described in U.S. Patent No. 10,207,296. Such an IL-XRF tube may include a separate X-ray source dedicated to one or more streams (e.g., a streamlining) of the material pieces being conveyed. In such a case, the one or more detectors 124 may be implemented as XRF detectors that detect fluorescent X-rays from the pieces of material 101 within each of the individualized streams.

[0048] In certain embodiments of the present disclosure, when each piece of material 101 passes near the radiation source 121, the sensor system 120 can emit an appropriate sensing signal towards the piece of material 101. One or more detectors 124 can be arranged and configured to sense / detect one or more properties from the piece of material 101 in a format appropriate for the type of sensor technology utilized. The one or more detectors 124 and associated detector electronics 125 capture these received sensed properties, perform signal processing, generate digitized information (e.g., spectral data) representing the sensed properties, and then, analyzed according to certain embodiments of the present disclosure, can be used to classify each of the pieces of material 101 (either alone or in combination with the vision system 110). This classification can be performed within the computer system 107 and then, according to the determined classification, can be utilized by the automatic control system 108 to activate one of the N (N>1) sorting devices 126…129 of the sorting device for sorting (e.g., diverting / discharging) the pieces of material 101 into one or more of the N (N>1) sorting containers 136…139. Four sorting devices 126…129 and four sorting containers 136…139 associated with these sorting devices are shown in FIG. 1 as merely non-limiting examples.

[0049] The sorting device can include any well-known mechanism for redirecting a selected piece of material 101 towards a desired location, including but not limited to redirecting the piece of material 101 from a conveyor belt system into a plurality of sorting containers. For example, the sorting device can utilize air jets, and each air jet can be assigned to one or more classifications. When one of the air jets (e.g., 127) receives a signal from the automatic control system 108, that air jet emits an air stream that redirects / discharges the piece of material 101 from the conveyor system 103 into the sorting bin (e.g., 137) corresponding to that air jet.

[0050] The material piece can also be redirected / discharged using other mechanisms, such as removing the material piece from the conveyor belt by a robot, pushing the material piece from the conveyor belt (e.g., using a paintbrush type plunger), creating an opening (e.g., a trapdoor) in the conveyor system 103 through which the material piece can fall, or using an air jet to redirect the material piece to another container as it falls off the edge of the conveyor belt. The term pusher device, as used herein, refers to any form of device that can be actuated using pneumatic, mechanical, or other means, such as a suitable type of mechanical pushing mechanism (such as an ACME screw drive), a pneumatic pushing mechanism, or an air jet pushing mechanism, to dynamically move an object on or from a conveyor system / device. Some embodiments may include a plurality of pusher devices arranged at different positions along the path of the conveyor system and / or having different orientations of the redirection path. In various different implementations, these sorting systems described herein can determine which pusher device (if any) to activate depending on the classification of the material piece performed by a machine learning system. Further, the determination of which pusher device to activate may be based on the detected presence and / or characteristics of other objects that may be present within the redirection path of the pusher device at the same time as the target item (e.g., a material piece to be treated confidentially). Further, even in a facility where the singulation along the conveyor system is not complete, the disclosed sorting system can recognize that a plurality of objects are not properly separated and dynamically select the pusher device to be activated from a plurality of pusher devices based on a pusher device that provides an optimal redirection path that may separate the proximate objects. In some embodiments, the object identified as the target object may represent the material to be redirected from the conveyor system. In other embodiments, the object identified as the target object may represent the material that should be allowed to remain on the conveyor system so that the non-target object is redirected instead.

[0051] In addition to the N sorting containers 136...139 into which the material pieces 101 are diverted / discharged, the system 100 can also include a container 140 that receives the material pieces 101 that were not diverted / discharged from the conveyor system 103 to any of the aforementioned sorting containers 136...139. For example, the material pieces 101 may not be diverted / discharged to one of the N sorting containers 136...139 from the conveyor system 103 if the classification of the material pieces 101 has not been determined (or simply if the sorting device was unable to properly divert / discharge the material pieces), if the material pieces 101 contain contaminants detected by the vision system 110 and / or the sensor system 120, or if there is no need to produce a specific chemical composition of a particular aggregate. Alternatively, the container 140 may be used to receive one or more classifications of material pieces that have not been intentionally assigned to any of the N sorting containers 136...139. These material pieces may be further sorted according to other characteristics and / or by another sorting system.

[0052] According to the specific requirements of a given specific aggregate chemical composition, multiple classifications can be mapped to a single sorting device and associated containers. In other words, there does not need to be a one-to-one correlation between the classifications and the containers. For example, the user may desire to sort specific classifications of materials into the same container in order to achieve a specific aggregate chemical composition. To achieve this sorting, when the material pieces 101 are classified as meeting one or more requirements for achieving a specific aggregate chemical composition, the same sorting device can operate to sort them into the same container. Such combined sorting can be applied to produce any desired combination of sorted material pieces (e.g., one or more specific aggregate chemical compositions). The mapping of the classifications can be programmed by the user (e.g., using a sorting algorithm (see, e.g., FIG. 4) operated by the computer system 107) to produce such a desired combination. Further, the classification of the material pieces can be user-definable and is not limited to specific known classifications of the material pieces.

[0053] In certain embodiments of the present disclosure, conveyor system 103 may be divided into a plurality of belts configured in series, such as, for example, two belts, where the first belt conveys a piece of material passing through vision system 110 and / or implemented sensor system 120, and the second belt conveys a particular sorted piece of material passing through sensor system 120 implemented for subsequent sorting. Further, such a second conveyor belt may be at a lower height than the first conveyor belt such that the piece of material falls from the first belt onto the second belt.

[0054] In certain embodiments of the present disclosure implementing sensor system 120, light source 121 may be disposed above the detection region (i.e., above conveyor system 103). However, certain embodiments of the present disclosure may still dispose light source 121 and / or detector 124 at other positions that produce acceptable sensed / detected physical properties.

[0055] It should be understood that the systems and methods described herein are mainly described in relation to the classification of solid-state pieces of material, but the present disclosure is not limited thereto. The systems and methods described herein can be applied to classify materials having a range of physical states including, but not limited to, liquids, melts, gases, or powders in a solid state, other states, and any suitable combinations thereof.

[0056] Regardless of the type of sensed characteristics / captured information of the material pieces, next, the information can be transmitted to a computer system (e.g., computer system 107) to identify and / or classify each of the material pieces and processed by a machine learning system. Such a machine learning system can implement any well-known machine learning system, including those implementing neural networks (e.g., artificial neural networks, deep neural networks, convolutional neural networks, recurrent neural networks, autoencoders, reinforcement learning, etc.), supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, self-learning, feature learning, sparse dictionary learning, anomaly detection, robot learning, correlation rule learning, fuzzy logic, artificial intelligence "AI"), deep learning algorithms, deep structure learning hierarchical learning algorithms, support vector machines ("SVM") (linear SVM, non-linear SVM, SVM regression, etc.), decision tree learning (classification and regression trees ("CART"), etc.), ensemble methods (ensemble learning, random forest, bagging and pasting, patches and subspaces, boosting, stacking, etc.), dimensionality reduction (projection, manifold learning, principal component analysis, etc.), and / or deep machine learning algorithms (such as those described and published on the deeplearning.net website, which is hereby incorporated by reference in its entirety, including all software, publications, and hyperlinks to available software referenced within this website).Non-limiting examples of publicly available machine learning software and libraries that can be utilized within embodiments of the present disclosure include Python, OpenCV, Inception, Theano, Torch, PyTorch, Pylearn2, Numpy, Blocks, TensorFlow, MXNet, Caffe, Lasagne, Keras, Chainer, Matlab Deep Learning, CNTK, MatConvNet (a MATLAB® toolbox implementing convolutional neural networks for computer vision applications), DeepLearnToolbox (a Matlab toolbox for deep learning (by Rasmus Berg Palm)), BigDL, Cuda-Convnet (a fast C++ / CUDA implementation of convolutional (more generally, feedforward) neural networks), Deep Belief Networks, RNNLM, RNNLIB-RNNLIB, matrbm, deeplearning4j, Eblearn.lsh, deepmat, MShadow, Matplotlib, SciPy, CXXNET, Nengo-Nengo, Eblearn, cudamat, Gnumpy, three-way factorized RBM and mcRBM, mPoT (Python code for training models of natural images using CUDAMat and Gnumpy), ConvNet, Elektronn, OpenNN, NeuralDesigner, Theano generalized Hebbian learning, Apache Singa, Lightnet, and SimpleDNN.

[0057] According to certain embodiments of the present disclosure, machine learning can be performed in two stages. For example, first, training is performed, which can be performed offline in that the system 100 is not utilized to actually classify / sort the pieces of material. The system 100 can be utilized to train the machine learning system in that a homogeneous set of pieces of material (also referred to herein as a control sample), i.e., having the same type or class of material, or falling within the same predetermined ratio, passes through the system 100 (e.g., by the conveyor system 103), and all such pieces of material need not be sorted, but can be collected in a common container (e.g., container 140). Alternatively, the training can be performed at another location away from the system 100, including using some other mechanism for collecting sensing information (characteristics) of the control set of pieces of material. In this training stage, the algorithms within the machine learning system extract features from the captured information (e.g., using image processing techniques well known in the art). Non-limiting examples of training algorithms include, but are not limited to, linear regression, gradient descent, feedforward, polynomial regression, learning curves, regularized learning models, logistic regression. It is in this training stage that the algorithms within the machine learning system learn the relationship between the material and its features / characteristics (e.g., those captured by the vision system and / or sensor system) and create a knowledge base for later classifying the mixture of pieces of material received by the system 100. Such a knowledge base may include one or more libraries, and each library may include parameters (e.g., neural network parameters) utilized by the machine learning system when classifying the pieces of material. For example, one particular library may include parameters configured in the training stage for recognizing and classifying a particular type or class of material, or one or more materials corresponding to a predetermined ratio.According to certain embodiments of the present disclosure, when such a library can be input into a machine learning system, a user of system 100 can adjust certain of the parameters (e.g., adjust the effectiveness of a threshold as to how well the machine learning system recognizes a particular piece of material from a mixture of materials) to adjust the operation of system 100.

[0058] Furthermore, a particular material (e.g., a chemical element, a compound, etc.) being included in a piece of material (e.g., a metal alloy, etc.), or a combination of particular chemical elements or compounds can result in distinguishable physical characteristics (e.g., visually recognizable properties, etc.) in the material. As a result, when a plurality of pieces of material including such a particular composition pass through the aforementioned training stage, the machine learning system can learn a way to distinguish such pieces of material from other pieces of material. Thus, a machine learning system configured according to certain embodiments of the present disclosure can be configured to sort pieces of material according to their respective chemical compositions. For example, such a machine learning system can be configured to sort various aluminum alloys according to the percentage of a particular alloy material contained within the aluminum alloy.

[0059] For example, FIG. 6 shows a captured or acquired image of an exemplary piece of material of a cast aluminum alloy that can be used in the aforementioned training stage. FIG. 7 shows a captured or acquired image of an exemplary piece of material of an extruded aluminum alloy that can be used in the aforementioned training stage. FIG. 8 shows a captured or acquired image of an exemplary piece of material of a wrought aluminum alloy that can be used in the aforementioned training stage. In the training stage, a plurality of pieces of material (control samples) of a specific (homogeneous) classification (type) of material are passed (e.g., by a conveyor system) through a vision system and / or one or more sensor systems so that an algorithm within the machine learning system detects, extracts, and learns what features (e.g., visually recognizable characteristics) represent such a type or class of material. In other words, an image of a piece of cast aluminum alloy material such as that shown in FIG. 6 can pass through such a training stage, whereby the algorithm within the machine learning system “learns” (is trained) the method of detecting, recognizing, and classifying pieces of material composed of cast aluminum alloy. When training a vision system (e.g., vision system 110), it is trained to visually identify pieces of material. Thereby, a library of parameters specific to pieces of cast aluminum alloy material is created. Next, by performing a similar process on an image of an aluminum alloy extruded material such as that shown in FIG. 7, a library of parameters specific to aluminum alloy extruded materials can be created. Also, by performing a similar process on an image of a wrought aluminum alloy material such as that shown in FIG. 8, a library of parameters specific to wrought aluminum alloy materials can be created. As can be seen from the exemplary image of the cast aluminum alloy shown in FIG. 6, such a cast aluminum alloy material has visually recognizable features such as sharply defined angles. As can be seen from the exemplary image of the aluminum alloy extruded material shown in FIG. 7, such an aluminum alloy extruded material has visually recognizable features such as rounded corners and a hammer texture. As can be seen from the exemplary image of the wrought aluminum alloy shown in FIG. 8, such a wrought aluminum alloy material has visually recognizable features such as bends in the material and a smoother texture than that present in casting and extrusion.

[0060] Embodiments of the present disclosure are not limited to the materials shown in FIGS. 6-8. For each type of material classified by the vision system, any number of exemplary material pieces of that type of material can be passed through the vision system. When the captured sensing information is provided as input data, the algorithms within the machine learning system use N classifiers, each of which can test for one of N different material types, classes, or fractions. Note that the machine learning system can be "educated" (trained) to detect any type, class, or fraction of material, including the type, class, fraction, or other materials within the MSW whose chemical composition results in visually distinguishable features.

[0061] After the parameters within the algorithm are established and the machine learning system has sufficiently learned (trained) the differences in material classification (e.g., visually recognizable differences) (e.g., within a user-defined statistical confidence level), then a library of different material classifications is implemented in a material classification and / or sorting system (e.g., system 100) used to identify and / or classify material pieces from a mixture of material pieces, and then, when sorting is performed (e.g., to produce a specific chemical composition of aggregate), such classified material pieces are sorted.

[0062] Techniques for constructing, optimizing, and utilizing machine learning systems are known to those skilled in the art as described in the relevant literature. Examples of such literature include the following publications, namely, Krizhevsky et al., "ImageNet Classification with Deep Convolutional Networks," Proceedings of the 25th International Conference on Neural Information Processing Systems, December 3-6, 2012, Lake Tahoe, Nevada, and LeCun et al., "Gradient-Based Learning Applied to Document Recognition," Proceedings of the IEEE, Institute of Electrical and Electronics Engineers (IEEE), November 1998, both of which are hereby incorporated by reference in their entirety.

[0063] In an exemplary technique, data captured by a sensor and / or a vision system for a particular material can be processed as an array of data values within a data processing system (e.g., the data processing system 3400 of FIG. 11 implementing (configured to implement) a machine learning system). For example, the data may be spectral data captured by a digital camera or other type of sensor system for a particular piece of material and processed as an array of data values (e.g., an image data packet). Each data value can be represented as a single numerical value or as a series of numerical values representing the value. These values may be multiplied by the weight parameters of a neuron (e.g., using a neural network), and a bias may be added. This can affect the non-linearity of the neuron. The resulting numerical value output by the neuron can be treated in the same way as a value by multiplying it by the weight value of a subsequent neuron in this output, optionally adding a bias, and again affecting the non-linearity of the neuron. Each iteration of such processing is known as a "layer" of the neural network. The final output of the final layer can be interpreted as the probability of the presence or absence of the material within the captured data related to the piece of material. Examples of such processing are described in detail in both of the aforementioned references, "ImageNet Classification with Deep Convolutional Networks" and "Gradient-Based Learning Applied to Document Recognition".

[0064] According to certain embodiments of the present disclosure in which a neural network is implemented as a final layer (the "classification layer"), the final set of neuron outputs is trained to represent the likelihood that the data captured by the piece of material is associated. In operation, if the likelihood that the data captured by the piece of material is associated exceeds a user-specified threshold, it is determined that the data captured by the piece of material is actually associated. These techniques can be extended to determine not only the presence of a certain type of material associated with specific captured data, but also whether a sub-region of the specific captured data belongs to one type of material or another type of material. This process is known as segmentation, and techniques using neural networks exist in the literature, such as neural networks known as "fully convolutional" neural networks and networks that include convolutional portions even if they are not fully convolutional (i.e., are partially convolutional). This enables the determination of the position and size of the material.

[0065] It should be understood that the present disclosure is not limited only to machine learning techniques. Other common techniques for material classification / identification can also be used. For example, a sensor system can utilize spectroscopic analysis techniques using a multispectral camera or a hyperspectral camera to examine the spectral emission (i.e., spectral imaging) of a material to provide a signal indicating the presence or absence of a certain type, class, or fraction of material. A template matching algorithm may also use the spectral image of the piece of material, in which a database of spectral images is compared with the acquired spectral image, and the presence or absence of a specific type of material is detected from that database. It is also possible to compare the histogram of the captured spectral image with a database of histograms. Similarly, a bag-of-words model can be used together with a feature extraction technique such as scale-invariant feature transform ("SIFT") to compare the features extracted between the captured spectral image and those in the database.

[0066] Accordingly, as disclosed herein, certain embodiments of the present disclosure provide an identification / classification of one or more different types, classes, or fractions of materials to determine which pieces of material need to be diverted (i.e., sorted) from a conveyor system into defined groups (e.g., according to the chemical composition of one or more predetermined specific aggregates). According to certain embodiments, machine learning techniques are utilized to train (i.e., configure) a neural network to identify one or more different types, classes, or fractions of materials. Spectral images and other types of sensed information are captured from materials (e.g., those moving on a conveyor system), and based on the identification / classification of such materials, the systems described herein can determine which pieces of material to leave on the conveyor system and which pieces of material should be diverted / removed from the conveyor system (e.g., either placed in a collection container or diverted to another conveyor system).

[0067] According to certain embodiments of the present disclosure, a machine learning system for an existing facility (e.g., system 100) can be dynamically reconfigured to identify / classify the characteristics of a new type, class, or fraction of materials by replacing a current set of neural network parameters with a new set of neural network parameters.

[0068] It should be noted here that according to certain embodiments of the present disclosure, the detected / captured features / characteristics of the material piece (e.g., spectral image) do not necessarily have to be merely physically identifiable or discernible characteristics. They may be abstract formulas that can only be expressed mathematically in some cases, or may not be expressible mathematically at all. Nevertheless, the machine learning system can be configured to analyze the spectral data to find patterns that can classify the control samples during the training phase. Further, the machine learning system can obtain a subsection of the captured information (e.g., spectral image, etc.) of the material piece and attempt to find the correlation between the pre-defined classifications.

[0069] According to certain embodiments of the present disclosure, instead of using the training phase in which the control samples of the material piece are passed by the vision system and / or the sensor system, the training of the machine learning system can be performed using labeling / annotation techniques, whereby the data / information of the material piece is captured by the vision / sensor system, and the user inputs the labels or annotations that identify each material piece, which are used to create the library that the machine learning system uses when classifying the material pieces within the mixture of the material pieces.

[0070] According to certain embodiments of the present disclosure, any perceived characteristic output by any of the sensor systems 120 disclosed herein can be input into the machine learning system for classifying and / or sorting the material. For example, in a machine learning system implementing supervised learning, the output of the sensor system 120 that uniquely characterizes a particular type or composition of material (e.g., a particular metal alloy) can be used to train the machine learning system.

[0071] FIG. 9 shows a flowchart diagram illustrating an exemplary embodiment of a process 3500 for classifying / sorting pieces of material using a vision system 110 and / or one or more sensor systems 120, according to a particular embodiment of the present disclosure. Process 3500 may be performed to classify a mixture of pieces of material into any combination of predetermined types, classes, and / or fractions, including generating a chemical composition of a predetermined particular aggregate. Process 3500 may be configured to operate within any embodiment of the present disclosure described herein, including system 100 of FIG. 1. As further described, process 3500 may be utilized within the system and process 400 of FIG. 4. The operation of process 3500 may be performed by hardware and / or software included within a computer system (e.g., computer system 3400 of FIG. 11) that controls the system (e.g., computer system 107, vision system 110, and / or sensor system 120 of FIG. 1).

[0072] In processing block 3501, the material piece 101 can be deposited on the conveyor system 103. In processing block 3502, as each material piece 101 moves through the system 100, the position of each material piece 101 on the conveyor system 103 is detected in order to track each material piece 101. This can be performed by the vision system 110 (e.g., by distinguishing the material piece 101 from the underlying conveyor system material while communicating with a conveyor system position detector (e.g., position detector 105)). Alternatively, a material tracking device 111 can also be used to track the material piece 101. Or, any system that can generate a light source (including but not limited to visible light, UV, and IR) and is equipped with a corresponding detector can be used to track the material piece 101. In processing block 3503, when the material piece 101 moves near one or more of the vision system 110 and / or the sensor system 120, the sensed information / characteristics of the material piece 101 are captured / acquired. In processing block 3504, a vision system as described above (e.g., implemented within the computer system 107) may perform preprocessing of the captured information and can detect (extract) the respective information of the material piece 101 using the captured information (e.g., from the background (e.g., the belt conveyor 103); in other words, preprocessing can be utilized to identify the difference between the material piece 101 and the background). Well-known image processing techniques such as dilation, thresholding, and contouring can be used to identify the material piece 101 as distinguishable from the background. In processing block 3505, segmentation can be performed. For example, the captured information may include information regarding one or more material pieces 101. Further, when an image is captured, a particular material piece 101 may be located at the seam of the conveyor belt 103. Thus, in such cases, it may be desirable to separate the image of each individual material piece 101 from the background of the image. In an exemplary technique of processing block 3505, the first step is to apply high contrast to the image, such that in this way, substantially all background pixels are reduced to black pixels, and at least some of the pixels regarding the material piece 101 are brightened to substantially all white pixels.The white image pixels of the material piece 101 are expanded to cover the size of the entire material piece 101. When this step is completed, the position of the material piece 101 becomes a high-contrast image with all white pixels arranged on a black background. Next, a contour algorithm can be utilized to detect the boundary of the material piece 101. The boundary information is saved and the boundary position is transferred to the original image. Next, segmentation is performed in a region of the original image that is larger than the previously defined boundary. In this way, the material piece 101 is identified and separated from the background.

[0073] In an optional processing block 3506, the material piece 101 may be conveyed along the conveyor system 103 near the material tracking and measuring device 111 and / or the sensor system 120 to determine the size and / or shape of the material piece 101. Such a material tracking and measuring device 111 may be configured to measure one or more dimensions of each material piece so that the system can calculate (determine) the approximate mass of each material piece. In processing block 3507, post-processing may be performed. The post-processing may include resizing the captured information / data in preparation for use in a machine learning system. The post-processing may also include changing specific characteristics (e.g., enhancing the contrast of the image, changing the background of the image, applying a filter, etc.) in a way that enhances the function of the machine learning system for classifying the material piece 101. In processing block 3509, the size of the data may be changed. Under certain circumstances, it may be desirable to change the size of the data to meet the data input requirements of a specific machine learning system such as a neural network. For example, in a neural network, an image data size much smaller than the size of an image captured by a general digital camera (e.g., 225×255 pixels or 299×299 pixels) may be required. Furthermore, the smaller the size of the input data, the shorter the processing time required to perform the classification. Therefore, the smaller the data size, the higher the throughput of the system 100 and the greater its value can be.

[0074] In processing blocks 3510 and 3511, each material piece 101 is identified / classified based on the sensed / detected features. For example, processing block 3510 may be composed of a neural network that uses one or more machine learning algorithms to compare the extracted features with the features stored in a previously generated knowledge base (e.g., generated during the training phase), and based on such comparison, assigns the most matching classification to each of the material pieces 101. The algorithms of the machine learning system may hierarchically process the captured information / data using automatically trained filters. The filter responses are properly combined at the next level of the algorithm until probabilities are obtained at the final step. In processing block 3511, these probabilities can be used for each of the N classifications to determine into which of the N sorting containers each material piece 101 should be sorted. Each of the N classifications may be related to the chemical composition of N different predetermined specific aggregates. For example, each of the N classifications can be assigned to one sorting container, and the material piece 101 under consideration is sorted into that container corresponding to the classification that returns a probability greater than a predefined threshold. In embodiments of the present disclosure, such a predefined threshold may be preset by the user. If none of the probabilities are greater than the predefined threshold, the specific material piece 101 may be classified into an outlier container (e.g., sorting container 140).

[0075] Next, in processing block 3512, sorting devices 126...129 are activated corresponding to the classification of the material piece 101. Between the time when the image of the material piece 101 is captured and the time when the sorting devices 126...129 are activated, the material piece 101 has moved to a position downstream of the conveyor system 103 near the vision system 110 and / or the sensor system 120 (e.g., at the conveying speed of the conveyor system). In an embodiment of the present disclosure, the activation of the sorting devices 126...129 is timed such that the sorting devices 126...129 are activated when the material piece 101 passes through the sorting devices 126...129 mapped to the classification of the material piece 101, and the material piece 101 is redirected / discharged from the conveyor system 103 to the associated sorting containers 136...139. In an embodiment of the present disclosure, the activation of the sorting devices 126...129 can be timed by respective position detectors that detect when the material piece 101 passes in front of the sorting devices 126...129 and send a signal enabling the activation of the sorting devices 126...129. In processing block 3513, the sorting containers 136...139 corresponding to the activated sorting devices 126...129 receive the redirected / discharged material piece 101.

[0076] FIG. 10 shows a flowchart diagram illustrating an exemplary embodiment of a process 1000 for classifying / sorting a material piece 101 according to a particular embodiment of the present disclosure. The process 1000 can be configured to operate within any embodiment of the present disclosure described herein, including the system 100 of FIG. 1. As will be further described, the process 1000 can be utilized within the system and process 400 of FIG. 4.

[0077] Process 1000 may be configured to operate in cooperation with process 3500. For example, according to certain embodiments of the present disclosure, process blocks 1003 and 1004 may be incorporated into process 3500 (e.g., operating in series or in parallel with process blocks 3503-3510) to combine the efforts of vision system 110 implemented in combination with a machine learning system having a sensor system (e.g., sensor system 120) that is not implemented in combination with the machine learning system for classifying and / or sorting material piece 101, including those according to the system and method 400 of FIG. 4.

[0078] The operations of process 1000 may be performed by hardware and / or software included within a computer system (e.g., computer system 3400 of FIG. 11) that controls various aspects of system 100 (e.g., computer system 107 of FIG. 1). In process block 1001, material piece 101 may be deposited on conveyor system 103. Next, in optional process block 1002, material piece 101 may be conveyed along conveyor system 103 within the vicinity of material tracking measurement device 111 and / or an optical imaging system to track each material piece and / or determine the size and / or shape of material piece 101. Such material tracking and measurement device 111 may be configured to measure one or more dimensions of each material piece such that the system can calculate (determine) the approximate mass of each material piece. In process block 1003, as material piece 101 moves near sensor system 120, material piece 101 may be interrogated or stimulated with EM energy (waves) or other types of stimuli suitable for the particular type of sensor technology utilized by sensor system 120. In process block 1004, the physical properties of material piece 101 are sensed / detected and captured by sensor system 120. In process block 1005, for at least a portion of material piece 101, the type of material is identified / classified (at least in part) based on the captured properties and can be combined with classification by a machine learning system in cooperation with vision system 110 (e.g., when executed in combination with process 3500).

[0079] Next, when sorting of the material pieces 101 is performed, in processing block 1006, sorting devices 126...129 corresponding to the classification of the material pieces 101 are activated. Between when the material piece is sensed and when the sorting devices 126...129 are activated, the material piece 101 moves from near the sensor system 120 to a position downstream of the conveyor system 103 at the conveyance speed of the conveyor system. In a particular embodiment of the present disclosure, the activation of the sorting devices 126...129 is timed such that when the material piece 101 passes through the sorting devices 126...129 mapped to the classification of the material piece 101, the sorting devices 126...129 are activated and the material piece 101 is diverted / discharged from the conveyor system 103 to its associated sorting containers 136...139. In a particular embodiment of the present disclosure, the activation of the sorting devices 126...129 can be timed by respective position detectors that detect when the material piece 101 passes in front of the sorting devices 126...129 and transmit a signal enabling the activation of the sorting devices 126...129. In processing block 1007, the sorting containers 136...139 corresponding to the activated sorting devices 126...129 receive the diverted / discharged material pieces.

[0080] According to various embodiments of the present disclosure, different types or classes of materials may be classified by different types of sensors for use in a machine learning system, respectively, and combined to classify material pieces in a scrap or waste stream.

[0081] According to various embodiments of the present disclosure, data (e.g., spectral data) from two or more sensors can be combined using one or more machine learning systems to perform classification of the material pieces.

[0082] According to various embodiments of the present disclosure, a plurality of sensor systems can be attached to a single conveyor system, and each sensor system can utilize a different machine learning system. According to various embodiments of the present disclosure, a plurality of sensor systems can be attached to different conveyor systems, and each sensor system can utilize a different machine learning system.

[0083] According to embodiments of the present disclosure, system 100 can be configured to output a collection of sorted materials having a particular chemical composition (i.e., the chemical composition of a given specific aggregate) in an aggregate (e.g., according to system and method 400 of FIG. 4). In other words, if such a collection of sorted materials were combined into a single object or mass, or at least theoretically could be combined (e.g., melted together or mixed in a solution), such a unique object or mass would have a particular chemical composition. Further, embodiments of the present disclosure can be configured to output a collection of materials having a particular chemical composition that does not exist within any individual piece of material supplied to system 100.

[0084] Non-limiting examples would be the production of aluminum alloys having a chemical composition according to a predetermined (e.g., designed by the user of system 100) combination of specific weight percentages (wt%) of aluminum, silicon, magnesium, iron, manganese, copper, zinc. Scrap pieces of aluminum alloys available for supply to system 100 may be those listed in the table of FIG. 2. And from the sorting of such available aluminum alloy scrap pieces, it may be desirable to produce an aluminum alloy having a chemical composition substantially equivalent to that listed in the table of FIG. 3. However, even if system 100 can be configured to distinguish each of the aluminum alloys listed in the table of FIG. 2 (i.e., by classifying each of the aluminum alloy pieces 101 according to either or both of processes 1000 and 3500), none of these aluminum alloys have a chemical composition equivalent to that listed in the table of FIG. 3. Thus, even if scrap pieces made of any of the aluminum alloys listed in the table of FIG. 2 are sorted, an aggregate of aluminum alloy scrap pieces having a chemical composition equivalent to that listed in the table of FIG. 3 as a whole cannot be obtained.

[0085] However, embodiments of the present disclosure can be configured to generate a set of aluminum alloy scrap pieces having a chemical composition equivalent to, or at least substantially equivalent to, the chemical composition of the aggregate listed in the table of FIG. 3. This is achieved by utilizing one or more of the vision system 110 and / or the sensor system 120 to classify, select, and sort combinations of multiple scrap pieces of the aluminum alloy of FIG. 2 in a ratio at which the chemical composition of the aggregate (also referred to herein as the chemical composition of a predetermined specific aggregate) is obtained for output.

[0086] Since individual aluminum alloy scrap pieces can have different sizes and thus different masses, the mass of each aluminum alloy scrap piece can be estimated using the material tracking and measurement device 111. For example, the system 100 can use the size of each scrap piece measured by the material tracking and measurement device 111 to determine (calculate) the mass of each scrap piece, or at least an approximate mass. The system 100 is configured to recognize and classify each scrap piece as belonging to one of the plurality of aluminum alloys listed in the table of FIG. 2. Since the specific chemical composition of each different aluminum alloy is known, the system 100 can use this information together with the determined size of each scrap piece to determine (calculate) the mass of each different chemical element contained in each aluminum alloy scrap piece, or at least an approximate mass.

[0087] To generate an aggregate of aluminum alloy scrap pieces having the chemical composition of the aggregate, the system 100 is then configured to sort and select the aluminum alloy scrap pieces supplied to the system 100 so that when combined, the combined mass of the sorted aluminum alloy scrap pieces achieves the chemical composition of the aggregate. In other words, if the combined mass of the aluminum alloy scrap pieces sorted and output by the system 100 dissolves in each other (is likely to dissolve at some point), the resulting melt will have the chemical composition of the aggregate, or at least be substantially close to the chemical composition of the aggregate within a desired accuracy threshold.

[0088] Thus, the system 100 may be configured to calculate on a running basis the contribution of each chemical element in the chemical composition of the aggregate to the individual mass as each aluminum alloy scrap piece is added to the sorted aggregate, whereby the system 100 can determine whether the next classified aluminum alloy scrap piece should be added to the aggregate (i.e., separated from the mixture of aluminum alloy scrap pieces).

[0089] FIG. 4 shows a flowchart block diagram of a system and process 400 configured in accordance with an embodiment of the present disclosure to produce an aggregate of pieces of material having a chemical composition of a given specific aggregate. The system and process 400 may be implemented as a computer program (or other type of algorithm) executed within system 100 (e.g., by computer system 107). The system and process 400 may be executed in conjunction with aspects of the system and process 3500 of FIG. 9, and / or the system and process 1000 of FIG. 10.

[0090] In processing block 401, system 100 receives or inputs a predetermined specific chemical composition of a given aggregate that is desirably generated at the output of one of the sorting devices 126...129 within system 100. In processing block 402, when each piece of material 101 is conveyed through the material tracking and measurement device 111, the material tracking and measurement device 111 will determine the respective size and / or shape of each piece of material 101 as described herein. In processing block 403, each piece of material 101 is assigned a classification by the vision system 110 and / or one or more sensor systems 120 in the manner described herein (see, for example, FIGS. 9 and 10). In processing block 404, system 100 will determine the respective chemical composition of each classified piece of material 101. This can be determined directly using one or more sensor systems 120, such as an XRF or LIBS system, that can measure and determine the weight percentages of various chemical elements within a particular piece of material. Alternatively, the respective chemical composition of each classified piece of material 101 may be determined indirectly, such as by being inferred as a result of the classification of the piece of material 101. For example, if the various different classes or types of pieces of material 101 supplied to system 100 are known (as described above with respect to FIG. 2), then the specific chemical composition of each class or type of piece of material 101 can be input into system 100 (e.g., stored in a database), and when a particular piece of material 101 is classified (e.g., by the vision system 110 and / or one or more sensor systems 120), its specific chemical composition is associated (in some way) with the determined classification. Further, in processing block 404, the respective mass of each piece of material 101 can be estimated based on the previously determined size and / or shape, and as a result, the approximate mass of each chemical element within the piece of material can be determined. This can be achieved because the relative masses of the chemical elements of various known types or classes of pieces of material are known and can be pre-input into system 100 in a similar manner as the known chemical compositions.

[0091] In processing block 405, system 100 sorts each of the material pieces 101 based on the determined chemical composition and mass to achieve a predetermined specific chemical composition of the aggregate. For example, system 100 may be configured to sort each of these material pieces 101 into a predetermined container (e.g., container 136) by a predetermined sorting device (e.g., sorting device 126) (e.g., redirect). The remainder of the material pieces 101 may be collected in container 140, or system 100 may be configured to sort certain ones of the material pieces 101 into another container (e.g., container 137) to achieve a second (e.g., different) predetermined specific chemical composition of the aggregate. Alternatively, system 100 may be configured to sort the remaining material pieces 101 based on any other type of desired classification, such as sorting the remaining material pieces 101 into two different classifications (e.g., low-temperature aluminum, extruded aluminum, and / or cast aluminum). In processing block 406, the sorted material pieces 101 for achieving the specific chemical composition of the aggregate are collected in a predetermined container (e.g., container 136).

[0092] Processing blocks 402-406 can be repeated as needed to achieve the chemical composition of a particular aggregate, to achieve the chemical composition of a particular aggregate within a particular accuracy threshold, or to achieve the chemical composition of a particular aggregate of a desired (predetermined) collected mass of material (which can be determined by counting the number of materials diverted towards the container). For example, since each material is sorted, the system can continuously determine (i.e., update) the chemical composition of the aggregate of the collected pieces of material and continue sorting until the updated chemical composition of the aggregate is within the threshold level of a predetermined particular aggregate chemical composition. Since each piece of material is classified, the system determines, for example, whether to divert that piece of material for collection, such as whether that piece of material will increase or decrease the total weight percentage of a particular chemical element within the already sorted and collected pieces of material. Further, the system may be configured not to divert a particular piece of material to the aggregate since such a piece of material contains contaminants that are not desired to be included within a predetermined particular chemical composition (e.g., a low-total-aluminum alloy piece containing an iron-containing material such as a bolt). Alternatively, another system can be implemented to remove pieces of material containing particular contaminants.

[0093] The material tracking and measuring device 111 may be a well-known one-dimensional or two-dimensional line scanner. In the case of a one-dimensional line scanner, the length along the travel direction of each piece of material is measured. If it can be assumed that the length and width of most pieces of material are approximately equal, such length measurements can be used to estimate the mass of each piece of material. When using a two-dimensional line scanner, both the length and width of each piece of material can be measured to determine the mass.

[0094] Alternatively, one or more cameras can be utilized in a well-known manner to image each piece of material and determine the approximate dimensions of each piece of material. Such cameras can be placed near the conveyor belt in front of the sorting device or downstream of the sorting device to image only the sorted pieces of material and determine the approximate mass.

[0095] If a substantial majority of the pieces of material can all be assumed to be approximately the same size and mass, such an implementation for determining the mass of each piece can be omitted.

[0096] Alternatively, a container for collecting the redirected pieces of material can be placed on a weighing scale that continuously weighs the collected pieces of material, so that when the pieces of material are sorted and collected within the container, the approximate weight and thus the mass of each piece of material are obtained. These masses can be utilized in the systems and processes 400 described herein.

[0097] According to certain embodiments of the present disclosure, at least a plurality of the system 100 can be continuously coordinated together to perform multiple iterations or layers of sorting. For example, when two or more systems 100 are coordinated in this way, a conveyor system passes the pieces of material through a first vision system (and, according to certain embodiments, a sensor system) configured to sort a first set of pieces of material of a mixture of materials by a sorter (e.g., the first automated control system 108 and associated one or more sorting devices 126... 129) into a first set of one or more containers (e.g., sorting containers 136... 139), and then passes the pieces of material through a second vision system (and, according to certain embodiments, another sensor system) configured to sort a second set of pieces of material of a mixture of materials by a second sorter into a second set of one or more sorting containers, which can be implemented with a single conveyor belt or multiple conveyor belts. Further discussion of such multi-stage sorting is described in U.S. Patent Application Publication No. 2022 / 0016675, which is incorporated herein by reference.

[0098] Such a series of systems 100 can include any number of such systems coordinated with each other in such a way. According to certain embodiments of the present disclosure, each successive vision system or sensor system can be configured to sort a different material than the previous vision system or sensor system, and ultimately, a collection of pieces of material having a predetermined specific chemical composition of the aggregate is produced.

[0099] Referring now to FIG. 11, a block diagram is shown illustrating a data processing (“computer”) system 3400 in which aspects of embodiments of the present disclosure may be implemented. (The terms “computer,” “system,” “computer system,” and “data processing system” may be used interchangeably herein.) Aspects of computer system 107, automation control system 108, sensor system 120, and / or vision system 110 may be configured similarly to computer system 3400. Computer system 3400 may use a local bus 3405. In particular, any suitable bus architecture may be utilized, such as a Peripheral Component Interconnect (“PCI”) local bus architecture, an Accelerated Graphics Port (“AGP”) architecture, or an Industry Standard Architecture (“ISA”). One or more processors 3415, volatile memory 3420, and non-volatile memory 3435 may be connected to local bus 3405 (e.g., via a PCI bridge (not shown)). An integrated memory controller and cache memory may be coupled to one or more processors 3415. One or more processors 3415 may include one or more central processing units and / or one or more graphics processing units 3401 and / or one or more tensor processing units. Additional connections to local bus 3405 may be made directly between components or through add-in boards. In the example shown, a communication (e.g., network (LAN)) adapter 3425, an I / O (e.g., Small Computer System Interface (“SCSI”) host bus) adapter 3430, and an expansion bus interface (not shown) may be connected to local bus 3405 by direct component connection. An audio adapter (not shown), a graphics adapter (not shown), and a display adapter 3416 (coupled to display 3440) may be connected to local bus 3405 (e.g., by an add-in board inserted into an expansion slot).

[0100] The user interface adapter 3412 can provide connections for the keyboard 3413, the mouse 3414, a modem / router (not shown), and additional memory (not shown). The I / O adapter 3430 can provide connections for the hard disk drive 3431, the solid state drive 3432, and a CD-ROM drive (not shown).

[0101] The operating system is executed on one or more processors 3415 and can be used to coordinate and control various components within the computer system 3400. In FIG. 11, the operating system may be a commercially available operating system. An object-oriented programming system (e.g., Java, Python, etc.) can be executed in conjunction with the operating system to provide calls from programs (e.g., Java, Python, etc.) running on the system 3400 to the operating system. Instructions for the operating system, the object-oriented operating system, and the programs may be located on a non-volatile memory 3435 storage device such as the hard disk drive 3431 or the solid state drive 3432, or may be loaded into the volatile memory 3420 for execution by the processor 3415.

[0102] Those skilled in the art will understand that the hardware in FIG. 11 may vary depending on the implementation. In addition to or instead of the hardware shown in FIG. 11, other internal hardware or peripheral devices such as flash ROM (or equivalent non-volatile memory) or an optical disk drive can also be used. Also, any of the processes of the present disclosure may be applied to a multiprocessor computer system or may be executed by a plurality of such systems 3400. For example, training of a machine learning system can be executed by a first computer system 3400 while the operation of the sorting system 100 can be executed by a second computer system 3400.

[0103] As another example, computer system 3400 may be a stand-alone system configured to be bootable without relying on a particular type of network communication interface, whether or not computer system 3400 includes some type of network communication interface. As a further example, computer system 3400 may be an embedded controller composed of ROM and / or flash ROM that provides non-volatile memory for storing operating system files or user-generated data.

[0104] The examples shown in FIG. 11 and the above examples do not imply architectural limitations. Further, the computer program form of the aspects of the present disclosure can reside on any computer-readable storage medium used by a computer system (i.e., floppy disk, compact disk, hard disk, tape, ROM, RAM, etc.).

[0105] As described herein, embodiments of the present disclosure may be implemented to perform various functions described for identifying, tracking, classifying, and / or sorting pieces of material. Such functions can be implemented in hardware and / or software, such as within one or more data processing systems (e.g., data processing system 3400 of FIG. 11), aspects of the aforementioned computer system 107, vision system 110, sensor system 120, and / or automatic control system 108. However, the functions described herein are not limited to implementation on a particular hardware / software platform.

[0106] As will be appreciated by one skilled in the art, aspects of the present disclosure may be embodied as a system, process, method, and / or computer program product. Accordingly, various aspects of the present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment (including firmware, resident software, microcode, etc.), or an embodiment combining software and hardware aspects that are generally referred to herein as a "circuit", "circuitry", "module", or "system". Further, aspects of the present disclosure may take the form of a computer program product embodied in one or more computer-readable storage media having computer-readable program code incorporated therein. (However, any combination of one or more computer-readable media may also be utilized. The computer-readable media may be a computer-readable signal medium or a computer-readable storage medium.)

[0107] A computer-readable storage medium may be, for example, an electronic, magnetic, optical, electromagnetic, infrared, biological, atomic, or semiconductor system, apparatus, controller, or device, or any suitable combination thereof, but is not limited thereto, and the computer-readable storage medium itself is not a transient signal. More specific examples (a non-exhaustive list) of computer-readable storage media include electrical connections having one or more wires, floppy disks of portable computers, hard disks, solid-state memories, random access memories ("RAM") (e.g., RAM 3420 in FIG. 11), read-only memories ("ROM") (e.g., ROM 3435 in FIG. 11), erasable programmable read-only memories ("EPROM") or flash memories, optical fibers, portable compact disk read-only memories ("CD-ROM"), optical storage devices, magnetic storage devices (e.g., hard drive 3431 in FIG. 11), or any suitable combination of the foregoing. In the context of this document, a computer-readable storage medium may be any tangible medium that can include or store a program used by or related to an instruction execution system, apparatus, controller, or device. Program code embodied on a computer-readable signal medium may be transmitted using any suitable medium including, but not limited to, wireless, wired, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0108] A computer-readable signal medium may include a propagated data signal having computer-readable program code incorporated therein, for example, as part of a baseband or a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electromagnetic, optical, or any suitable combination thereof. A computer-readable signal medium may be any computer-readable medium that can communicate, propagate, or transport a program for use by or related to an instruction execution system, apparatus, controller, or device, rather than a computer-readable storage medium.

[0109] The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, processes, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, segment, or portion of code that includes one or more executable program instructions for implementing the specified logical function. It should also be noted that, depending on the implementation, the functions shown in the blocks may be executed in an order different from that shown in the figures. For example, two blocks shown consecutively may actually be executed substantially simultaneously, or depending on the relevant functions, the blocks may be executed in the reverse order.

[0110] In the description herein, a flowcharted technique may be described as a series of consecutive operations. The order of the operations and the parties performing the operations can be freely changed without departing from the scope of the teachings. The operations can be added, deleted, or changed in several ways. Similarly, the order of the operations can be re-ordered or looped. Further, processes, methods, algorithms, etc. may be described in a consecutive order, but such processes, methods, algorithms, or any combination thereof may be operable to be executed in a different order. Additionally, some operations within a process, method, or algorithm may be executed at least sometimes simultaneously (e.g., operations executed in parallel), and may be executed in whole, in part, or in any combination thereof.

[0111] Modules implemented in software and executed by various types of processors (e.g., GPU 3401, CPU 3415) may include one or more physical or logical blocks of computer instructions that can be organized, for example, as objects, procedures, or functions. However, the executable files of the identified modules need not be physically located together, and may include heterogeneous instructions stored in different locations, which, when logically combined, incorporate the module and achieve the specified purpose of the module. In fact, the modules of executable code may be a single instruction, or may be multiple instructions, and may be distributed across multiple different code segments, between different programs, and across multiple memory devices. Similarly, operational data (e.g., the material classification library described herein) may be identified and illustrated within a module herein, embodied in any suitable form, and organized within any suitable type of data structure. The operational data may be collected as a single data set or may be distributed across different locations including different storage devices. The data may provide an electronic signal on a system or network.

[0112] These program instructions are provided to one or more processors and / or controllers of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus (e.g., a controller) so that the instructions executed via the processor of the computer or other programmable data processing apparatus (e.g., GPU 3401, CPU 3415) create circuitry or means for implementing the functions / operations specified in the blocks of the flowchart and / or block diagram, thereby producing a machine. In certain embodiments, the computer program instructions may be configured to send sorting instructions to a sorting device to sort specific pieces of material from a plurality of pieces of material in order to generate a set of pieces of material having a predetermined specific chemical composition of the aggregate.

[0113] Note also that each block of the block diagram and / or flowchart diagram, and combinations of blocks in the block diagram and / or flowchart diagram, can be implemented by a dedicated hardware-based system that performs the specified function or operation (which may include, for example, one or more graphics processing units (such as GPU 3401)), or by a combination of dedicated hardware and computer instructions. For example, a module can be implemented as a hardware circuit that includes off-the-shelf semiconductors such as custom VLSI circuits, gate arrays, logic chips, transistors, controllers, or other discrete components. A module can also be implemented in programmable hardware devices such as field programmable gate arrays, programmable array logic, programmable logic devices, and the like.

[0114] The computer program code, i.e., instructions, for performing the operations of the aspects of the present disclosure can be described in any combination of one or more programming languages, including object-oriented programming languages such as Java, Smalltalk, Python, C++, the "C" programming language or similar programming languages, or conventional procedural programming languages such as any of the machine learning software disclosed herein. The program code can be executed entirely on the user's computer system as a stand-alone software package, partly on the user's computer system, partly on a separate computer system used by the user, partly on a remote computer system (e.g., a computer system used for training a sensor system), or entirely on a remote computer system or server. In the latter scenario, the remote computer system can be connected to the user's computer system via any type of network, including a local area network ("LAN") or a wide area network ("WAN"), or can also be connected to an external computer system (e.g., via the Internet using an Internet service provider).

[0115] These program instructions can also be stored in a computer-readable storage medium that can cause a computer system, other programmable data processing apparatus, a controller, or other device to function in a particular manner so that the instructions stored in the computer-readable medium produce a product that includes instructions for implementing the functions / operations specified in the blocks of the flowchart and / or block diagram.

[0116] One or more databases can be included in the host to store data of various implementations and provide access to the data. Those skilled in the art will understand that, for security reasons, any database, system, or component of the present disclosure may include any combination of databases or components located in a single location or multiple locations, and each database or system may include any of various suitable security functions such as firewalls, access codes, encryption, decryption, etc. The database can be any type of database, such as relational, hierarchical, object-oriented, etc. Common database products that can be used for database implementation include IBM's DB2, database products available from Oracle Corporation, Microsoft Access from Microsoft Corporation, or other database products. The database can be organized in a suitable manner, such as data tables and lookup tables.

[0117] The association of specific data (e.g., between classified pieces of material and their known chemical compositions, or between classified pieces of material and their calculated approximate masses) can be achieved through any data association technique known and practiced in the art. For example, the association can be performed manually or automatically. Automatic association techniques can include, for example, database searches, database merges, GREP, AGREP, SQL, etc. The association step can be achieved, for example, by a database merge function that uses the key fields of the respective data tables of a manufacturer and a retailer. The key fields divide the database according to the high-level classes of the objects defined by the key fields. For example, a specific class can be specified as the key field for both a first data table and a second data table, and the two data tables can be merged based on the class data of the key fields. In these embodiments, it is preferred that the data corresponding to each key field of the merged data table is the same. However, data tables having data similar to the key fields, although not identical, can also be merged, for example, using AGREP.

[0118] As used herein, reference is made to a device being "configured to" or a device being "configured" to perform some function. It is to be understood that this can include selecting pre-defined logic blocks and logically associating them so as to provide a specific logic function, including a monitoring or control function. It may also include programming the computer software-based logic of a control device, wiring individual hardware components, or any or all of the foregoing combinations.

[0119] In the description of this specification, to provide a complete understanding of the embodiments of the present disclosure, many specific details such as examples of programming, software modules, user selections, network transactions, database queries, database structures, hardware modules, hardware circuits, hardware chips, controllers, etc. are provided. However, those skilled in the art will recognize that the present disclosure can be implemented without one or more specific details, or using other methods, components, materials, etc. In other cases, well-known structures, materials, or operations may not be illustrated or described in detail to avoid obscuring aspects of the present disclosure.

[0120] Those skilled in the art should understand that various settings and parameters of the components of system 100 (including neural network parameters) can be customized, optimized, and reconfigured over time based on the type of materials to be classified and sorted, the desired classification and sorting results, the type of equipment being used, the empirical results of previous classifications, the available data, and other factors.

[0121] References throughout this specification to "one embodiment", "an embodiment", or similar terms mean that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present disclosure. Thus, the appearances of "in one embodiment", "in an embodiment", "an embodiment", "a particular embodiment", "various embodiments", and similar terms throughout this specification may all refer to the same embodiment, but not necessarily. Furthermore, the described features, structures, aspects, and / or characteristics of the present disclosure can be combined in any suitable manner in one or more embodiments. Accordingly, even if a function was initially claimed to function in a particular combination, in some cases, one or more features can be deleted from the combination, and the claimed combination can be directed to a sub-combination or a variation of a sub-combination.

[0122] Advantages, benefits, and problem solutions are described herein with respect to specific embodiments. However, advantages, benefits, problem solutions, and elements where advantages, benefits, or solutions may occur or become more prominent should not be construed as important, necessary, or essential features or elements of some or all of the claims. Further, components described herein are not necessary for the practice of the disclosure unless explicitly stated as being essential or important.

[0123] Although this specification contains many details, these should not be construed as limiting the scope of the disclosure or the claims, but rather as descriptions of features specific to particular implementations of the disclosure. Headings in this specification may sometimes not be intended to limit the disclosure, embodiments of the disclosure, or other matters disclosed under the heading.

[0124] As used herein, the term "or" is intended to be inclusive, so that "A or B" includes A or B, as well as both A and B. When used in the context of a list of entities, the term "and / or" as used herein refers to entities that exist alone or in combination. Thus, for example, the phrase "A, B, C, and / or D" includes A, B, C, and D individually, as well as any combination and sub-combination of A, B, C, and D.

[0125] The terms used herein are for the purpose of describing particular embodiments only and are not intended to limit the disclosure. When used herein, the singular forms "a", "an", and "the" may be intended to include the plural as well, unless the context clearly dictates otherwise.

[0126] Corresponding structures, materials, acts, and equivalents of all means or step plus function elements in the claims below are intended to include any structures, materials, or acts for performing the functions in combination with other claimed elements as specifically claimed.

[0127] As used herein, each of the terms "controller", "processor", "memory", "neural network", "interface", "separator", "separating device", "separating apparatus", "device", "extrusion mechanism", "extrusion device", "imaging sensor", "bin", "container", "system", "circuit", etc. refers to a non-conventional device element that would be recognized and understood by one of ordinary skill in the art and is not used herein as a placeholder term or a placeholder term for the purpose of invoking 35 U.S.C. 112(f).

[0128] As used herein with respect to a specified property or situation, "substantially" refers to a degree of deviation that is sufficiently small so as not to measurably impair the specified property or situation. The exact degree of allowable deviation may depend in some cases on the particular situation.

[0129] As used herein, for convenience, a plurality of items, structural elements, components, exemplary fractions, and / or materials may be presented in a common list. However, these lists should be interpreted as if each member of the list were individually identified as a separate and unique member. Thus, the individual members of such lists should not be construed as being substantially equivalent to other members of the same list based solely on their representation within a common group without indication to the contrary.

[0130] Unless otherwise defined, all technical and scientific terms used herein (such as acronyms used for chemical elements in the periodic table) shall have the same meaning as commonly understood by one of ordinary skill in the art to which the presently disclosed subject matter belongs. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety unless a particular passage is cited. In case of conflict, the present specification, including definitions, will control. Further, the materials, methods, and examples are illustrative only and not intended to be limiting.

[0131] Unless otherwise described herein, many details regarding specific materials, processing acts, and circuits are conventional and can be found in textbooks and other sources of information on computing, electronics, and software technologies.

[0132] Unless otherwise indicated, all numbers expressing quantities of ingredients, reaction conditions, etc. used in this specification and the claims should be understood as being modified in all instances by the term “about”. Accordingly, unless indicated to the contrary, the numerical parameters set forth in this specification and the appended claims are approximations that may vary depending upon the desired properties sought to be obtained by the presently disclosed subject matter. As used herein, the term “about” when referring to a value or amount of mass, weight, time, volume, concentration, or percentage, such variations are appropriate for carrying out the disclosed methods, and in some embodiments include variations of ±20%, in some embodiments ±10%, in some embodiments ±5%, in some embodiments ±1%, in some embodiments ±0.5%, and in some embodiments ±0.1% from a particular amount. As used herein, the term “similar” may refer to values within a particular offset or percentage of each other (e.g., 1%, 2%, 5%, 10%, etc.).

Explanation of Reference Numerals

[0133] 100 System 101 Piece of Material 103 Conveyor system, conveyor belt 104 Conveyor system motor 105 Position detector 106 Tumbler / vibrator / singulator 107 Computer system 108 Automatic control system 109 Real-time camera 110 Optical recognition system, vision system 111 Material tracking measurement device 112 Control system 120 Sensor system 121 Radiation source 122 Power supply 124 Detector 125 Detector electronics 126…129 Separation device 136…139 Separation container 140 Separation container 3400 Computer system, data processing system 3401 Graphics processing unit, GPU 3405 Local bus 3412 User interface adapter 3413 Keyboard 3414 Mouse 3415 Processor (CPU) 3416 Display adapter 3420 Volatile memory, RAM 3425 Network (LAN) adapter 3430 I / O adapter 3431 Hard disk drive 3432 Tape drive 3435 Non-volatile memory, ROM 3440 Display

Claims

1. Determining an approximate mass of each of a plurality of material pieces, wherein at least one of the plurality of material pieces has a different material classification from other material pieces; Classifying each of the plurality of material pieces as belonging to one of a plurality of different material classifications; Separating a specific material piece from among the plurality of material pieces according to the determined approximate mass and classification of each of the plurality of material pieces, wherein the combination of the separated material pieces generates an aggregate of material pieces having a chemical composition of a predetermined specific aggregate; A method comprising the steps.

2. The method according to claim 1, wherein the separating step includes redirecting the specific material piece among the material pieces into a container.

3. The method according to claim 2, wherein the separating step includes continuously determining the chemical composition of the aggregate of the redirected material pieces.

4. Determining an approximate mass of each of a plurality of material pieces, wherein at least one of the plurality of material pieces has a different material classification from other material pieces; Classifying each of the plurality of material pieces as belonging to one of a plurality of different material classifications; Separating a specific material piece from among the plurality of material pieces according to the determined approximate mass and classification of each of the plurality of material pieces, wherein the separation generates an aggregate of material pieces having a chemical composition of a predetermined specific aggregate; Including, The separating step includes redirecting the specific material piece among the material pieces into a container, continuously determining the chemical composition of the aggregate of the redirected material pieces, and redirecting the next material piece into the container to increase the weight percentage of a specific chemical element of the chemical composition of the aggregate of the redirected material pieces. A method comprising the steps.

5. Determining an approximate mass of each of a plurality of material pieces, wherein at least one of the plurality of material pieces has a different material classification from other material pieces; Classifying each of the plurality of material pieces as belonging to one of a plurality of different material classifications; Separating a specific piece of material from the plurality of pieces of material according to the determined approximate mass and classification of each piece of material of the plurality of pieces of material, the separating comprising generating an aggregate of pieces of material having a chemical composition of a predetermined specific aggregate. comprising The separating step includes redirecting the specific piece of material among the pieces of material into a container, continuously determining the chemical composition of the aggregate of the redirected piece of material, and not redirecting the next piece of material into the container in order to reduce the weight percentage of a specific chemical element of the chemical composition of the aggregate of the redirected piece of material. A method.

6. The method according to claim 3, wherein the separating step includes not redirecting the next piece of material into the container because the chemical composition of the predetermined specific aggregate contains undesirable contaminants.

7. A step of determining the approximate mass of each piece of material of a plurality of pieces of material, at least one of the plurality of pieces of material having a different material classification from other pieces of material, classifying each piece of material of the plurality of pieces of material as belonging to one of a plurality of different material classifications, Separating a specific piece of material from the plurality of pieces of material according to the determined approximate mass and classification of each piece of material of the plurality of pieces of material, the separating comprising generating an aggregate of pieces of material having a chemical composition of a predetermined specific aggregate. comprising The separating step includes redirecting the specific piece of material among the pieces of material into a container and continuously determining the chemical composition of the aggregate of the redirected piece of material, and the separating step continues until the chemical composition of the aggregate of a predetermined minimum number of redirected pieces of material equals a threshold level of the chemical composition of the predetermined specific aggregate. A method.

8. A step of determining the approximate mass of each piece of material of a plurality of pieces of material, at least one of the plurality of pieces of material having a different material classification from other pieces of material, classifying each piece of material of the plurality of pieces of material as belonging to one of a plurality of different material classifications, A step of separating a specific material piece from among the plurality of material pieces according to the determined approximate mass and classification of each material piece of the plurality of material pieces, wherein the separation generates an aggregate of material pieces having a chemical composition of a predetermined specific aggregate. comprising A method, wherein the aggregate of material pieces having a chemical composition of a predetermined specific aggregate includes at least one material piece having a material classification different from other material pieces in the aggregate. **Claim 9** The method according to claim 1, wherein the plurality of material pieces includes material pieces having different metal alloy compositions. **Claim 10** The method according to claim 1, wherein the chemical composition of the predetermined specific aggregate is different from the chemical composition of each of the plurality of material pieces. **Claim 11** A step of determining an approximate mass of each material piece of a plurality of material pieces, wherein at least one of the plurality of material pieces has a material classification different from other material pieces, a step of classifying each material piece of the plurality of material pieces as belonging to one of a plurality of different material classifications, a step of separating a specific material piece from among the plurality of material pieces according to the determined approximate mass and classification of each material piece of the plurality of material pieces, wherein the separation generates an aggregate of material pieces having a chemical composition of a predetermined specific aggregate. comprising A method, wherein the chemical composition of the predetermined specific aggregate is different from the chemical composition of all aggregates of the plurality of material pieces. **Claim 12** A step of determining an approximate mass of each material piece of a plurality of material pieces, wherein at least one of the plurality of material pieces has a material classification different from other material pieces, a step of classifying each material piece of the plurality of material pieces as belonging to one of a plurality of different material classifications, a step of separating a specific material piece from among the plurality of material pieces according to the determined approximate mass and classification of each material piece of the plurality of material pieces, wherein the separation generates an aggregate of material pieces having a chemical composition of a predetermined specific aggregate. comprising A method, wherein the aggregate of material pieces includes material pieces having different material classifications. **Claim 13** The method according to claim 12, wherein the aggregate of material pieces includes at least one of the material pieces having a material classification different from other material pieces. Step 14 of determining the approximate mass of each of a plurality of material pieces, wherein at least one of the plurality of material pieces has a material classification different from other material pieces. Step of classifying each of the plurality of material pieces as belonging to one of a plurality of different material classifications. Step of separating a specific material piece among the plurality of material pieces according to the determined approximate mass and classification of each of the plurality of material pieces, wherein the separation generates an aggregate of material pieces having a chemical composition of a predetermined specific aggregate. Including The plurality of material pieces include brazed aluminum alloy pieces and cast aluminum alloy pieces. The aggregate of material pieces includes at least one brazed aluminum alloy piece and at least one cast aluminum alloy piece. The chemical composition of the predetermined specific aggregate is different from the chemical composition of the brazed aluminum alloy piece, and the chemical composition of the predetermined specific aggregate is different from the chemical composition of the cast aluminum alloy piece.

15. The method according to claim 1, wherein the classifying step includes processing image data captured from each of the plurality of material pieces through a machine learning system.

16. A sensor configured to capture one or more characteristics of each of a mixture of material pieces, wherein the mixture of material pieces includes material pieces having different material classifications. A data processing system configured to classify each of the material pieces of the mixture of material pieces as belonging to one of a plurality of different material classifications. A separation device configured to separate a specific material piece among the material pieces from the mixture of material pieces according to the classification of each of the material pieces of the mixture of material pieces, wherein an aggregate of material pieces having a chemical composition of a predetermined specific aggregate is generated by the combined separated material pieces. A system comprising

17. The sensor is a camera, and the one or more captured characteristics are captured by the camera configured to capture respective images of the mixture of the pieces of material as the mixture of the pieces of material is conveyed past the camera. The camera is configured to capture respective visual images of the mixture of the material and generate image data, and the characteristics are visually observable characteristics. The system according to claim 16.

18. The data processing system comprises a machine learning system implementing a neural network configured to classify each piece of material of the mixture of the pieces of material as belonging to one of a plurality of different material classifications based on the captured visually observable characteristics. The system according to claim 17.

19. The system according to claim 16, further comprising an apparatus configured to determine an approximate mass of each piece of material of a plurality of pieces of material, and the separation being performed according to the determined approximate mass and classification of each piece of material.

20. The system according to claim 19, wherein the apparatus comprises a line scanner configured to measure an approximate size of each piece of material.

21. A computer program product stored on a computer-readable storage medium, which when executed by a data processing system, determines an approximate mass of each piece of material of a plurality of pieces of material, at least one of the plurality of pieces of material having a different material classification from other pieces of material; classifies each piece of material of the plurality of pieces of material as belonging to one of a plurality of different material classifications; instructs to separate specific ones of the pieces of material from the plurality of pieces of material to generate an aggregate of pieces of material having a chemical composition of a predetermined specific aggregate, the separation being performed according to the determined approximate mass and classification of each piece of material of the plurality of pieces of material, the aggregate of pieces of material including pieces of material having different material classifications; A computer program product that performs a process including the above.

22. The computer program product according to claim 21, wherein the classifying includes processing image data captured from each of the plurality of pieces of material through a machine learning system.

23. The computer program product according to claim 21, wherein the chemical composition of the predetermined specific aggregate is different from the chemical composition of each of the plurality of material pieces.

Images