Devices, methods, systems, and non-transitory computer-readable media for automated resin verification

Abstract

An automated resin verification system includes a spectrometer configured to generate spectral data from a resin sample, a vision sensor with built-in artificial intelligence configured to scan a Certificate of Analysis (COA) and extract material identification information, a memory including a trained machine learning algorithm for classifying resin types and logic for controlling silo operations, and an electronic processor. The spectrometer operates in the near-infrared wavelength range of 700-1100 nm to capture spectral signatures that distinguish between different resin types such as polyethylene terephthalate (PET), recycled polyethylene terephthalate (rPET), and high-density polyethylene (HDPE). The system employs a dual verification approach wherein the spectral classification result is compared against the material type indicated on the COA. When the spectral classification matches the COA verification, the system automatically opens the appropriate silo for resin unloading. When the classifications do not match, the system prevents silo opening, reducing the risk of resin mix-ups.

Description

DEVICES, METHODS, SYSTEMS, AND NON-TRANSITORY COMPUTER- READABLE MEDIA FOR AUTOMATED RESIN VERIFICATION CROSS-REFERENCE TO RELATED APPLICATION

[0001] This application claims priority to, and the benefit of, U.S. Provisional Application No. 63 / 775,818, filed on March 21, 2025, which is hereby incorporated by reference in its entirety. FIELD OF THE INVENTION

[0002] The present disclosure relates to automated material verification systems, and more particularly to systems and methods for automated resin testing, classification, and sorting using near-infrared spectroscopy and artificial intelligence.BACKGROUND

[0003] Industrial facilities that receive bulk resin materials, such as polyethylene terephthalate (PET), recycled polyethylene terephthalate (rPET), and high-density polyethylene (HDPE), face significant challenges in verifying and sorting incoming materials. Conventional resin handling processes involve manual tasks that may require personnel to walk substantial distances and spend considerable time away from primary duties. For example, a manual resin handling process may involve walking 3.5 miles per week and spending 5.1 hours per week on resin receiving tasks, combined with driver wait times of approximately 20 minutes to begin unloading the bulk resin materials.

[0004] These manual processes create inefficiencies that impact productivity and operational efficiency. Additionally, the risk of resin mix-ups poses a serious threat to operational integrity, as mixing incompatible resin types can result in significant financial losses, potentially up to $300,000 or more per incident. Such mix-ups may occur when personnel incorrectly identify resin types or direct materials to incorrect storage silos.SUMMARY OF THE INVENTION

[0005] Existing verification approaches may include manual inspection, handheld scanning devices, or semi-automated systems that scan documentation without testing the actualmaterial. However, these approaches may still require significant manual intervention, may not verify the physical material against accompanying documentation, or may lack the precision needed to distinguish between similar resin types such as PET and rPET. Accordingly, there is a need for improved systems and methods for automated resin verification that can accurately classify resin types, verify material against documentation, and automatically control material routing to appropriate storage locations. Additionally, reducing driver wait times to less than 5 minutes represents a significant operational improvement target.

[0006] Embodiments described herein relate to an automated resin verification system. In some aspects, the system includes a spectrometer configured to generate spectral data from a resin sample, a vision sensor with built-in artificial intelligence configured to scan and verify a Certificate of Analysis (COA), a memory including a machine learning algorithm for classifying resin types and logic for controlling silo operations, and an electronic processor configured to control the spectrometer and the vision sensor to automate resin testing, classification, verification, and sorting.

[0007] In some aspects, the spectrometer operates in the near-infrared (NIR) wavelength range of 700-1100 nm to capture spectral signatures that distinguish between different resin types. The trained machine learning algorithm may employ one or more classification models including Support Vector Machines (SVM), Random Forest, K-Nearest Neighbors (KNN), Neural Networks, Gradient Boosting Machines (GBM), or XGBoost to classify resin samples as PET, rPET, or HDPE based on the spectral data.

[0008] In some aspects, the system employs a dual verification approach wherein the spectral classification result from the spectrometer analysis is compared against the material type indicated on the COA. When the spectral classification matches the COA verification, the system automatically opens the appropriate silo for resin unloading. When the classifications do not match, the system prevents silo opening, thereby reducing the risk of resin mix-ups.

[0009] In some aspects, the system may further include a pneumatic tube system configured to transport resin samples and documentation from a driver station to a quality laboratory for compliance and record-keeping purposes.

[0010] In some aspects, the system may include a radar level sensor configured to detect and monitor silo levels during unloading operations.

[0011] Other aspects of the embodiments will become apparent by consideration of the detailed description and accompanying drawings.BRIEF DESCRIPTION OF THE DRAWINGS

[0012] FIG. 1 illustrates a block diagram of an automated resin verification system, according to various aspects of the present disclosure.

[0013] FIG. 2 illustrates a block diagram of a server-side processing device of the system of FIG. 1, according to various aspects of the present disclosure.

[0014] FIG. 3 illustrates a spectrometer setup for spectral signature data collection, according to various aspects of the present disclosure.

[0015] FIG. 4 illustrates spectral signatures for resin types in the near-infrared wavelength range, according to various aspects of the present disclosure.

[0016] FIG. 5 illustrates an interior of a resin control office configured for automated resin verification operations, according to various aspects of the present disclosure.

[0017] FIG. 6 illustrates another view of the resin control office from an elevated angle, according to various aspects of the present disclosure.

[0018] FIG. 7 illustrates a spectrometer enclosure assembly and touchscreen monitor arranged on a stainless steel workstation within the resin control office, according to various aspects of the present disclosure.

[0019] FIG. 8 illustrates a first close-up view of a spectrometer enclosure assembly with a cover positioned over a sample holder, according to various aspects of the present disclosure.

[0020] FIG. 9 illustrates a second close-up view of the spectrometer enclosure assembly without the cover showing an empty sample holder, according to various aspects of the present disclosure.

[0021] FIG. 10 illustrates a third close-up view of the spectrometer enclosure assembly without the cover showing a sample holder containing resin pellets for testing, according to various aspects of the present disclosure.

[0022] FIG. 11 illustrates vision sensors with built-in artificial intelligence for Certificate of Analysis scanning, according to various aspects of the present disclosure.

[0023] FIG. 12 illustrates a pneumatic tube carrier for transferring samples and documentation to a quality laboratory, according to various aspects of the present disclosure.

[0024] FIG. 13 illustrates radar level sensors for detecting and monitoring silo levels, according to various aspects of the present disclosure.

[0025] FIG. 14 illustrates silo connection points configured for automated resin unloading operations, according to various aspects of the present disclosure.

[0026] FIG. 15 illustrates a flowchart of a method for automated resin verification, according to various aspects of the present disclosure.DETAILED DESCRIPTION

[0027] Before any examples are explained in detail, it is to be understood that the examples are not limited in application to the details of the configuration and arrangement of components set forth in the following description or illustrated in the accompanying drawings. The examples are capable of being practiced or of being carried out in various ways. Also, it is to be understood that the phraseology and terminology used herein are for the purpose of description and should not be regarded as limiting. The use of “including,” “comprising,” or “having” and variations thereof are meant to encompass the items listed thereafter and equivalents thereof as well as additional items. Unless specified or limited otherwise, the terms “mounted,” “connected,” “supported,” and “coupled” and variations thereof are used broadly and encompass both direct and indirect mountings, connections, supports, and couplings.

[0028] In addition, it should be understood that examples may include hardware, software, and electronic components or modules that, for purposes of discussion, may be illustrated and described as if the majority of the components were implemented solely in hardware. However, one of ordinary skill in the art, and based on a reading of this detailed description, would recognize that, in at least one example, the electronic-based aspects may be implemented in software (e.g., stored on non-transitory computer-readable medium) executable by one or more processing units, such as a microprocessor and / or application specific integrated circuits (“ASICs”). As such, it should be noted that a plurality of hardware and software based devices, as well as a plurality of different structural components, may be utilized to implement the examples. For example, “servers,” “computing devices,” “controllers,” “processors,” etc., described in the specificationcan include one or more processing units, one or more computer-readable medium modules, one or more input / output interfaces, and various connections (e.g., a system bus) connecting the components.

[0029] FIG. 1 illustrates a system 100, according to various aspects of the present disclosure. The system 100 includes a spectrometer 102, a vision sensor 104, a pneumatic tube system 106, a network 130, a server 135, a database 140, and a user interface device 150 (e.g., a computing device with a touchscreen display).

[0030] Each of the spectrometer 102, the vision sensor 104, and the pneumatic tube system 106 is configured to communicatively connect to the server 135 through the network 130 and provide information to the server 135. For example, the server 135 is configured to communicate with the spectrometer 102, the vision sensor 104, and the pneumatic tube system 106 over a wired connection or a wireless connection. In some examples, the wired connection may include an Ethernet connection or other suitable wired connection. In some examples, the connection may include a wide area network (“WAN”) (e.g., a TCP / IP based network), a local area network (“LAN”), a neighborhood area network (“NAN”), a home area network (“HAN”), or personal area network (“PAN”) employing any of a variety of communications protocols, such as Wi-Fi, Bluetooth, ZigBee, etc.

[0031] In the system 100 of FIG. 1, the server 135 represents a server that is hosting an automated resin verification system. The server 135 is configured to receive and analyze data from the spectrometer 102, the vision sensor 104, and the pneumatic tube system 106. In some examples, the server 135 is configured to store data in the database 140 for future retrieval and analysis. The user interface device 150 may be used, for example, by an operator or analyst to review a status of the automated resin verification system.

[0032] The network 130 is, for example, a wide area network (“WAN”) (e.g., a TCP / IP based network), a local area network (“LAN”), a neighborhood area network (“NAN”), a home area network (“HAN”), a personal area network (“PAN”) employing any of a variety of communications protocols, such as Wi-Fi, Bluetooth, ZigBee, etc., or a wireless cellular network. The wireless cellular network may be a Global System for Mobile Communications (“GSM”) network, a General Packet Radio Service (“GPRS”) network, a Code Division Multiple Access (“CDMA”) network, an Evolution-Data Optimized (“EV-DO”) network, an EnhancedData Rates for GSM Evolution (“EDGE”) network, a 3GSM network, a 4GSM network, a 4G LTE network, a 5GNew Radio network, a Digital Enhanced Cordless Telecommunications (“DECT”) network, a Digital AMPS (“IS-136 / TDMA”) network, or an Integrated Digital Enhanced Network (“iDEN”) network, or other suitable wireless cellular network.

[0033] The user interface device 150 includes a combination of digital and analog input or output devices required to achieve a desired level of control and monitoring for the system 100. For example, the user interface device 150 may be a desktop computer, a laptop computer, a tablet computer, a personal digital assistant (“PDA”), a mobile phone (e.g., a smart phone), a touchscreen display, or other suitable user interface device.

[0034] FIG. 2 illustrates the server-side of the system 100 with respect to the server 135 of FIG. 1, according to various aspects of the present disclosure. The server 135 is electrically and / or communicatively connected to a variety of modules or components of the system 100. For example, the server 135 is connected to the database 140. The server 135 includes a controller 200, a power supply module 205, and a network communications module 210. The controller 200 includes combinations of hardware and software that are operable to, for example, process sensor data to perform operations to operate an automated resin verification system. The controller 200 includes a plurality of electrical and electronic components that provide power and operational control to the components and modules within the controller 200 and / or the system 100. For example, the controller 200 (i.e., an electronic processor) includes, among other things, a processing unit 215 (e.g., a microprocessor, a microcontroller, or another suitable programmable device), a memory 220, input units 225, and output units 230. The processing unit 215, the memory 220, the input units 225, and the output units 230, as well as the various modules connected to the controller 200 are connected by one or more control and / or data buses (e.g., common bus 250). The control and / or data buses are shown schematically in FIG. 2 for illustrative purposes.

[0035] The memory 220 is a non-transitory computer readable medium and includes, for example, a program storage area and a data storage area. The program storage area and the data storage area can include combinations of different types of memory, such as read-only memory (“ROM”), random access memory (“RAM”) (e.g., dynamic RAM [“DRAM”], synchronous DRAM [“SDRAM”], etc.), electrically erasable programmable read-only memory (“EEPROM”),flash memory, a hard disk, an SD card, or other suitable magnetic, optical, physical, electronic memory devices, or other data structures. The processing unit 215 is connected to the memory 220 and executes software instructions that are capable of being stored in a RAM of the memory 220 (e.g., during execution), a ROM of the memory 220 (e.g., on a generally permanent basis), or another non-transitory computer readable data storage medium such as another memory or a disc.

[0036] In some examples, the controller 200 or network communications module 210 includes one or more communications ports (e.g., Ethernet, serial advanced technology attachment [“SATA”], universal serial bus [“USB”], integrated drive electronics [“IDE”], etc.) for transferring, receiving, or storing data associated with the system 100 or the operation of the system 100. In some examples, the network communications module 210 includes an application programming interface (“API”) for the server 135. Software included in the implementation of the system 100 can be stored in the memory 220 of the controller 200. The software includes, for example, firmware, one or more applications, program data, filters, rules, one or more program modules, and other executable instructions. The controller 200 is configured to retrieve from memory and execute, among other things, instructions related to the control methods and processes described herein. In some examples, the controller 200 includes a plurality of processing units 215 and / or a plurality of memories 220 for retrieving from memory and executing the instructions related to the control methods and processes described herein.

[0037] The power supply module 205 supplies a nominal AC or DC voltage to the controller 200 or other components or modules of the system 100. The power supply module 205 is powered by, for example, mains power having nominal line voltages between 100V and 240V AC and frequencies of approximately 50-60Hz. The power supply module 205 is also configured to supply lower voltages to operate circuits and components within the controller 200 or system 100.

[0038] The controller 200 can include various modules and submodules related to implementing the system 100.

[0039] FIG. 3 illustrates a spectrometer system 300 for spectral signature data collection, according to various aspects of the present disclosure. The spectrometer system 300 includes a computing device 302 (e.g., a laptop computing device) connected via a cable to a spectrometer unit 304. The spectrometer unit 304 is connected to a sample analysis chamber 306 with a hinged lid that can be opened to place resin samples for testing. A sample holder is positioned within theanalysis chamber to receive resin pellets. A light source unit 308 is connected to the sample analysis chamber 306 via a cable and provides near-infrared illumination in the wavelength range of 700-1100 nm for spectral analysis.

[0040] In some examples, the spectrometer system 300 is one example of the spectrometer 102 of FIG. 1. The spectrometer system 300 is configured to analyze resin samples and generate spectral data. In some examples, the spectrometer unit 304 includes a near-infrared (NIR) light emitting diode (LED) operating in the wavelength range of 700-1100 nm, a diffraction grating for separating light into its component wavelengths, photodiodes configured to detect reflected light in the 700-1100 nm range, a microcontroller for processing signals, an analog-to-digital converter for converting analog signals to digital data, a PLC communication interface for transmitting classification results, optical fiber or tubing for directing light, a sample holder for receiving resin pellets, a housing enclosing the optical components, a power supply, and connecting wires.

[0041] In operation, a resin sample is placed in the sample holder. The sample holder may be a circular, cup-shaped container configured to hold resin pellets. The sample holder is positioned such that the NIR light source illuminates the resin sample. The photodiodes detect the reflected light and generate spectral data representing the reflectance or absorbance characteristics of the resin sample across the NIR wavelength range. This spectral data is then processed by the spectrometer unit 304, using the trained machine learning algorithm, to classify the resin type.

[0042] In some examples, the system 300 includes an enclosure that houses the spectrometer 102, the light source, and associated electronics (e.g., FIGS. 5-10). The enclosure may include a removable cover that can be placed over the sample holder during testing. Testing may be performed with the cover in place or removed, as the system may maintain accuracy in either configuration. A user interface, such as a touchscreen display, may be mounted on or near the enclosure to allow an operator to initiate testing by pressing a test button. Upon pressing the test button, the spectrometer unit 304 performs spectral analysis and the machine learning algorithm classifies the material, typically within approximately 10 seconds.

[0043] FIG. 4 illustrates spectral signatures 402-406 for resin types in the near-infrared wavelength range, according to various aspects of the present disclosure. The spectral signatures demonstrate that each resin type possesses unique absorption or reflectance characteristics withinthe 700-1100 nm wavelength range, enabling differentiation between a PET spectral signature 402, an HDPE spectral signature 404, and a rPET spectral signature 406 based on their distinct peak positions and intensities.

[0044] In some examples, the PET spectral signature 402 exhibits a single prominent peak at approximately 1050 nm. The HDPE spectral signature 404 exhibits a distinct peak centered at approximately 900 nm. The rPET spectral signature 406 exhibits two separate peaks: a smaller peak at approximately 750 nm and a larger peak at approximately 1030 nm. These characteristic spectral differences form the basis for automated material identification and classification using near-infrared spectroscopy.

[0045] The variation between PET and rPET spectral signatures may result from chemical degradation causing changes in polymer chains and formation of new byproducts, additives and contaminants introduced during recycling, molecular structure changes including variations in chain length and molecular weight, physical changes such as altered color and physical state, blending with other materials, processing condition variations including temperature, pressure, and technique, and aging and environmental factors such as exposure to light, heat, and pollutants.

[0046] The machine learning algorithm is trained using spectral data from known resin samples and processes the spectral data to classify the resin samples as rPET, PET, or HDPE. The classification may employ one or more machine learning models, including Support Vector Machines (SVM), Random Forest, K-Nearest Neighbors (KNN), Neural Networks, Gradient Boosting Machines (GBM), and XGBoost. The algorithm employs feature extraction techniques to analyze key spectral characteristics, such as peak wavelengths and intensity values, which are used for distinguishing between different resin types. These techniques may identify subtle differences in spectral signatures that are unique to rPET, PET, and HDPE. In some aspects, the algorithm may be retrained with additional scans to improve classification accuracy over time.

[0047] FIG. 5 illustrates an interior of a resin control office 500 configured for automated resin verification operations, according to various aspects of the present disclosure. The resin control office 500 illustrates another example of the spectrometer 102 and the vision sensor 104 of FIG. 1. The resin control office 500 includes a work surface 502 positioned against a back wall. On the work surface 502, a touchscreen monitor 504 (e.g., similar to the computing device302 of FIG. 3) displays a user interface for resin verification operations. Adjacent to the touchscreen monitor 504 is a spectrometer enclosure assembly 508 (e.g., housing the spectrometer unit 304, sample analysis chamber 306, light source unit 308, and sample holder 310 of FIG. 3) with a hinged lid and a cover that covers the sample holder. Mounted on a vertical support structure on the right side of the work surface 502 is a vision sensor assembly 516 (e.g., similar to the vision sensor 104 of FIG. 1) with a control panel 510 including a touchscreen display and a push button. Above the vision sensor assembly 516, a human-machine interface (HMI) display 512 is mounted providing operational feedback. The vision sensor assembly 516 is positioned to scan a Certificate of Analysis (COA) placed within a designated scanning area 518.

[0048] FIG. 6 illustrates another view of the resin control office 500 from an elevated angle, according to various aspects of the present disclosure. FIG. 6 illustrates another view of the system 100 of FIG. 1. A stainless steel work surface 502 supports the touchscreen monitor 504 displaying a resin verification interface and the spectrometer enclosure assembly 508. Mounted on a vertical support frame adjacent to the stainless steel work surface 502 is the vision sensor assembly 516 containing a touchscreen display 512 showing operational data, a green push button 524 for initiating operations, and keys for access control. The vision sensor assembly 516 and the designated scanning area 518 for COA verification are visible on the right side of the work surface 502.

[0049] FIG. 7 illustrates the spectrometer enclosure assembly 508 and the touchscreen monitor 504 arranged on the stainless steel workstation 520 within the resin control office 500, according to various aspects of the present disclosure. FIG. 7 illustrates components of the system 100 of FIG. 1. The spectrometer enclosure assembly 508 is a rectangular housing with sample analysis opening 526 on its front face. A cable 528 extends from the spectrometer enclosure assembly 508 for electrical and data connections. Adjacent to the spectrometer enclosure assembly 508, the touchscreen monitor 504 shows a user interface with a blue background displaying text indicating resin verification functionality.

[0050] FIGs. 8-10 illustrate close-up views of the spectrometer enclosure assembly 508, according to various aspects of the present disclosure. The spectrometer enclosure assembly 508 houses components of the spectrometer 102 of the system 100 of FIG. 1. The spectrometerenclosure assembly 508 comprises a rectangular housing 530 with a hinged lid 532 and latching mechanisms 534 on the sides. A sample analysis compartment 536 is visible on the front lower portion of the housing 530. As shown in FIG. 8, a cover 538 may be placed over a circular material holder opening. As shown in FIG. 9, the cover 538 may be removed to reveal an empty circular cup-shaped sample holder 902 (e.g., similar to the sample holder 310 of FIG. 3) configured to receive resin pellets for spectral analysis. As shown in FIG. 10, the sample holder 902 may contain a plurality of resin pellets 1002 visible as round balls, and the sample holder 902 area is illuminated by the near-infrared light source (e.g., the light source unit 308 of FIG. 3) during spectral analysis.

[0051] FIG. 11 illustrates vision sensors 1102 and 1104 with built-in artificial intelligence for Certificate of Analysis (COA) scanning, according to various aspects of the present disclosure. The vision sensor units 1102 and 1104 include a rectangular housing with a camera lens positioned centrally on the front face, surrounded by multiple illumination elements configured to provide consistent lighting conditions for capturing images regardless of ambient lighting variations. Status indicator lights are visible on the housing, and connector ports on the bottom portion provide electrical connections.

[0052] FIG. 12 illustrates a pneumatic tube carrier 1200 for transferring samples and documentation to a quality laboratory, according to various aspects of the present disclosure. The pneumatic tube carrier 1200 includes end caps 1202 with openings for aerodynamic purposes or pressure equalization, a transparent cylindrical section 1204 for visibility of contents, and sealing rings 1206 for airtight seals necessary for pneumatic tube system operation. In operation, resin samples and COA documents are placed into the pneumatic tube carrier 1200 by the driver and transported to the quality laboratory. This process may comply with standards for storage of samples and delivery documentation.

[0053] FIG. 13 illustrates radar level sensors 1302 and 1304 for detecting and monitoring silo levels, according to various aspects of the present disclosure. The radar level sensors 1302 and 1304 include a metallic housing with threaded mounting connections for secure installation in industrial environments. The radar level sensors 1302 and 1304 may include illuminated indicator rings suggesting operational status and integrated digital display screens showing graphical interfaces with measurement readouts. The radar level sensors 1302 and 1304 are configured todetect and measure material levels within storage silos, enabling continuous monitoring and data gathering during unloading operations.

[0054] In some examples, the system 100 includes multiple silos for storing different resin types. For example, the system may include one silo for HDPE one silo for PET, and one silo for rPET materials. In some examples, two silos may be used interchangeably for PET or rPET. Each silo connection point may include a cylindrical coupling mechanism with colored bands for identification purposes and signal light towers containing red and green indicator lights to communicate operational status. The green light illuminates when the corresponding silo is authorized and open for unloading, while the red light indicates the connection is locked or unavailable.

[0055] In some examples, the system 100 is configured to store data including the spectrometer scans, the material that was predicted, and the date and time of verification. This data may be stored locally in the database 140 or uploaded to a cloud-based storage system for compliance and record-keeping purposes.

[0056] FIG. 14 illustrates silo connection points configured for automated resin unloading operations, according to various aspects of the present disclosure. Three silo connection points are arranged in a row: HDPE silo 1 designated for high-density polyethylene material, PET silo 2 designated for polyethylene terephthalate material, and PET silo 3 designated for polyethylene terephthalate or recycled polyethylene terephthalate material. Each silo connection point includes a cylindrical coupling mechanism with colored bands for identification purposes, a control box with indicator lights, and pneumatic valve assemblies. Signal light towers are mounted above each control box containing red and green indicator lights to communicate operational status. The green light illuminates when the corresponding silo is authorized and open for unloading, while the red light indicates the connection is locked or unavailable. The pneumatic valve assemblies include lever mechanisms for securing hose connections during unloading operations. This configuration relates to the silo arrangement described with respect to FIG. 1, restricting drivers to use only the authorized silo after the automated resin verification system confirms material classification and COA verification.

[0057] FIG. 15 illustrates a flowchart of a method 1500 for automated resin verification, according to various aspects of the present disclosure. The method 1500 is performed by an electronic processor of a computing device, such as the server 135 of FIGs. 1-2.

[0058] The method 1500 includes controlling a spectrometer to generate spectral data from a resin sample (at block 1502). For example, a driver arriving with resin material enters a resin control office and obtains a sample bag delivered with the load. The driver pours the resin sample into a sample holder (e.g., a circular-shaped cup), filling the sample holder (e.g., to the top and flush for consistent spectral analysis). The sample holder is positioned beneath the spectrometer, and the door of the resin control office may be shut to minimize ambient light interference with the spectrometer. An operator presses a green "Test" button on a touchscreen monitor to initiate testing, and the electronic processor controls the spectrometer to generate spectral data. In some examples, the spectral data comprises near-infrared spectral data in a wavelength range of 700-1100 nm.

[0059] The method 1500 includes classifying the resin sample using a machine learning algorithm based on the spectral data (at block 1504). The machine learning algorithm may employ one or more classification models including Support Vector Machines (SVM), Random Forest, K-Nearest Neighbors (KNN), Neural Networks, Gradient Boosting Machines (GBM), or XGBoost. The electronic processor classifies the material as polyethylene terephthalate (PET), recycled polyethylene terephthalate (rPET), or high-density polyethylene (HDPE), typically within approximately 10 seconds.

[0060] The method 1500 includes controlling a vision sensor with built-in artificial intelligence to scan a Certificate of Analysis (COA) and extract material identification information (at block 1506). For example, the operator locates the COA document from paperwork delivered with the load and places the COA within a designated scanning area defined by a blue-outlined boundary within an aluminum extrusion frame structure. The operator presses a green button on a human-machine interface (HMI) to initiate COA scanning.

[0061] The method 1500 includes comparing the classification of the resin sample with the material identification information extracted from the COA (at block 1508).

[0062] The electronic processor determines whether the classification matches the material identification information (at decision block 1510). When the classification matches, the method 1500 proceeds to transmitting a control signal to a Programmable Logic Controller (PLC) to unlock a silo corresponding to the classified resin type (at block 1512). Only one silo is unlocked for the driver to connect to and unload, preventing connection to incorrect silos. When the classification does not match, the method 1500 proceeds to preventing silo unlocking, thereby reducing the risk of resin mix-ups (at block 1514).

[0063] In some examples, the method 1500 further includes transporting, with a pneumatic tube system, the resin sample and the COA from the testing station to a quality laboratory for compliance and record-keeping purposes. The driver may then place the resin sample and COA documents into a pneumatic tube carrier for transport to the quality laboratory for compliance and record -keeping purposes.

[0064] Referring back to FIG. 1, the automated resin verification system 100 provides several operational improvements. By automating the testing, classification, and verification process, the system may eliminate motion waste associated with manual processes and reduce driver wait times. The dual verification approach, combining spectral analysis with COA verification, reduces the risk of resin mix-ups that could result in significant financial losses.

[0065] The system 100 is applicable in various industrial settings where resin materials are used, including plastics manufacturing for sorting and verifying different types of plastic resins, recycling facilities for automating the sorting of recyclable materials based on resin type, and chemical processing where precise material classification is relevant to product quality.

[0066] In some examples, the machine learning algorithm is trained using spectral data from known resin samples. The algorithm may be retrained with additional scans to improve classification accuracy overtime. The system may be configured to classify additional resin types beyond PET, rPET, and HDPE by training the algorithm on spectral data for the additional materials.

[0067] In some embodiments, the system 100 may be extended to verify other materials beyond plastic resins. For example, similar spectral analysis and verification techniques may beapplied to verify different types of liquids, such as different milk types (skim milk, 1%, 2%, whole milk, etc.), by training the machine learning algorithm on spectral data for the different liquid types.

[0068] Example 1: a system comprising: a spectrometer configured to generate spectral data from a resin sample; a memory including a trained machine learning algorithm for classifying resin types based on the spectral data and logic for controlling silo operations; a vision sensor with built-in artificial intelligence configured to scan a Certificate of Analysis (COA) and extract material identification information; and an electronic processor configured to: control the spectrometer to generate the spectral data from the resin sample, classify the resin sample using the trained machine learning algorithm based on the spectral data, control the vision sensor to scan the COA and extract the material identification information, compare a classification of the resin sample with the material identification information from the COA, and automatically open a silo corresponding to the classified resin type when the classification matches the material identification information.

[0069] Example 2: the system of example 1, wherein the spectrometer is a near-infrared spectrometer configured to operate in a wavelength range of 700-1100 nm.

[0070] Example 3: the system of examples 1 or 2, wherein the trained machine learning algorithm comprises one or more of: Support Vector Machines (SVM), Random Forest, K-Nearest Neighbors (KNN), Neural Networks, Gradient Boosting Machines (GBM), or XGBoost.

[0071] Example 4: the system of any of examples 1-3, wherein the logic for controlling silo operations comprises Programmable Logic Controller (PLC) logic configured to receive the classification and the material identification information and transmit control signals to open the silo.

[0072] Example 5: the system of any of examples 1-4, wherein the electronic processor is further configured to prevent silo opening when the classification does not match the material identification information from the COA.

[0073] Example 6: the system of any of examples 1-5, wherein the trained machine learning algorithm is configured to classify the resin sample as one of: polyethylene terephthalate (PET), recycled polyethylene terephthalate (rPET), or high-density polyethylene (HDPE).

[0074] Example 7: the system of any of examples 1-6, further comprising a pneumatic tube system configured to transport the resin sample and the COA from a testing station to a quality laboratory.

[0075] Example 8: a method comprising: generating, with a spectrometer controlled by an electronic processor, spectral data from a resin sample; classifying, with the electronic processor, the resin sample using a trained machine learning algorithm based on the spectral data; scanning, with a vision sensor with built-in artificial intelligence controlled by the electronic processor, a Certificate of Analysis (COA) to extract material identification information; comparing, with the electronic processor, a classification of the resin sample with the material identification information from the COA; and automatically opening, with the electronic processor, a silo corresponding to the classified resin type when the classification matches the material identification information.

[0076] Example 9: the method of example 8, wherein generating the spectral data comprises generating near-infrared spectral data in a wavelength range of 700-1100 nm.

[0077] Example 10: the method of examples 8 or 9, wherein classifying the resin sample comprises classifying the resin sample using one or more of: Support Vector Machines (SVM), Random Forest, K-Nearest Neighbors (KNN), Neural Networks, Gradient Boosting Machines (GBM), or XGBoost.

[0078] Example 11: the method of any of examples 8-10, further comprising preventing silo opening when the classification does not match the material identification information from the COA

[0079] Example 12: the method of any of examples 8-11, wherein classifying the resin sample comprises classifying the resin sample as one of: polyethylene terephthalate (PET), recycled polyethylene terephthalate (rPET), or high-density polyethylene (HDPE).

[0080] Example 13: the method of any of examples 8-12, further comprising transmitting the classification and the material identification information to a Programmable Logic Controller (PLC) configured to control silo operations.

[0081] Example 14: the method of any of examples 8-13, further comprising transporting, with a pneumatic tube system, the resin sample and the COA from a testing station to a quality laboratory.

[0082] Example 15 : a non-transitory computer-readable medium comprising instructions that, when executed by an electronic processor, cause the electronic processor to perform a set of operations comprising: controlling a spectrometer to generate spectral data from a resin sample; classifying the resin sample using a trained machine learning algorithm based on the spectral data; controlling a vision sensor with built-in artificial intelligence to scan a Certificate of Analysis (COA) and extract material identification information; comparing a classification of the resin sample with the material identification information from the COA; and automatically opening a silo corresponding to the classified resin type when the classification matches the material identification information.

[0083] Example 16: the non-transitory computer-readable medium of example 15, wherein the spectral data comprises near-infrared spectral data in a wavelength range of 700-1100 nm.

[0084] Example 17: the non-transitory computer-readable medium of examples 15 or 16, wherein the set of operations further comprises preventing silo opening when the classification does not match the material identification information from the COA.

[0085] Example 18: the non-transitory computer-readable medium of any of examples 15-17, wherein classifying the resin sample comprises classifying the resin sample as one of: polyethylene terephthalate (PET), recycled polyethylene terephthalate (rPET), or high-density polyethylene (HOPE).

[0086] Example 19: a device comprising: a memory including a trained machine learning algorithm for classifying resin types based on spectral data and logic for controlling silo operations; and an electronic processor configured to: receive spectral data from a spectrometer that has analyzed a resin sample, classify the resin sample using the trained machine learning algorithm based on the spectral data, receive material identification information from a vision sensor that has scanned a Certificate of Analysis (COA), compare a classification of the resin sample with the material identification information from the COA, and generate a control signalto open a silo corresponding to the classified resin type when the classification matches the material identification information.

[0087] Example 20: the device of example 19, wherein the electronic processor is further configured to prevent generation of the control signal when the classification does not match the material identification information from the COA.

[0088] Accordingly, it is to be understood that the above description is intended to be illustrative and not restrictive. Many embodiments and applications other than the examples provided would be apparent upon reading the above description. The scope should be determined, not with reference to the above description, but should instead be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled. It is anticipated and intended that future developments will occur in the technologies discussed herein, and that the disclosed systems and methods will be incorporated into such future embodiments. In sum, it should be understood that the application is capable of modification and variation.

[0089] All terms used in the claims are intended to be given their broadest reasonable constructions and their ordinary meanings as understood by those knowledgeable in the technologies described herein unless an explicit indication to the contrary in made herein. In particular, use of the singular articles such as “a,” “the,” “said,” etc. should be read to recite one or more of the indicated elements unless a claim recites an explicit limitation to the contrary. It is to be understood that the claims are not limited to the precise configuration and components illustrated in the exemplary embodiments. A reference to an element by the indefinite article "a" or "an" does not exclude the possibility that more than one of the element is present, unless the context clearly requires that there be one and only one of the elements. Similarly, the use of "at least one" of an element is intended to be open-ended, meaning one or more of the element, and does not preclude the presence of other elements or steps. Furthermore, where a claim recites a group of items in a list or Markush group format (e.g., one selected from the group consisting of), it is intended to cover the use of one or more of the listed items, either individually or in any combination, unless otherwise specified. The scope of the disclosure is therefore to be determined not by the specific examples provided, but by the appended claims and their legal equivalents.

[0090] As used herein, terms such as "approximately," "about," or "substantially" when used in reference to a given measurement or value are intended to account for tolerances, measurement error, rounding, and manufacturing variances that are acceptable in the relevant technical field. Unless otherwise indicated, such terms are intended to encompass values within ±10% of the stated value, or within an industry-accepted tolerance range for the specified dimension or parameter. Similarly, expressions such as "equal to" or "equals" are not intended to require mathematical or absolute identity unless expressly stated and should be understood to allow for deviations that do not materially affect the performance or function described.

[0091] The Abstract is provided to allow the reader to quickly ascertain the nature of the technical disclosure. It is submitted with the understanding that it will not be used to interpret or limit the scope or meaning of the claims. In addition, in the foregoing Detailed Description, it can be seen that various features are grouped together in various embodiments for the purpose of streamlining the disclosure. This method of disclosure is not to be interpreted as reflecting an intention that the claimed embodiments require more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive subject matter lies in less than all features of a single disclosed embodiment. Thus, the following claims are hereby incorporated into the Detailed Description, with each claim standing on its own as a separately claimed subject matter.

[0092] Thus, embodiments described herein provide, among other things, systems, methods, devices, and computer readable media for automated resin verification.

Claims

CLAIMSWhat is claimed is:

1. A system comprising:a spectrometer configured to generate spectral data from a resin sample;a memory including a trained machine learning algorithm for classifying resin types based on the spectral data and logic for controlling silo operations;a vision sensor with built-in artificial intelligence configured to scan a Certificate of Analysis (COA) and extract material identification information; andan electronic processor configured to:control the spectrometer to generate the spectral data from the resin sample, classify the resin sample using the trained machine learning algorithm based on the spectral data,control the vision sensor to scan the COA and extract the material identification information,compare a classification of the resin sample with the material identification information from the COA, andautomatically open a silo corresponding to the classified resin type when the classification matches the material identification information.

2. The system of claim 1, wherein the spectrometer is a near-infrared spectrometer configured to operate in a wavelength range of 700-1100 nm.

3. The system of claim 1, wherein the trained machine learning algorithm comprises one or more of: Support Vector Machines (SVM), Random Forest, K-Nearest Neighbors (KNN), Neural Networks, Gradient Boosting Machines (GBM), or XGBoost.

4. The system of claim 1, wherein the logic for controlling silo operations comprises Programmable Logic Controller (PLC) logic configured to receive the classification and the material identification information and transmit control signals to open the silo.

5. The system of claim 1, wherein the electronic processor is further configured to prevent silo opening when the classification does not match the material identification information from the COA.

6. The system of claim 1, wherein the trained machine learning algorithm is configured to classify the resin sample as one of: polyethylene terephthalate (PET), recycled polyethylene terephthalate (rPET), or high-density polyethylene (HDPE).

7. The system of claim 1 , further comprising a pneumatic tube system configured to transport the resin sample and the COA from a testing station to a quality laboratory.

8. A method comprising:generating, with a spectrometer controlled by an electronic processor, spectral data from a resin sample;classifying, with the electronic processor, the resin sample using a trained machine learning algorithm based on the spectral data;scanning, with a vision sensor with built-in artificial intelligence controlled by the electronic processor, a Certificate of Analysis (COA) to extract material identification information;comparing, with the electronic processor, a classification of the resin sample with the material identification information from the COA; andautomatically opening, with the electronic processor, a silo corresponding to the classified resin type when the classification matches the material identification information.

9. The method of claim 8, wherein generating the spectral data comprises generating nearinfrared spectral data in a wavelength range of 700-1100 nm.

10. The method of claim 8, wherein classifying the resin sample comprises classifying the resin sample using one or more of: Support Vector Machines (SVM), Random Forest, K-Nearest Neighbors (KNN), Neural Networks, Gradient Boosting Machines (GBM), or XGBoost.

11. The method of claim 8, further comprising preventing silo opening when the classification does not match the material identification information from the COA.

12. The method of claim 8, wherein classifying the resin sample comprises classifying the resin sample as one of: polyethylene terephthalate (PET), recycled polyethylene terephthalate (rPET), or high-density polyethylene (HDPE).

13. The method of claim 8, further comprising transmitting the classification and the material identification information to a Programmable Logic Controller (PLC) configured to control silo operations.

14. The method of claim 8, further comprising transporting, with a pneumatic tube system, the resin sample and the COA from a testing station to a quality laboratory.

15. A non-transitory computer-readable medium comprising instructions that, when executed by an electronic processor, cause the electronic processor to perform a set of operations comprising:controlling a spectrometer to generate spectral data from a resin sample;classifying the resin sample using a trained machine learning algorithm based on the spectral data;controlling a vision sensor with built-in artificial intelligence to scan a Certificate of Analysis (COA) and extract material identification information;comparing a classification of the resin sample with the material identification information from the COA; andautomatically opening a silo corresponding to the classified resin type when the classification matches the material identification information.

16. The non-transitory computer-readable medium of claim 15, wherein the spectral data comprises near-infrared spectral data in a wavelength range of 700-1100 nm.

17. The non-transitory computer-readable medium of claim 15, wherein the set of operations further comprises preventing silo opening when the classification does not match the material identification information from the COA.

18. The non-transitory computer-readable medium of claim 15, wherein classifying the resin sample comprises classifying the resin sample as one of: polyethylene terephthalate (PET), recycled polyethylene terephthalate (rPET), or high-density polyethylene (HDPE).

19. The non-transitory computer-readable medium of claim 15, wherein the set of operations further comprises transmitting the classification and the material identification information to a Programmable Logic Controller (PLC) configured to control silo operations.

20. The non-transitory computer-readable medium of claim 15, wherein the set of operations further comprises controlling a pneumatic tube system to transport the resin sample and the COA from a testing station to a quality laboratory.

Images