System and method for monitoring chain health

Abstract

Systems and methods for monitoring chain health of a chain having a plurality of chain links in a closed loop as the chain is cycled. The system includes a camera positioned statically relative to the chain, the camera configured to generate frames to include a chain link of the plurality of chain links, and a controller module operably coupled to the camera and programmed to repeat the following steps: receive a frame; convert the frame to a black and white frame having high contrast; identify a chain link type of the chain link in the black and white frame based on a reference plurality of predefined skyline templates; and store the chain link with the chain link type and a corresponding time marker in a cache, a storage sequence. The system continues until the chain is found to have cycled back to a starting chain link, a chain link type of the chain link is identified, and a fingerprint of the chain.

Description

Technical Field

[0001] This invention generally relates to chains used in the manufacture of heavy equipment, and more specifically, to systems and methods for monitoring the health of conveyor chains. Background Technology

[0002] As used herein, a chain includes a conveyer chain and comprises multiple links assembled into a closed loop. Conveyer chains are typically configured to circulate on support structures in manufacturing environments and are frequently used to move objects or equipment.

[0003] Chain health directly impacts manufacturing efficiency. As a result of routine wear and tear, and due to defects or damage, chains can break unexpectedly or become unsuitable for their intended use / unstable; technical issues can lead to disruptions in large-scale manufacturing environments. The duration of such disruptions can be a function of the time spent locating and repairing defects or damage. Conveyor chains can exceed half a mile in length, making it a time-sensitive situation to pinpoint the defective link for manual intervention.

[0004] Therefore, there is a need to improve the systems and methods used for monitoring chain health. In addition to addressing the related problems, the following disclosure provides technical solutions to these technical problems. Furthermore, other desirable features and characteristics of the system and method will become apparent from the following detailed description and appended claims, taken in conjunction with the accompanying drawings and the foregoing background art. Summary of the Invention

[0005] In one embodiment, a method is provided for monitoring the chain health of a chain having multiple links in a closed loop during chain looping. The method includes: receiving a frame from a camera, which is statically positioned relative to the chain during chain looping, the camera being configured to generate frames to include links from the multiple links; converting the frame into a high-contrast black-and-white frame; identifying the link type of the links in the black-and-white frame based on references to multiple predefined skyline templates; and storing the links having link types and corresponding time stamps in a cache.

[0006] In one embodiment, the method further includes: locating long lines in a black and white frame; straightening the black and white frame when it is determined, based on the corresponding slope of the long lines, that the black and white frame needs to be straightened; isolating the frame skyline in the black and white frame; and wherein identifying the link type further includes comparing the frame skyline with multiple predefined skyline templates.

[0007] In the embodiment, the method further includes: for each of the plurality of predefined skyline templates, determining a fitting coefficient based on cross-frame skyline walking of the predefined skyline template; determining a stretching coefficient based on cross-frame skyline stretching of the predefined skyline template; wherein the link type is further identified based on the fitting coefficient and the stretching coefficient; and the link having the fitting coefficient and the stretching coefficient is further stored in a cache.

[0008] In an embodiment, the method further includes: determining that a link is defective when a black and white frame indicates that a link is missing a component or has a surface area defect based on the link type; and generating an alarm indicating that a link is defective.

[0009] In one embodiment, the method further includes: receiving user input defining a fault mode and fault mode conditions; determining that a fault mode condition exists in a black and white frame; and generating an alarm for the chain link and the fault mode.

[0010] In one embodiment, the method further includes: receiving user input defining a fault mode and fault mode conditions; determining that a fault mode condition exists in a black and white frame; and generating an alarm for the chain link and the fault mode.

[0011] In an embodiment, the method further includes: after storing the link, designating the link as the starting link; and (a) receiving at least one next frame from the camera, the at least one next frame including the next link among a plurality of links; (b) converting the at least one next frame into a next black and white frame with high contrast; (c) identifying the link type of the next link in the next black and white frame; and (d) storing the next link having the link type and a corresponding time stamp in a cache; (e) constructing a chain segment comprising a sequence of the link type of the next link and the link type of at least one previous link; and repeating (a)-(d) until it is found that the chain has looped back to the starting link based on the sequence in the chain segment.

[0012] In one embodiment, the method further includes identifying the chain type of the chain based on the sequence.

[0013] In one embodiment, a fingerprint of the chain, created in a unique chronological order, is also included, comprising all links of the chain.

[0014] In one embodiment, the fingerprint of the chain is also stored in a cache.

[0015] In one embodiment, the method further includes identifying missing or removed chain links based on the chain's fingerprint and generating a notification accordingly.

[0016] In one embodiment, the method further includes identifying defective links in the chain based on the chain's fingerprint and generating a notification accordingly.

[0017] In one embodiment, a system is provided for monitoring the health of a chain having multiple links in a closed loop during chain circulation. The system includes: a camera statically positioned relative to the chain during chain circulation, the camera being configured to generate frames to include links from the multiple links; and a controller module operatively coupled to the camera and programmed to: receive the frames; convert the frames into high-contrast black-and-white frames; identify the link types of the links in the black-and-white frames based on references to a plurality of predefined skyline templates; and store the links having link types and corresponding timestamps in a cache.

[0018] In an embodiment, the controller module is also programmed to: locate long lines in a black and white frame; straighten the black and white frame when it is determined, based on the respective slopes of the long lines, that the black and white frame needs to be straightened; isolate the frame skyline in the black and white frame; and further identify the chain type by comparing the frame skyline with multiple predefined skyline templates.

[0019] In an embodiment, the controller module is also programmed to: determine a fitting coefficient based on cross-frame skyline walking of the predefined skyline template for each of a plurality of predefined skyline templates; determine a stretching coefficient based on cross-frame skyline stretching of the predefined skyline template; further identify the link type according to the fitting coefficient and the stretching coefficient; and further store the links with fitting coefficient and stretching coefficient in a cache.

[0020] In an embodiment, the controller module is also programmed to: determine that a link is defective when a black and white frame indicates that a link is missing a component or has a surface area defect based on the link type; and generate an alarm for the defective link.

[0021] In this embodiment, the controller module is also programmed to: receive user input defining fault modes and fault mode conditions; determine the existence of fault mode conditions based on black and white frames; and generate alarms for the chain and fault modes.

[0022] In this embodiment, the controller module is also programmed to: receive user input defining fault modes and fault mode conditions; determine the presence of fault mode conditions in black and white frames; and generate alarms for the chain links and fault modes.

[0023] In an embodiment, the controller module is further programmed to: after storing a link as the starting link, (a) receive a next frame from the camera, the next frame including the next link among a plurality of links; (b) convert the next frame into a next black and white frame with high contrast; (c) identify the link type of the next link in the next black and white frame; and (d) store the next link having the link type and a corresponding time stamp in a cache; (e) construct a chain segment including the link type of the next link and the link type of at least one previous link; and repeat (a)-(d) until it is found that the chain has looped back to the starting link based on the sequence in the chain segment.

[0024] In an embodiment, the controller module is also programmed to: identify the chain type of the chain based on a sequence; and create a fingerprint of the chain in a unique chronological order, the fingerprint including all the links of the chain. Attached Figure Description

[0025] Exemplary embodiments will now be described in conjunction with the following accompanying drawings, wherein the same reference numerals denote the same elements, wherein:

[0026] Figures 1 to 2 This is a diagram illustrating examples of links in a chain according to various embodiments;

[0027] Figure 3 This is a simplified diagram of a link for marking a link assembly according to various embodiments;

[0028] Figure 4 This is a schematic diagram of a system according to various embodiments, which is used to monitor the chain health of a chain with multiple links in a closed loop during chain circulation;

[0029] Figures 5 to 6 This is a process flowchart illustrating an example method according to various embodiments for monitoring the chain health of a chain having multiple links in a closed loop during chain circulation;

[0030] Figure 7 It is a description Figures 5 to 6 A diagram illustrating some steps in the method; and

[0031] Figure 8 and Figure 9 This is a diagram illustrating the walking skyline template according to various embodiments and determining the stretching factor and fitting factor. Detailed Implementation

[0032] The following detailed description is merely exemplary in nature and is not intended to limit application and use. Furthermore, it is not intended to be bound by any express or implied theory presented in the foregoing technical field, background, overview, or the following detailed description.

[0033] This document describes embodiments of the present disclosure based on functional and / or logical block components and various processing steps. It should be understood that such block components can be implemented by any number of hardware, software, and / or firmware components configured to perform specified functions. For example, embodiments of the present disclosure may employ various integrated circuit components, such as memory elements, digital signal processing elements, logic elements, lookup tables, etc., which can perform various functions under the control of one or more microprocessors or other control devices.

[0034] As used herein, the term "module" can refer to any hardware, software, firmware, electronic control components, processing logic, and / or processor device that provides functionality belonging to the module, individually or in any combination. In various embodiments, a module includes one or more of the following: application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), electronic circuitry, a computer system including a processor (shared, dedicated, or grouped) and memory executing one or more software or firmware programs, combinational logic circuitry, and / or other suitable components that provide functionality belonging to the module.

[0035] For the sake of brevity, conventional techniques related to signal processing, data transmission, signaling, control, machine learning models, radar, lidar, image analysis, and other functional aspects of the system (as well as the various operating components of the system) may not be described in detail herein. Furthermore, the connecting lines shown in the various figures contained herein are intended to represent exemplary functional relationships and / or physical couplings between the various elements. It should be noted that many alternative or additional functional relationships or physical connections may exist in the embodiments of this disclosure.

[0036] As mentioned above, technical problems can arise when using conveyor chains. Routine wear, defects, and damage can cause the chain to break unexpectedly or become unsuitable / unstable, potentially disrupting the manufacturing environment. The duration of such interruptions can be a function of the time spent locating and repairing defects or damage. Furthermore, since some conveyor chains exceed half a mile in length, locating identified defective links for manual intervention is a time-sensitive situation.

[0037] As used herein, a link is an assembled link comprising one or more link components. As those skilled in the art will understand, multiple links assembled into a closed loop form a chain. In various embodiments, the chain is a conveyor chain configured to circulate on a device to move objects.

[0038] An exemplary embodiment provides a technical solution to these problems in the form of a system for monitoring the chain health of a chain with multiple links in a closed loop during chain circulation (hereinafter referred to as System 402). Figure 4As used herein, in addition to detecting damaged and defective link assemblies, links, and chain segments, monitoring chain health also includes identifying the link type of each link in the chain, identifying the sequence of links in the chain, and identifying the entire sequence of links of a given chain. The provided embodiments monitor chain health at two levels (individual links and the entire chain) via the fingerprinting method described herein.

[0039] Advantageously, these embodiments can also detect configuration problems in the chain. Thus, the embodiments can determine when the chain itself is not faulty, but is not configured to follow the desired pattern (e.g., the desired pattern of a type 1 link A followed by a type 2 link B, but the embodiments detect a type 1 link C and a type 2 link D).

[0040] As background and context, Figures 1 to 2 Some non-limiting examples of links in a chain according to various embodiments are shown, and Figure 3 Components are provided to simplify and label chain links for further reference. It is understood that, depending on the link type, each chain may have its own corresponding sequence of links, as well as its own number of links. Figures 1 to 2 The image shows three exemplary link types: brackets indicate the boundaries of the front locking link 102, the partition link 114, and the rear locking link 104.

[0041] As described above, each link comprises multiple components, including one or more link plates. Link plates take various shapes, with form following function; however, for a given chain, link plates typically have the same length (e.g., length 310 equals length 312). In various chains, the link is four inches, measured from the center of the first bolt to the center of the second bolt. In other embodiments, the link is six inches. Non-limiting examples of link plates used herein include slide plate 106, top plate 116, partition plate 108, one or more link plates 118, and one or more link plates 120. Bolts 110 extend through the various link plates, pivotally connecting them to each other. Nuts 112 secure the components of the link in place. In some embodiments, partition plate 108 is integral with link plate 118, as shown in connecting segment 306. In some embodiments, bolt 110 is simply an extension of a dogbone hinge that joins two links together; thus, in some chain configurations, such as those without a slide plate, the dogbone hinge functions as bolt 110.

[0042] exist Figures 1 to 2In the image, the front locking link 102 includes a front locking claw 302 (via bolt 110 and nut 112) bolted to the slide plate 106. The rear locking link 104 includes a rear locking claw 304 bolted to the slide plate 106. Each slide plate 106 is separated from the top plate 116 by a partition 108.

[0043] It is understood that the exemplary chain represented by these links is a non-limiting example. Other chain configurations exist, such as those using locking claws; and chains that use pushers to slide things. Furthermore, sometimes the links are supported by wheels from above a trolley or below what appears to be a small electric train.

[0044] The dashed line represents frame 308 provided by the camera, as combined with... Figure 4 Further description. For example... Figure 4 As shown, an embodiment of system 402 includes a camera 406, a controller module 404, and one or more external sources 420, wherein the camera 406 is statically positioned relative to the chain 401 as the chain 401 cycles.

[0045] Camera 406 is configured to capture a continuous video feed and divide the video feed into frames. Camera 406 thereby generates frames 308 to include links (e.g., links 410, 412, 414, and 416) of a plurality of links comprising chain 401. In various embodiments, camera 406 is located in camera system 408 and is configured to capture frames of chain 401 (in response to configurable timing control), convert sensed information into frames of frame data, and transmit the frames to controller module 404, as shown by connection 418, which may be wired or wireless. In various systems 402, camera 406 may be an optical camera, an infrared camera, or other sensing camera. In various embodiments, camera 406 receives timing configuration and other commands (wired or wireless) from controller module 404. In other embodiments, camera system 408 is also wired or wirelessly connected to external source 420 and may receive the aforementioned timing configuration and other commands from controller module 404 or one or more external sources 420.

[0046] External source 420 may include user interface 422 or central command 424 (e.g., fleet control). External source 420 may include the Internet. User interface 422 may be part of a complete human-machine interface (HMI) and typically includes a display device and a user input device (UI). In various embodiments, the user input device may be any combination of a keyboard, cursor control device, voice input device, gesture input device, etc.

[0047] Various embodiments of system 402 may include one or more other components and / or systems that typically communicate with each other via an external source 420.

[0048] exist Figure 4 In this embodiment, central management of tasks / operations is provided by controller module 404. In other embodiments, centralized management of tasks / operations may be generated by controller circuitry, a PLA, custom circuitry, etc. In various embodiments, controller module 404 is communicatively coupled to camera system 408, particularly camera 406, via connection 418, and communicatively coupled to external source 420 via bus 426. In various embodiments, controller module 404 is configured to transmit commands, control, and power for various components of system 402.

[0049] In various embodiments, such as Figure 4 As shown, controller module 404 is implemented as an enhanced computer system, including a computer-readable storage device or medium, memory 54 for storing instructions, algorithms and / or programs 56, and operating parameters 58, such as pre-programmed models and configuration requirements. Controller module 404 also includes a processor 50 for executing program 56 and input / output interface (I / O) 52. The computer-readable storage device or medium, memory 54, may include volatile and non-volatile memory such as read-only memory (ROM), random access memory (RAM), and keep-alive memory (KAM). KAM is a permanent or non-volatile memory that can be used to store various operating variables when processor 50 is powered off. Memory 54 can be implemented using any of several known memory devices, such as PROM (programmable read-only memory), EPROM (electrical PROM), EEPROM (electrically erasable PROM), flash memory, or any other electrical, magnetic, optical, or combined memory device capable of storing data, some of which represent executable instructions used by processor 50 when controlling the vehicle. In various embodiments, processor 50 is configured to implement system 402. The processor 50 can also utilize the memory 54 to cache data, temporarily store comparison and analysis results, etc. During method initialization or installation operations, the information in the memory 54 can be organized and / or imported from external sources; it can also be programmed via the user I / O interface.

[0050] In various embodiments, multiple predefined skyline templates are stored in parameter 58. Each predefined skyline template is unique for its respective link type and has the ideal geometry for that link type, including silhouette, length, and surface area smoothness requirements. As described in more detail below, controller module 404 uses the multiple predefined skyline templates to compare with the links in the frame to identify link types, make determinations about link integrity and health, and make determinations about the overall chain health.

[0051] Input / output interface (I / O) 52 may be operatively coupled to processor 50 via a bus and allows for both intra-module and extra-module communication. Input / output interface (I / O) 52 may include one or more wired and / or wireless network interfaces and may be implemented using any suitable methods and means. In various embodiments, input / output interface (I / O) 52 includes hardware and software to support one or more communication protocols for wireless communication between processor 50 and external sources such as satellites, clouds, communication towers, and ground stations. In various embodiments, input / output interface (I / O) 52 supports communication with technicians and / or one or more storage interfaces for direct connection to storage devices.

[0052] During operation of system 402, processor 50 loads and executes one or more algorithms, instructions, and rules embodied in program 56, and also controls the overall operation of system 402. During operation of system 402, processor 50 may receive data via bus 426 from connection 418 or external source 420. In various embodiments of system 402, controller module 404 may: perform operations belonging to system 402 according to algorithms; perform operations according to state machine logic; and perform operations according to logic in a programmable logic array.

[0053] While exemplary embodiments of system 402 are described in the context of controller module 404 implemented as a full-featured enhanced computer system, those skilled in the art will recognize that the mechanisms of this disclosure can be distributed as a program product including program 56 and predefined parameters. Such a program product may include an instruction arrangement organized into multiple interdependent program code modules, each configured to implement a separate process and / or perform a separate algorithmic operation, arranged to manage data flow through system 402. Each program code module may each include an ordered list of executable instructions for implementing the logical functions of the process executed by system 402. When executed by a processor (e.g., processor 50), the instructions in the program code modules cause the processor to receive and process signals and perform the logic, calculations, methods, and / or algorithms as described herein.

[0054] Once developed, the program code modules constituting the program product can be stored and distributed individually or together using one or more types of non-transitory computer-readable signal-bearing media (e.g., non-transitory computer-readable media) that can be used to store and distribute instructions. Such a program product can take many forms, and this disclosure applies equally to the type of computer-readable signal-bearing medium used for distribution. Examples of signal-bearing media include recordable media such as floppy disks, hard disks, memory cards, and optical disks, and transmission media such as digital and analog communication links. It should be understood that in some embodiments, cloud-based storage and / or other technologies can also be used as storage and time-based viewing license request program products.

[0055] Now go to Figures 5 to 6 And continue to refer to Figures 1 to 4 This describes various method steps for monitoring the chain health of chains with multiple links in a closed loop, typically as shown in method 500. (Reference) Figures 7 to 8 Let me explain the steps in method 500.

[0056] For illustrative purposes, the following description of method 500 may be referenced in conjunction with the above. Figures 1 to 4 The mentioned components. It should be understood that method 500 can include any number of additional or alternative operations and tasks. Figures 5 to 6 The tasks shown do not need to be performed in the order shown, and method 500 can be incorporated into more comprehensive programs or methods, such as energy-saving or security applications, with additional functionality not described in detail here. Furthermore, if the overall functionality is to remain intact, then... Figures 5 to 6 One or more tasks shown may be omitted from an embodiment of method 500.

[0057] At 502, the method begins by receiving frame 308 from camera 406 (e.g., Figure 7 (702), Camera 406 is statically positioned relative to chain 401 during chain looping, and Camera 406 is configured to generate frame 702 to include one of a plurality of chain links. In some embodiments, a whiteboard is placed in the background of the chain to ensure contrast for generating a frame skyline. At 504, the method converts the frame into a high-contrast black and white frame (e.g., Figure 7 (704). In this embodiment, contrast enhancement is performed by scanning the frame from left to right and from top to bottom, finding the edges between black and white and enhancing them. As used herein, high contrast is a conventional photographic term that means a limited range of tones, in this case, black and white, with bright white and dark shadows, and a contrast greater than a threshold contrast.

[0058] At 506, the method identifies the chain type of a chain in a black-and-white frame based on reference to multiple predefined skyline templates. For example, black-and-white frame 706 has indicated skyline template 708, which has been determined as the best match or fit for the black-and-white image. In various embodiments, task 506 includes one or more tasks. Some embodiments include identifying individual components of a chain in step 506. Some embodiments include locating individual components and features to identify the chain. Once the correct component of the chain type is determined to exist, the chain type can be identified.

[0059] Some embodiments perform rotation correction (also known as straightening) at 506. For rotation correction, method 500 locates long lines in the black-and-white frame. Long lines are defined as substantially straight horizontal lines that extend substantially across the frame, or lines longer than a threshold length. Reference Figure 3 For example, the long lines may include one or more of lines 710, 712, 714, 716, and 718. The slope of the long lines can then be compared to a slope threshold (e.g., ±2% from the horizontal direction) to determine if they are horizontally flat as required for the links to be considered straight rather than twisted. When it is determined that the black-and-white frame needs to be straightened based on the corresponding slope of the long lines, the method can correct this by rotating or straightening the black-and-white frame based on the determined slope.

[0060] Once any necessary straightening has been performed, method 500 can move to the step of isolating the frame skyline in the black and white frame. Isolating the frame skyline may include representing the outline using a single vector. Identifying the link type may further include comparing the frame skyline with multiple predefined skyline templates. An exemplary skyline template 708 is shown in... Figure 7 As shown in the image.

[0061] Another aspect of identifying link types in 506 may include determining fit and stretch (see [link]). Figure 8 and Figure 9 In an embodiment, at 506, method 500 can determine fitting coefficients for each of the N templates to identify chain configurations. For example, for each of a plurality of predefined skyline templates, system 402 can determine fitting coefficients based on cross-frame skyline walking skyline template 708 and detecting over / under-limit spaces or gaps. This is in Figure 8The diagram illustrates a skyline template 708 with a link length 802. The skyline template 708 is initially positioned at 804 and moves incrementally from left to right to the next position 806, continuing in increments of 808. These increments are configurable and can be stored along with other parameters 58. The fitting coefficient becomes 1 when the error or margin between the skyline template and the frame skyline becomes zero (perfect fit). A configurable fitting threshold can be used to compare the fitting coefficients, and when the fitting coefficient is greater than the threshold, the skyline template is determined to match the frame skyline. In this embodiment, the fitting coefficient is used to identify the link template. Figure 9 In the image at the top of the page, the skyline template 902 (dashed line) walks across the frame 904 from left to right, as shown in the image from top to bottom.

[0062] The stretch factor can be determined similarly. Once the link type has been determined, the skyline template is incrementally stretched across the frame skyline; for example, the skyline template 708 is stretched to have a length of 810 and made to walk across the frame skyline when over- or under-limit space or gaps are detected. The more the skyline template must be stretched to match the frame skyline (greater stretch), the lower the stretch factor. The stretch factor can be compared to a configurable stretch threshold, and when the stretch factor is less than the stretch threshold, it is determined that the frame skyline is stretched to a defective degree. In some embodiments, identifying the link type may further include selecting the skyline template that best matches the link in the frame based on the fitting coefficient and the stretch factor. Figure 9 In the image below, the skyline template 906 (dashed line) is stretched across frames 908 from left to right (each starting from the same walking point on the left) in a series of consecutive images from top to bottom.

[0063] In some embodiments, the fitting coefficient and stretching coefficient are determined simultaneously by the system. For example, using a skyline template of a first length (e.g., 100), the skyline template is slid (walked) across frames in increasing steps, and any error (difference between the frame skyline and the skyline template in each step of the cross-frame walk) is recorded. The template is then stretched (gradually increasing its length) to a length of 101, and the skyline template is again walked across frame skylines in frames, with the error recorded. This process is repeated to iterate through all length values ​​within a predefined range (e.g., 100-120) associated with the corresponding skyline template. After these multiple walks are completed, the system examines all collected error values ​​and determines the combination of 'walk' position and stretching length that produces the lowest error. The combined walk and stretch can be described in a programming loop as follows:

[0064] Loop WALKING_START from MIN_WALK_START to MAX_WALK_START

[0065] Loop STRETCH_FACTOR from MIN_STRETCH_FACTOR to MAX_STRETCH_FACTOR

[0066] Compare template to frame-skyline from WALKING_START with STRETCH_FACTOR

[0067] End Loop

[0068] End Loop

[0069] Use the best WALKING_START and STRETCH_FACTOR

[0070] The translation is as follows:

[0071] The loop from MIN_WALK_START to MAX_WALK_START, named WALKING_START.

[0072] The loop STRETCH_FACTOR runs from MIN_STRETCH_FACTOR to MAX_STRETCH_FACTOR.

[0073] Compare the templates and frames of WALKING_START and STRETCH_FACTOR to Skyline

[0074] End the loop

[0075] End the loop

[0076] Use the optimal WALKING_START and STRETCH_FACTOR

[0077] At 508, in various embodiments, the method uses a black-and-white frame 704 to determine if a link is defective. In some embodiments, the determination made at 508 is based on a predefined link length (e.g., such that a tensile defect can be identified as a defect). In some embodiments, the determination at 508 is based on a predefined surface area smoothness (e.g., such that a crack or deformation can be identified as a defect). In some embodiments, the determination made at 508 is based on noticing a missing, defective, or damaged component in the link (e.g., a broken bolt). Missing links will be described in detail below. In some embodiments, in addition to predefined parameters, a user fault mode is received via user interface 422, and the determination at 508 is further made based on criteria provided by the user fault mode. At 510, in various embodiments, when a defect is determined, an alarm is generated, which may be an email notification, a text notification, a graphical user interface object, or a light illuminated on a display. In various embodiments, the email notification includes frames from a video feed from camera 406.

[0078] In 512, the method continues to store links with their link types and corresponding timestamps in a cache. In embodiments where other determinations about links have been made, such as fit coefficients, stretching coefficients, the presence of defects, and, if so, the failure mode to which they are attributed, these determinations can also be stored along with the link, link type, and timestamp. All this data stored with the links can be referred to as the link's metadata. The fingerprint of the chain described herein is created from the metadata about each link in the chain, specifically the link type and stretching coefficient of the links in the chain.

[0079] Steps 502-512 can be used to identify the link type and detect the health of each link in the chain, and notifications and alerts can be generated based on the determination. Now proceeding to steps 602-620, which may optionally branch from step 512, method 500 can provide additional information about the entire chain.

[0080] At 602, the method may mark a link as the starting link and continue processing additional links. In some embodiments, multiple frames correspond to one link, so the method then moves to finding a new link. At 604, at least one next frame (e.g., at least one incrementing frame) is received, and at 606, method 500 may perform steps 504-510 on the at least one next frame to identify a link in the at least one next frame. At 608, the incrementing frame is stored along with the previous frame, and these consecutive frames form the beginning of the chain sequence.

[0081] Understandably, as the chain continues to loop, it will eventually return to the starting segment, and the current segment will begin to match the segments that have been saved in the sequence in history; this can be referred to as a fold or fingerprint overlap in the cache, thus creating a fingerprint of the chain.

[0082] In 610, the method uses a fingerprint to determine if it has "looped back to the starting link". In essence, the method folds the link samples over themselves to determine if it has found the starting link again. To do this, the method compares the first N links with the second N links in a loop, then compares the first N+2 links with the second N+2 links, and so on. The method then uses the lowest Average_Error calculated by these loops to determine the actual number of links in the optimal N-representation chain. Example loops are provided:

[0083] Loop N=SMALLEST_LEN to LARGEST_LEN by 2

[0084] Cumulative_Error = 0

[0085] Loop i = 1 to N

[0086] Cumulative_Error=Cumulative_Error+abs(link_stretches[i]–link_stretches[i+N])

[0087] End Loop

[0088] Average_Error=Cumulative_Error / N

[0089] End Loop

[0090] The loop N = SMALLEST_LEN to LARGEST_LEN divided by 2

[0091] Cumulative_Error = 0

[0092] Loop i = 1 to N

[0093] Cumulative_Error=Cumulative_Error+abs(link_stretches[i]-link_stretches[i+N])

[0094] End the loop

[0095] Average_Error=Cumulative_Error / N

[0096] End the loop

[0097] Therefore, at 610, if method 500 determines that the incrementing frame does not include the starting link, it returns to 604 and loops until it determines that it has returned to the starting link.

[0098] At 612, the chain type can be identified and stored along with the chain link information, at least based on the link sequence. In some embodiments, a sequence of as few as six links can be used to identify the chain and the starting link. At 614, the chain can be fingerprinted, meaning that a unique link / link type is organized in its sequence for that particular chain, possibly including stretching coefficients and fitting coefficients.

[0099] Not only do the link sequences form a unique fingerprint within the chain, but in various embodiments, the method uses the average stretch of different chain segments to form a unique fingerprint. Thus, for example, if the chain pattern repeats itself every 50 links (e.g., FW latch, spacer, FW latch, spacer, BW latch, spacer, then 22 pairs of top plates / spacers), embodiments additionally calculate the average stretch of the repeating segments of the repeating length and store it along with the chain's fingerprint information. Embodiments define this as a chain segment fingerprint to distinguish it from the link fingerprints described above.

[0100] At 618, chain fingerprints can be stored, providing a useful historical reference as the chain continues to loop, since method 500 and system 402 can now locate links in the chain and compare each currently sensed link with its previous data. Based on the comparison with previous data, method 500 and system 402 can identify at 618 that a link has been removed, and at 620 identify links that have been stretched or become defective. Furthermore, based on the fingerprint identification, method 500 and system 402 can predict the current position of the chain in the loop and when a particular link will return to the video feed, i.e., the amount of time between returning to the camera. This predictive capability greatly improves the efficiency of finding and repairing defective links among available solutions.

[0101] As described above, method 500 and system 402 can identify at 618 that a link has been removed. Once the system identifies the chain, it can identify when the camera missed a link and further identify which link was missed. This is achieved in three ways: #1 - when there are physically non-adjacent links, such as a top plate followed by another top plate instead of a gap link; #2 - when a pattern length error is found, i.e., there are 48 links instead of 50 between pattern repetitions; or #3 - by inserting a "wildcard" link into the sample sequence of the chain, which effectively shifts the sequence of samples that follow it, thereby repairing the fingerprint match.

[0102] Inserting a wildcard link means inserting a wildcard link that was ignored when calculating the average error. The wildcard link is initially inserted at the first suspected location of the missing link, then at the second and third suspected locations, and so on. After each insertion, the overall fit of the fingerprint is compared to the target of the chain, and the fingerprint comparison error is calculated. When the fingerprint comparison error suddenly decreases, the system 402 detects that the wildcard link has been inserted in the correct location; that is, a decrease in the fingerprint comparison error indicates that the correct location of the missing link has been found / identified. More simply, the system 402 continues to loop the chain after finding the initial link; loops the wildcard link at multiple suspected locations of the missing link; calculates the fingerprint comparison error at each of the multiple suspected locations; compares each fingerprint comparison error with others; and detects the correct location of the missing link when the fingerprint comparison error is less than the remaining fingerprint comparison errors by more than a threshold.

[0103] This concept can be further extended; for example, there may be more than one missing segment. The number of wildcard segments may be greater than 1, indicating that the segment sequence is missing. Furthermore, it is possible that N discontinuous segments are missing. In this case, the system can repair a fingerprint with more than one missing segment sequence by traversing a set of samples with more than one wildcard segment, while still being able to identify the location of the sampled segment by comparing it with historical fingerprints.

[0104] Furthermore, as those skilled in the art will understand, the frame skyline recognition provided by these embodiments offers a flexible health monitoring method, because in practice, chains may cycle at different speeds and suddenly start and stop, etc. Therefore, in large-scale manufacturing environments, a readily available chain health monitoring system that relies solely on time calibration is highly disadvantageous and cumbersome.

[0105] Therefore, a system and method have been described for monitoring the chain health of a chain with multiple links in a closed loop during chain circulation.

[0106] While at least one exemplary embodiment has been described in the foregoing detailed description, it should be understood that numerous variations exist. It should also be understood that the one or more exemplary embodiments are merely examples and are not intended to limit the scope, applicability, or configuration of this disclosure in any way. Rather, the foregoing detailed description will provide those skilled in the art with a convenient roadmap for implementing one or more exemplary embodiments. Various changes may be made to the function and arrangement of the elements without departing from the scope of this disclosure as set forth in the appended claims and their legal equivalents.

Claims

1. A method for monitoring the chain health of a chain having multiple links in a closed loop during chain circulation, comprising: Frames are received from a camera, which is statically positioned relative to the chain as the chain loops, and the camera is configured to generate the frames to include links from the plurality of chain links; The frame is converted into a black and white frame with high contrast, where high contrast is greater than a threshold contrast. The link type of the link in the black and white frame is identified by referencing multiple predefined skyline templates; The chain segment having the chain segment type and the corresponding time stamp is stored in the cache; Locate a long line in the black and white frame, where the long line is longer than the threshold length; When it is determined that the black and white frame needs to be straightened based on the corresponding slope of the long line, the black and white frame is straightened. The frame skyline is isolated in the black and white frame; The identification of the chain segment type also includes comparing the frame skyline with the plurality of predefined skyline templates; For each of the plurality of predefined skyline templates The fitting coefficients are determined by making the predefined skyline template walk across the frame skyline; The stretching factor is determined by stretching the predefined skyline template across the frame skyline. The type of link is further identified based on the fitting coefficient and the stretching coefficient; as well as The link having the fitting coefficient and the stretching coefficient is further stored in the cache.

2. The method according to claim 1, further comprising: When the black and white frame indicates that the link is missing a component or has a surface area defect based on the link type, the link is determined to be defective; as well as An alert is generated indicating that the link is defective.

3. The method according to claim 1, further comprising: After storing the linked section, the linked section is represented as the starting linked section, and (a) Receive at least one next frame from the camera, the at least one next frame including the next link of the plurality of links; (b) Convert the at least one next frame into a next black and white frame with high contrast; (c) Identify the link type of the next link in the next black-and-white frame; as well as (d) Store the next link having the link type and the corresponding timestamp in the cache; (e) Constructing a chain segment comprising a sequence of chain type of the next chain segment and chain type of at least one previous chain segment; and Repeat (a)-(d) until it is found that the chain has cycled back to the starting link based on the sequence in the chain segment.

4. The method of claim 3, further comprising creating a fingerprint of the chain in a unique time series, the fingerprint including all of the plurality of links of the chain.

5. The method according to claim 4, further comprising: The missing or removed links of the chain are identified based on the chain's fingerprint, and an alarm is generated accordingly. or The chain is identified by fingerprinting, indicating that there are defective links in the chain, and an alarm is generated accordingly.

6. A system for monitoring the chain health of a chain having multiple links in a closed loop during chain circulation, comprising: A camera, which is statically positioned relative to the chain as the chain loops, is configured to generate frames to include links from the plurality of chain links; as well as A controller module, operatively coupled to the camera and programmed to: Receive the frame; The frame is converted into a black and white frame with high contrast, where high contrast is greater than a threshold contrast. The link type of the link in the black and white frame is identified by referencing multiple predefined skyline templates; The chain segment having the chain segment type and the corresponding time stamp is stored in the cache; Locate a long line in the black and white frame, where the long line is longer than the threshold length; When it is determined that the black and white frame needs to be straightened based on the corresponding slope of the long line, the black and white frame is straightened. The frame skyline is isolated in the black and white frame; The chain segment type is further identified by comparing the frame skyline with the plurality of predefined skyline templates; For each of the plurality of predefined skyline templates The fitting coefficients are determined by making the predefined skyline template walk across the frame skyline; The stretching factor is determined by stretching the predefined skyline template across the frame skyline. The link type is further identified based on the fitting coefficient and the stretching coefficient; as well as The link having the fitting coefficient and the stretching coefficient is further stored in the cache.

Images