DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 04/05/2024 and 06/03/2025 and submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Priority
Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
Claims 17-20 are interpreted to invoke 35 U.S.C. 112(f).
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation(s) is/are: “a detection module” in claim 17.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof. Original specifications paragraphs [0052-0053] discloses computer hardware processors or their equivalents thereof to execute the functions executed by the detection module.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Cho et al. (KR 20190051624 A, as provided) in view of Burkhardt et al. (US Pub No. 20210142467 A1).
Regarding Claim 1,
Cho discloses A method of detecting abnormal transport of a defective electrode plate, which is performed by at least one processor, the method comprising: receiving a plurality of images associated with transport of a defective electrode plate from one or more cameras installed on a path of a secondary battery assembly process; (Cho, Abstract, means to solve the problem, discloses present invention relates to an electrode plate conveying device of a secondary battery manufacturing apparatus. The device comprises: a robot arm configured to pick up an electrode plate from a magazine loaded with the electrode plate and convey the electrode plate to an electrode plate lamination unit; an electrode plate inspection unit provided on a path for conveying the electrode plate and configured to determine whether the electrode plate is defective and has an alignment error; and a control unit for determining the conveying path of the electrode plate according to whether the electrode plate is defective and has the alignment error. The electrode conveying device of a secondary battery manufacturing apparatus having an electrode inspection function according to the present invention is capable of determining the defects and alignments of the electrode plate while conveying the electrode plate in the stacking of the electrode plates, and destroying or correcting the defects and errors, thereby reducing the defect rate; the image in which the plate inspection unit takes a picture of the plate is analyzed and fault acceptance and rejection can be judged. Moreover, the robot arm can be controlled so that the control unit passes the check position in which the plate inspection unit can obtain the image and the plate is transferred. In the meantime, the plate inspection unit can obtain the image of the plate at the moment when the plate passes through the check position; image of conveying path that conveys defective electrode plates are obtained installed on the path with camera sensors) and
detecting abnormal transport of the defective electrode plate on the path based on the images (Cho, Abstract, discloses present invention relates to an electrode plate conveying device of a secondary battery manufacturing apparatus. The device comprises: a robot arm configured to pick up an electrode plate from a magazine loaded with the electrode plate and convey the electrode plate to an electrode plate lamination unit; an electrode plate inspection unit provided on a path for conveying the electrode plate and configured to determine whether the electrode plate is defective and has an alignment error; and a control unit for determining the conveying path of the electrode plate according to whether the electrode plate is defective and has the alignment error. The electrode conveying device of a secondary battery manufacturing apparatus having an electrode inspection function according to the present invention is capable of determining the defects and alignments of the electrode plate while conveying the electrode plate in the stacking of the electrode plates, and destroying or correcting the defects and errors, thereby reducing the defect rate; alignment error (abnormal transport) of the defective electrode plate is determined on the conveyer path).
Cho does not explicitly disclose using a machine learning model based on unsupervised learning.
Burkhardt discloses using a machine learning model based on unsupervised learning. (Burkhardt, [0008-0009], discloses computer system is provided for generating training datasets to train a model that is used to analyze a manufacturing process. The computer system includes non-transitory memory configured to store at least first and second base images. The computer system also includes a processing system that includes at least one hardware processor. The processing system is configured or otherwise programmed to generate a plurality of training images that are each generated by combining the first and second base images together, wherein each of the plurality of images is generated by randomly varying a location of at least the first base image with respect to the second base image. The processing system is configured or otherwise programmed to train a model based on the plurality of training images and corresponding labels associated with each of the plurality of training images. The processing system is configured or otherwise programmed to receive a plurality of real images of the manufacturing process. The processing system is configured or otherwise programmed to generate, for each of the plurality of real images, a prediction by using the model to predict whether a defect is contained within a corresponding one of the plurality of real images; a non-transitory computer readable storage medium comprising an application program for use with a computer system that generates training datasets for neural networks, the application program comprising instructions that cause the computer system operate in the following manner. Loading a weld seam image and second image of a plate that is to be welded. Generating a preparation image that includes at least two instances of the plate separated by a gap. Generating a training image by sequentially applying a plurality of instances of the weld seam image over the gap, wherein each successive one of the plurality of instances overlaps a prior instance. The location at which each of the plurality of instances is applied over the gap may be varied based on a randomness factor. The generation of training images is repeated until a plurality of training images have been generated. This may be thousands or even millions of different images. The plurality of training images may be varied due to application of the randomness factor that is used in the creation of each one of the plurality of training images. Thus, the plurality of training images may be diverse and robust in nature and include many different examples of defective welds between the plates, including examples that would normally be quite rare; combination of images are used as normal and not normal are used to train the machine learning classifier to be compared with path of transport to determine abnormal transport of objects).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to combine the teachings of Cho in view of Burkhardt having a method of installed cameras on transfer path of electrode plates stacked from magazine to stacking place and simultaneously detecting any defects in the electrode plated or alignment errors in the transfer path of the electrode plates, with the teachings of Burkhardt having, by the training module, training a machine learning classifier to be trained on defective and nondefective normal and operation of manufacturing and packaging video images in order to accurately identify operation of transport path and detecting any defect in objects to be inspected in applications including manufacturing.
Regarding Claim 2,
The combination of Cho and Burkhardt further discloses wherein the one or more cameras are installed at a position adjacent wherein the one or more cameras are installed at a position adjacent to at least one of a no good (NG) box and an out belt on the path (Cho, description, discloses sensor unit functions as the electrode plate inspection unit 300. The electrode plate inspecting unit 300 performs sensing for determining whether or not the electrode plate 1000 is defective and for measuring the alignment error in a state where the robot arm 200 picks up the electrode plate 1000. The sensor unit may include an image sensor and an infrared sensor. The image sensor is configured to capture one side of the electrode plate 1000 to acquire an image. The infrared sensor senses an edge portion of the electrode plate 1000 and provides data to determine whether the electrode plate 1000 is aligned. The robot arm 200 is attracted to the upper surface of the electrode plate 1000 using the attachment 201 and is transported in a state that the lower surface is opened. Further, although not shown, the apparatus may further include an illumination unit (not shown) so as to provide an appropriate illumination for photographing. The illuminating unit may be located on the same side as the camera or on the opposite side of the camera with respect to the plate to be inspected. However, the sensor unit may be configured to include only the camera so as to simultaneously perform defect and alignment errors using the acquired image; The control unit 600 is configured to determine the presence or absence of defects and misalignment of the substrate picked up by the robot arm 200 and to control the operation of the robot arm 200 according to the determination result. The control unit 600 controls the operation of the robot arm 200 in accordance with a first transferring step of picking up the electrode plate 1000 from the magazine 110 and moving the sensor unit to the inspection position I where the sensor unit is located, It is possible to control them separately in steps. When the alignment error of the electrode plate 1000 occurs in the second transfer step, the robot arm 200 is controlled to reflect the misalignment of the electrode plate 1000 in the transfer operation of the robot arm 200. The function of the controller 600 will be described later in detail. However, since the configuration of the control unit 600 can be variously configured, a description of the configuration is omitted; alarm unit 700 is configured to notify the user when the structural alignment between the units is disturbed when the electrode plate 1000 is stacked. The alarm unit 700 is configured to generate a warning including a time and an auditory sense to the user. However, although not shown, the function of the alarm unit 700 may be performed by providing visual information such as a pop-up through a display on an integrated control system without any other configuration; magazine transfer unit 100 is configured to transfer the magazine 110 to a position where the robot arm 200 to be described later can pick up the magazine. Here, the magazine 110 refers to an instrument such as a tray capable of stacking a plurality of electrode plates 1000 cut in a uniform size from an electrode plate cutting unit (not shown). The magazine transfer unit 100 may include a first magazine transfer unit 100 on which the positive electrode plate 1000 is mounted and a second magazine transfer unit 100 on which the negative electrode plate 1000 is mounted. Each magazine transfer unit 100 may be configured to include a transfer line through which the magazine 110 can move, and each magazine 110 may be loaded with dozens to hundreds of cut polar plates 1000. One end of the magazine transfer unit 100 is configured to load the cut electrode plate 1000 from the electrode plate cutting unit and the other end is connected to the magazine transfer unit 100 through a magazine 110); Each of the magazines 110 may be formed with an opening on the lower surface thereof so as to determine whether the electrode plate 1000 is exhausted from a sensor unit to be described later; cameras are installed on the transport path at various orientations to determine defect and alignment errors). (Burkhardt, [0033], Fig. 1, Fig.3, Fig. 9, Fig.11, discloses in shipping a decorative box of a product, an inspection as to whether a part of a target object forms a plane is performed using information obtained from an image. The description will be given of an information processing apparatus used in an inspection as to whether a lid of an outer casing being a target object is slanting. In the inspection, if the target object does not satisfy a condition (the lid is not closed or is slanting), the information processing apparatus presents, to a person or a robot, an instruction for correcting the orientation of the lid. For the sake of simplicity, the lid described here is assumed to have a smooth and flat top surface (top surface in a case where the product is placed in a correct up-down direction, and a surface squarely facing an imaging apparatus); object is dropped in incorrect box; object position is checked for if it positioned correctly to be placed in right or wrong box). to at least one of a no good (NG) box and an out belt on the path). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 3,
The combination of Cho and Burkhardt further discloses wherein the detecting of abnormal transport of the defective electrode plate on the path includes: detecting at least one of stacking of the defective electrode plate in the no good box, abnormal drop of the defective electrode plate in the no good box, an abnormal position of the defective electrode plate in the no good box and an abnormal position in the out belt, based on the plurality of images. (Cho, description, discloses sensor unit functions as the electrode plate inspection unit 300. The electrode plate inspecting unit 300 performs sensing for determining whether or not the electrode plate 1000 is defective and for measuring the alignment error in a state where the robot arm 200 picks up the electrode plate 1000. The sensor unit may include an image sensor and an infrared sensor. The image sensor is configured to capture one side of the electrode plate 1000 to acquire an image. The infrared sensor senses an edge portion of the electrode plate 1000 and provides data to determine whether the electrode plate 1000 is aligned. The robot arm 200 is attracted to the upper surface of the electrode plate 1000 using the attachment 201 and is transported in a state that the lower surface is opened. Further, although not shown, the apparatus may further include an illumination unit (not shown) so as to provide an appropriate illumination for photographing. The illuminating unit may be located on the same side as the camera or on the opposite side of the camera with respect to the plate to be inspected. However, the sensor unit may be configured to include only the camera so as to simultaneously perform defect and alignment errors using the acquired image; The control unit 600 is configured to determine the presence or absence of defects and misalignment of the substrate picked up by the robot arm 200 and to control the operation of the robot arm 200 according to the determination result. The control unit 600 controls the operation of the robot arm 200 in accordance with a first transferring step of picking up the electrode plate 1000 from the magazine 110 and moving the sensor unit to the inspection position I where the sensor unit is located, It is possible to control them separately in steps. When the alignment error of the electrode plate 1000 occurs in the second transfer step, the robot arm 200 is controlled to reflect the misalignment of the electrode plate 1000 in the transfer operation of the robot arm 200. The function of the controller 600 will be described later in detail. However, since the configuration of the control unit 600 can be variously configured, a description of the configuration is omitted; alarm unit 700 is configured to notify the user when the structural alignment between the units is disturbed when the electrode plate 1000 is stacked. The alarm unit 700 is configured to generate a warning including a time and an auditory sense to the user. However, although not shown, the function of the alarm unit 700 may be performed by providing visual information such as a pop-up through a display on an integrated control system without any other configuration; magazine transfer unit 100 is configured to transfer the magazine 110 to a position where the robot arm 200 to be described later can pick up the magazine. Here, the magazine 110 refers to an instrument such as a tray capable of stacking a plurality of electrode plates 1000 cut in a uniform size from an electrode plate cutting unit (not shown). The magazine transfer unit 100 may include a first magazine transfer unit 100 on which the positive electrode plate 1000 is mounted and a second magazine transfer unit 100 on which the negative electrode plate 1000 is mounted. Each magazine transfer unit 100 may be configured to include a transfer line through which the magazine 110 can move, and each magazine 110 may be loaded with dozens to hundreds of cut polar plates 1000. One end of the magazine transfer unit 100 is configured to load the cut electrode plate 1000 from the electrode plate cutting unit and the other end is connected to the magazine transfer unit 100 through a magazine 110); Each of the magazines 110 may be formed with an opening on the lower surface thereof so as to determine whether the electrode plate 1000 is exhausted from a sensor unit to be described later; cameras are installed on the transport path at various orientations to determine defect and alignment errors). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 4,
The combination of Cho and Burkhardt further discloses wherein the one or more cameras are connected to a power over Ethernet (POE) hub, and the POE hub is connected to at least one of an equipment control device associated with the secondary battery assembly process, (Cho, Description, Fig. 1, discloses concept of the first the electrode plate transfer apparatus of the secondary battery manufacturing apparatus including the electrode plate inspection function according to the present invention includes a robot arm 200, a sensor unit, a control unit 600, and an alarm unit 700; sensor unit functions as the electrode plate inspection unit 300. The electrode plate inspecting unit 300 performs sensing for determining whether or not the electrode plate 1000 is defective and for measuring the alignment error in a state where the robot arm 200 picks up the electrode plate 1000. The sensor unit may include an image sensor and an infrared sensor. The image sensor is configured to capture one side of the electrode plate 1000 to acquire an image. The infrared sensor senses an edge portion of the electrode plate 1000 and provides data to determine whether the electrode plate 1000 is aligned. The robot arm 200 is attracted to the upper surface of the electrode plate 1000 using the attachment 201 and is transported in a state that the lower surface is opened. Further, although not shown, the apparatus may further include an illumination unit (not shown) so as to provide an appropriate illumination for photographing. The illuminating unit may be located on the same side as the camera or on the opposite side of the camera with respect to the plate to be inspected. However, the sensor unit may be configured to include only the camera so as to simultaneously perform defect and alignment errors using the acquired image; cameras on transport path are installed in communication with control unit (POE) hub). a video recording device and a computing device associated with the machine learning model. (Burkhardt, [0096], discloses the approach of image generation discussed herein may use randomness to generate a balanced set of images with and without defects. Such an approach advantageously may provide a training dataset for machine learning purposes that performs better other techniques of image generation; machine learning model processes video recorded images). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 5,
The combination of Cho and Burkhardt further discloses wherein the machine learning model is trained by an unsupervised learning method using a learning image frame set associated with normal transport of the defective electrode plate captured on the path of the secondary battery assembly process. (Burkhardt, [0008-0009], discloses computer system is provided for generating training datasets to train a model that is used to analyze a manufacturing process. The computer system includes non-transitory memory configured to store at least first and second base images. The computer system also includes a processing system that includes at least one hardware processor. The processing system is configured or otherwise programmed to generate a plurality of training images that are each generated by combining the first and second base images together, wherein each of the plurality of images is generated by randomly varying a location of at least the first base image with respect to the second base image. The processing system is configured or otherwise programmed to train a model based on the plurality of training images and corresponding labels associated with each of the plurality of training images. The processing system is configured or otherwise programmed to receive a plurality of real images of the manufacturing process. The processing system is configured or otherwise programmed to generate, for each of the plurality of real images, a prediction by using the model to predict whether a defect is contained within a corresponding one of the plurality of real images; a non-transitory computer readable storage medium comprising an application program for use with a computer system that generates training datasets for neural networks, the application program comprising instructions that cause the computer system operate in the following manner. Loading a weld seam image and second image of a plate that is to be welded. Generating a preparation image that includes at least two instances of the plate separated by a gap. Generating a training image by sequentially applying a plurality of instances of the weld seam image over the gap, wherein each successive one of the plurality of instances overlaps a prior instance. The location at which each of the plurality of instances is applied over the gap may be varied based on a randomness factor. The generation of training images is repeated until a plurality of training images have been generated. This may be thousands or even millions of different images. The plurality of training images may be varied due to application of the randomness factor that is used in the creation of each one of the plurality of training images. Thus, the plurality of training images may be diverse and robust in nature and include many different examples of defective welds between the plates, including examples that would normally be quite rare; combination of images are used as normal and not normal are used to train the machine learning classifier to be compared with path of transport to determine abnormal transport of objects). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 6,
The combination of Cho and Burkhardt further discloses wherein the machine learning model is trained to receive an image and to output whether the image is included in a group generated based on the learning image frame set. (Burkhardt, [0008-0009], discloses computer system is provided for generating training datasets to train a model that is used to analyze a manufacturing process. The computer system includes non-transitory memory configured to store at least first and second base images. The computer system also includes a processing system that includes at least one hardware processor. The processing system is configured or otherwise programmed to generate a plurality of training images that are each generated by combining the first and second base images together, wherein each of the plurality of images is generated by randomly varying a location of at least the first base image with respect to the second base image. The processing system is configured or otherwise programmed to train a model based on the plurality of training images and corresponding labels associated with each of the plurality of training images. The processing system is configured or otherwise programmed to receive a plurality of real images of the manufacturing process. The processing system is configured or otherwise programmed to generate, for each of the plurality of real images, a prediction by using the model to predict whether a defect is contained within a corresponding one of the plurality of real images; a non-transitory computer readable storage medium comprising an application program for use with a computer system that generates training datasets for neural networks, the application program comprising instructions that cause the computer system operate in the following manner. Loading a weld seam image and second image of a plate that is to be welded. Generating a preparation image that includes at least two instances of the plate separated by a gap. Generating a training image by sequentially applying a plurality of instances of the weld seam image over the gap, wherein each successive one of the plurality of instances overlaps a prior instance. The location at which each of the plurality of instances is applied over the gap may be varied based on a randomness factor. The generation of training images is repeated until a plurality of training images have been generated. This may be thousands or even millions of different images. The plurality of training images may be varied due to application of the randomness factor that is used in the creation of each one of the plurality of training images. Thus, the plurality of training images may be diverse and robust in nature and include many different examples of defective welds between the plates, including examples that would normally be quite rare; combination of images are used as normal and not normal are used to train the machine learning classifier to be compared with path of transport to determine abnormal transport of objects). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 7,
The combination of Cho and Burkhardt further discloses wherein the detecting of the abnormal transport of the defective electrode plate on the path includes: determining abnormal transport of the defective electrode plate for each of the images using the machine learning model; and determining that the defective electrode plate is abnormally transported if a number of images determined to be abnormal transport (Cho, Description, Fig.8, discloses an alignment error according to the cumulative number. It should be noted that the graph shown is extreme for the sake of explanation and does not exactly correspond to the position error occurring in the actual plate 1000 manufacturing process; number of errors are counted) of the defective electrode plate among the images exceeds a predetermined threshold. (Burkhardt, [0008-0009], discloses computer system is provided for generating training datasets to train a model that is used to analyze a manufacturing process. The computer system includes non-transitory memory configured to store at least first and second base images. The computer system also includes a processing system that includes at least one hardware processor. The processing system is configured or otherwise programmed to generate a plurality of training images that are each generated by combining the first and second base images together, wherein each of the plurality of images is generated by randomly varying a location of at least the first base image with respect to the second base image. The processing system is configured or otherwise programmed to train a model based on the plurality of training images and corresponding labels associated with each of the plurality of training images. The processing system is configured or otherwise programmed to receive a plurality of real images of the manufacturing process. The processing system is configured or otherwise programmed to generate, for each of the plurality of real images, a prediction by using the model to predict whether a defect is contained within a corresponding one of the plurality of real images; a non-transitory computer readable storage medium comprising an application program for use with a computer system that generates training datasets for neural networks, the application program comprising instructions that cause the computer system operate in the following manner. Loading a weld seam image and second image of a plate that is to be welded. Generating a preparation image that includes at least two instances of the plate separated by a gap. Generating a training image by sequentially applying a plurality of instances of the weld seam image over the gap, wherein each successive one of the plurality of instances overlaps a prior instance. The location at which each of the plurality of instances is applied over the gap may be varied based on a randomness factor. The generation of training images is repeated until a plurality of training images have been generated. This may be thousands or even millions of different images. The plurality of training images may be varied due to application of the randomness factor that is used in the creation of each one of the plurality of training images. Thus, the plurality of training images may be diverse and robust in nature and include many different examples of defective welds between the plates, including examples that would normally be quite rare; combination of images are used as normal and not normal are used to train the machine learning classifier to be compared with path of transport to determine abnormal transport of objects). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 8,
The combination of Cho and Burkhardt further discloses counting a number of normally discharged defective electrode plates based on the images. (Burkhardt, [0050-0052] , discloses Large input data requirements are further complicated in connection with certain real world situations that occur infrequently. For example, fraud detection and anomaly detection tend to have real world datasets that are highly imbalanced as the nubmer of anomalies within a given dataset tend to be vastly outnumbered by “normal” samples within the dataset. This imbalance tends to hold for welding defect detection as it deals with detecting defects, which are (hopefully) rare incidents. Thus, if real world images were to be used exclusively to train models then the defective weld samples would tend to be a fraction of good samples (e.g., those without defects). Such an imbalance could then fail to properly train the model for detecting accurate defect representations; potential issue is also that even in a case where there is enough data, annotation of large number of images can still be a challenging problem. Specifically, with problems like detecting welding defects, the deep learning models may require the location of the defects within each image. Manually annotating a large set of images requires lot of time and effort and is also prone to errors; the problem of acquiring a large enough dataset can be a problem with Generative Adversarial Networks (GANs) as well. This is because such implementations tend to require large datasets in order to learn the underlying data distribution of source dataset and generated images. For images to be near-realistic GANs would have to be trained on very large datasets; machine learning classifier is trained to determine defective images and total number of defective images output by the classifier is counted as normal or abnormal). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 9,
The combination of Cho and Burkhardt further discloses wherein the counting of the number of normally discharged defective electrode plates includes counting the number of normally discharged defective electrode plates (Cho, background, discloses Generally, a secondary battery is a battery which can be repeatedly used through a discharge process of converting chemical energy into electrical energy and a charging process in the reverse direction. Examples of the secondary battery include a nickel-cadmium (Ni-Cd) battery, a nickel- A lithium-metal battery, a lithium-ion (Ni-Ion) battery, and a lithium-ion polymer battery. The secondary battery is composed of an anode, a cathode, an electrolyte, and a separator, and stores and generates electricity using voltage difference between different anode and cathode materials. Here, the discharge is to move electrons from a cathode having a high voltage to a cathode having a low voltage (generating electricity as much as the voltage difference of the anode), and charging means transferring the electrons again from the anode to the cathode where the anode material receives electrons and lithium ions And returned to the original metal oxide. That is, when the secondary battery is charged, the charge current flows as the metal atoms move from the anode to the cathode through the separator, and when discharged, the metal atoms move from the cathode to the anode and the discharge current flows; the rechargeable battery can be divided into a winding type and a lamination type. In the lamination type, a positive electrode plate and a negative electrode plate, which are cut to a predetermined size, are alternately laminated to fabricate an electrode assembly. However, there is a problem that the defective electrode plate is not recognized and the electrode plate is stacked like the normal electrode plate, or the alignment of the electrode plate is disturbed during the stacking process of the electrode plate, resulting in defects; number of pixels present in discharge image are determined in secondary battery) based on a number of image pixels corresponding to a defective electrode plate in regions of interest of the images. (Burkhardt, [0050-0052], discloses Large input data requirements are further complicated in connection with certain real world situations that occur infrequently. For example, fraud detection and anomaly detection tend to have real world datasets that are highly imbalanced as the nubmer of anomalies within a given dataset tend to be vastly outnumbered by “normal” samples within the dataset. This imbalance tends to hold for welding defect detection as it deals with detecting defects, which are (hopefully) rare incidents. Thus, if real world images were to be used exclusively to train models then the defective weld samples would tend to be a fraction of good samples (e.g., those without defects). Such an imbalance could then fail to properly train the model for detecting accurate defect representations; potential issue is also that even in a case where there is enough data, annotation of large number of images can still be a challenging problem. Specifically, with problems like detecting welding defects, the deep learning models may require the location of the defects within each image. Manually annotating a large set of images requires lot of time and effort and is also prone to errors; the problem of acquiring a large enough dataset can be a problem with Generative Adversarial Networks (GANs) as well. This is because such implementations tend to require large datasets in order to learn the underlying data distribution of source dataset and generated images. For images to be near-realistic GANs would have to be trained on very large datasets; machine learning classifier is trained to determine defective images and total number of defective images output by the classifier is counted as normal or abnormal). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 10,
The combination of Cho and Burkhardt further discloses generating statistical data associated with transport of the defective electrode plate based on the number of normally discharged defective electrode plates. (Cho, background, discloses Generally, a secondary battery is a battery which can be repeatedly used through a discharge process of converting chemical energy into electrical energy and a charging process in the reverse direction. Examples of the secondary battery include a nickel-cadmium (Ni-Cd) battery, a nickel- A lithium-metal battery, a lithium-ion (Ni-Ion) battery, and a lithium-ion polymer battery. The secondary battery is composed of an anode, a cathode, an electrolyte, and a separator, and stores and generates electricity using voltage difference between different anode and cathode materials. Here, the discharge is to move electrons from a cathode having a high voltage to a cathode having a low voltage (generating electricity as much as the voltage difference of the anode), and charging means transferring the electrons again from the anode to the cathode where the anode material receives electrons and lithium ions And returned to the original metal oxide. That is, when the secondary battery is charged, the charge current flows as the metal atoms move from the anode to the cathode through the separator, and when discharged, the metal atoms move from the cathode to the anode and the discharge current flows; the rechargeable battery can be divided into a winding type and a lamination type. In the lamination type, a positive electrode plate and a negative electrode plate, which are cut to a predetermined size, are alternately laminated to fabricate an electrode assembly. However, there is a problem that the defective electrode plate is not recognized and the electrode plate is stacked like the normal electrode plate, or the alignment of the electrode plate is disturbed during the stacking process of the electrode plate, resulting in defects; number of pixels present in discharge image are determined in secondary battery). (Burkhardt, [0050-0052], discloses Large input data requirements are further complicated in connection with certain real world situations that occur infrequently. For example, fraud detection and anomaly detection tend to have real world datasets that are highly imbalanced as the nubmer of anomalies within a given dataset tend to be vastly outnumbered by “normal” samples within the dataset. This imbalance tends to hold for welding defect detection as it deals with detecting defects, which are (hopefully) rare incidents. Thus, if real world images were to be used exclusively to train models then the defective weld samples would tend to be a fraction of good samples (e.g., those without defects). Such an imbalance could then fail to properly train the model for detecting accurate defect representations; potential issue is also that even in a case where there is enough data, annotation of large number of images can still be a challenging problem. Specifically, with problems like detecting welding defects, the deep learning models may require the location of the defects within each image. Manually annotating a large set of images requires lot of time and effort and is also prone to errors; the problem of acquiring a large enough dataset can be a problem with Generative Adversarial Networks (GANs) as well. This is because such implementations tend to require large datasets in order to learn the underlying data distribution of source dataset and generated images. For images to be near-realistic GANs would have to be trained on very large datasets; machine learning classifier is trained to determine defective images and total number of defective images output by the classifier is counted as normal or abnormal). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 11,
The combination of Cho and Burkhardt further discloses wherein the statistical data associated with the transport of the defective electrode plate includes data obtained by comparing a number of defective electrode plates determined by a vision inspector with the number of normally discharged defective electrode plates. Cho, background, discloses Generally, a secondary battery is a battery which can be repeatedly used through a discharge process of converting chemical energy into electrical energy and a charging process in the reverse direction. Examples of the secondary battery include a nickel-cadmium (Ni-Cd) battery, a nickel- A lithium-metal battery, a lithium-ion (Ni-Ion) battery, and a lithium-ion polymer battery. The secondary battery is composed of an anode, a cathode, an electrolyte, and a separator, and stores and generates electricity using voltage difference between different anode and cathode materials. Here, the discharge is to move electrons from a cathode having a high voltage to a cathode having a low voltage (generating electricity as much as the voltage difference of the anode), and charging means transferring the electrons again from the anode to the cathode where the anode material receives electrons and lithium ions And returned to the original metal oxide. That is, when the secondary battery is charged, the charge current flows as the metal atoms move from the anode to the cathode through the separator, and when discharged, the metal atoms move from the cathode to the anode and the discharge current flows; the rechargeable battery can be divided into a winding type and a lamination type. In the lamination type, a positive electrode plate and a negative electrode plate, which are cut to a predetermined size, are alternately laminated to fabricate an electrode assembly. However, there is a problem that the defective electrode plate is not recognized and the electrode plate is stacked like the normal electrode plate, or the alignment of the electrode plate is disturbed during the stacking process of the electrode plate, resulting in defects; number of pixels present in discharge image are determined in secondary battery). (Burkhardt, [0050-0052], discloses Large input data requirements are further complicated in connection with certain real world situations that occur infrequently. For example, fraud detection and anomaly detection tend to have real world datasets that are highly imbalanced as the nubmer of anomalies within a given dataset tend to be vastly outnumbered by “normal” samples within the dataset. This imbalance tends to hold for welding defect detection as it deals with detecting defects, which are (hopefully) rare incidents. Thus, if real world images were to be used exclusively to train models then the defective weld samples would tend to be a fraction of good samples (e.g., those without defects). Such an imbalance could then fail to properly train the model for detecting accurate defect representations; potential issue is also that even in a case where there is enough data, annotation of large number of images can still be a challenging problem. Specifically, with problems like detecting welding defects, the deep learning models may require the location of the defects within each image. Manually annotating a large set of images requires lot of time and effort and is also prone to errors; the problem of acquiring a large enough dataset can be a problem with Generative Adversarial Networks (GANs) as well. This is because such implementations tend to require large datasets in order to learn the underlying data distribution of source dataset and generated images. For images to be near-realistic GANs would have to be trained on very large datasets; machine learning classifier is trained to determine defective images and total number of defective images output by the classifier is counted as normal or abnormal). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 12,
The combination of Cho and Burkhardt further discloses counting a number of defective electrode plates detected to be abnormal transport based on the images. (Cho, background, discloses Generally, a secondary battery is a battery which can be repeatedly used through a discharge process of converting chemical energy into electrical energy and a charging process in the reverse direction. Examples of the secondary battery include a nickel-cadmium (Ni-Cd) battery, a nickel- A lithium-metal battery, a lithium-ion (Ni-Ion) battery, and a lithium-ion polymer battery. The secondary battery is composed of an anode, a cathode, an electrolyte, and a separator, and stores and generates electricity using voltage difference between different anode and cathode materials. Here, the discharge is to move electrons from a cathode having a high voltage to a cathode having a low voltage (generating electricity as much as the voltage difference of the anode), and charging means transferring the electrons again from the anode to the cathode where the anode material receives electrons and lithium ions And returned to the original metal oxide. That is, when the secondary battery is charged, the charge current flows as the metal atoms move from the anode to the cathode through the separator, and when discharged, the metal atoms move from the cathode to the anode and the discharge current flows; the rechargeable battery can be divided into a winding type and a lamination type. In the lamination type, a positive electrode plate and a negative electrode plate, which are cut to a predetermined size, are alternately laminated to fabricate an electrode assembly. However, there is a problem that the defective electrode plate is not recognized and the electrode plate is stacked like the normal electrode plate, or the alignment of the electrode plate is disturbed during the stacking process of the electrode plate, resulting in defects; number of pixels present in discharge image are determined in secondary battery). (Burkhardt, [0050-0052], discloses large input data requirements are further complicated in connection with certain real world situations that occur infrequently. For example, fraud detection and anomaly detection tend to have real world datasets that are highly imbalanced as the nubmer of anomalies within a given dataset tend to be vastly outnumbered by “normal” samples within the dataset. This imbalance tends to hold for welding defect detection as it deals with detecting defects, which are (hopefully) rare incidents. Thus, if real world images were to be used exclusively to train models then the defective weld samples would tend to be a fraction of good samples (e.g., those without defects). Such an imbalance could then fail to properly train the model for detecting accurate defect representations; potential issue is also that even in a case where there is enough data, annotation of large number of images can still be a challenging problem. Specifically, with problems like detecting welding defects, the deep learning models may require the location of the defects within each image. Manually annotating a large set of images requires lot of time and effort and is also prone to errors; the problem of acquiring a large enough dataset can be a problem with Generative Adversarial Networks (GANs) as well. This is because such implementations tend to require large datasets in order to learn the underlying data distribution of source dataset and generated images. For images to be near-realistic GANs would have to be trained on very large datasets; machine learning classifier is trained to determine defective images and total number of defective images output by the classifier is counted as normal or abnormal). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 13,
The combination of Cho and Burkhardt further discloses wherein the counting of the number of defective electrode plates detected to be abnormal transport includes: determining a type of abnormal transport of the defective electrode plate detected to be abnormal transport; and counting the number of defective electrode plates for each type of the determined abnormal transport. (Cho, background, discloses Generally, a secondary battery is a battery which can be repeatedly used through a discharge process of converting chemical energy into electrical energy and a charging process in the reverse direction. Examples of the secondary battery include a nickel-cadmium (Ni-Cd) battery, a nickel- A lithium-metal battery, a lithium-ion (Ni-Ion) battery, and a lithium-ion polymer battery. The secondary battery is composed of an anode, a cathode, an electrolyte, and a separator, and stores and generates electricity using voltage difference between different anode and cathode materials. Here, the discharge is to move electrons from a cathode having a high voltage to a cathode having a low voltage (generating electricity as much as the voltage difference of the anode), and charging means transferring the electrons again from the anode to the cathode where the anode material receives electrons and lithium ions And returned to the original metal oxide. That is, when the secondary battery is charged, the charge current flows as the metal atoms move from the anode to the cathode through the separator, and when discharged, the metal atoms move from the cathode to the anode and the discharge current flows; the rechargeable battery can be divided into a winding type and a lamination type. In the lamination type, a positive electrode plate and a negative electrode plate, which are cut to a predetermined size, are alternately laminated to fabricate an electrode assembly. However, there is a problem that the defective electrode plate is not recognized and the electrode plate is stacked like the normal electrode plate, or the alignment of the electrode plate is disturbed during the stacking process of the electrode plate, resulting in defects; number of pixels present in discharge image are determined in secondary battery). (Burkhardt, [0050-0052], discloses Large input data requirements are further complicated in connection with certain real world situations that occur infrequently. For example, fraud detection and anomaly detection tend to have real world datasets that are highly imbalanced as the nubmer of anomalies within a given dataset tend to be vastly outnumbered by “normal” samples within the dataset. This imbalance tends to hold for welding defect detection as it deals with detecting defects, which are (hopefully) rare incidents. Thus, if real world images were to be used exclusively to train models then the defective weld samples would tend to be a fraction of good samples (e.g., those without defects). Such an imbalance could then fail to properly train the model for detecting accurate defect representations; potential issue is also that even in a case where there is enough data, annotation of large number of images can still be a challenging problem. Specifically, with problems like detecting welding defects, the deep learning models may require the location of the defects within each image. Manually annotating a large set of images requires lot of time and effort and is also prone to errors; the problem of acquiring a large enough dataset can be a problem with Generative Adversarial Networks (GANs) as well. This is because such implementations tend to require large datasets in order to learn the underlying data distribution of source dataset and generated images. For images to be near-realistic GANs would have to be trained on very large datasets; machine learning classifier is trained to determine defective images and total number of defective images output by the classifier is counted as normal or abnormal). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 14,
The combination of Cho and Burkhardt further discloses generating statistical data associated with transport of the defective electrode plates, based on the number of defective electrode plates for each type of abnormal transport. (Cho, background, discloses Generally, a secondary battery is a battery which can be repeatedly used through a discharge process of converting chemical energy into electrical energy and a charging process in the reverse direction. Examples of the secondary battery include a nickel-cadmium (Ni-Cd) battery, a nickel- A lithium-metal battery, a lithium-ion (Ni-Ion) battery, and a lithium-ion polymer battery. The secondary battery is composed of an anode, a cathode, an electrolyte, and a separator, and stores and generates electricity using voltage difference between different anode and cathode materials. Here, the discharge is to move electrons from a cathode having a high voltage to a cathode having a low voltage (generating electricity as much as the voltage difference of the anode), and charging means transferring the electrons again from the anode to the cathode where the anode material receives electrons and lithium ions And returned to the original metal oxide. That is, when the secondary battery is charged, the charge current flows as the metal atoms move from the anode to the cathode through the separator, and when discharged, the metal atoms move from the cathode to the anode and the discharge current flows; the rechargeable battery can be divided into a winding type and a lamination type. In the lamination type, a positive electrode plate and a negative electrode plate, which are cut to a predetermined size, are alternately laminated to fabricate an electrode assembly. However, there is a problem that the defective electrode plate is not recognized and the electrode plate is stacked like the normal electrode plate, or the alignment of the electrode plate is disturbed during the stacking process of the electrode plate, resulting in defects; number of pixels present in discharge image are determined in secondary battery). (Burkhardt, [0050-0052], discloses Large input data requirements are further complicated in connection with certain real world situations that occur infrequently. For example, fraud detection and anomaly detection tend to have real world datasets that are highly imbalanced as the nubmer of anomalies within a given dataset tend to be vastly outnumbered by “normal” samples within the dataset. This imbalance tends to hold for welding defect detection as it deals with detecting defects, which are (hopefully) rare incidents. Thus, if real world images were to be used exclusively to train models then the defective weld samples would tend to be a fraction of good samples (e.g., those without defects). Such an imbalance could then fail to properly train the model for detecting accurate defect representations; potential issue is also that even in a case where there is enough data, annotation of large number of images can still be a challenging problem. Specifically, with problems like detecting welding defects, the deep learning models may require the location of the defects within each image. Manually annotating a large set of images requires lot of time and effort and is also prone to errors; the problem of acquiring a large enough dataset can be a problem with Generative Adversarial Networks (GANs) as well. This is because such implementations tend to require large datasets in order to learn the underlying data distribution of source dataset and generated images. For images to be near-realistic GANs would have to be trained on very large datasets; machine learning classifier is trained to determine defective images and total number of defective images output by the classifier is counted as normal or abnormal). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 15,
The combination of Cho and Burkhardt further discloses wherein the statistical data associated with the transport of the defective electrode plates includes data obtained by comparing a number of defective electrode plates occurring on a path of another secondary battery assembly process with the number of defective electrode plates detected to be abnormal transport. Cho, background, discloses Generally, a secondary battery is a battery which can be repeatedly used through a discharge process of converting chemical energy into electrical energy and a charging process in the reverse direction. Examples of the secondary battery include a nickel-cadmium (Ni-Cd) battery, a nickel- A lithium-metal battery, a lithium-ion (Ni-Ion) battery, and a lithium-ion polymer battery. The secondary battery is composed of an anode, a cathode, an electrolyte, and a separator, and stores and generates electricity using voltage difference between different anode and cathode materials. Here, the discharge is to move electrons from a cathode having a high voltage to a cathode having a low voltage (generating electricity as much as the voltage difference of the anode), and charging means transferring the electrons again from the anode to the cathode where the anode material receives electrons and lithium ions And returned to the original metal oxide. That is, when the secondary battery is charged, the charge current flows as the metal atoms move from the anode to the cathode through the separator, and when discharged, the metal atoms move from the cathode to the anode and the discharge current flows; the rechargeable battery can be divided into a winding type and a lamination type. In the lamination type, a positive electrode plate and a negative electrode plate, which are cut to a predetermined size, are alternately laminated to fabricate an electrode assembly. However, there is a problem that the defective electrode plate is not recognized and the electrode plate is stacked like the normal electrode plate, or the alignment of the electrode plate is disturbed during the stacking process of the electrode plate, resulting in defects; number of pixels present in discharge image are determined in secondary battery). (Burkhardt, [0050-0052], discloses Large input data requirements are further complicated in connection with certain real world situations that occur infrequently. For example, fraud detection and anomaly detection tend to have real world datasets that are highly imbalanced as the nubmer of anomalies within a given dataset tend to be vastly outnumbered by “normal” samples within the dataset. This imbalance tends to hold for welding defect detection as it deals with detecting defects, which are (hopefully) rare incidents. Thus, if real world images were to be used exclusively to train models then the defective weld samples would tend to be a fraction of good samples (e.g., those without defects). Such an imbalance could then fail to properly train the model for detecting accurate defect representations; potential issue is also that even in a case where there is enough data, annotation of large number of images can still be a challenging problem. Specifically, with problems like detecting welding defects, the deep learning models may require the location of the defects within each image. Manually annotating a large set of images requires lot of time and effort and is also prone to errors; the problem of acquiring a large enough dataset can be a problem with Generative Adversarial Networks (GANs) as well. This is because such implementations tend to require large datasets in order to learn the underlying data distribution of source dataset and generated images. For images to be near-realistic GANs would have to be trained on very large datasets; machine learning classifier is trained to determine defective images and total number of defective images output by the classifier is counted as normal or abnormal). Additionally, the rational and motivation to combine the references Cho and Burkhardt as applied in rejection of claim 1 apply to this claim.
Regarding Claim 16,
The combination of Cho and Burkhardt further discloses A non-transitory computer-readable recording medium storing instructions for execution by one or more processors that, when executed by the one or more processors, cause the one or more processors to perform the method according to claim 1.(Burkhardt, [0009-0010], discloses a computer system is provided for generating training datasets to train a model that is used to analyze a manufacturing process. The computer system includes non-transitory memory configured to store at least first and second base images. The computer system also includes a processing system that includes at least one hardware processor. The processing system is configured or otherwise programmed to generate a plurality of training images that are each generated by combining the first and second base images together, wherein each of the plurality of images is generated by randomly varying a location of at least the first base image with respect to the second base image. The processing system is configured or otherwise programmed to train a model based on the plurality of training images and corresponding labels associated with each of the plurality of training images. The processing system is configured or otherwise programmed to receive a plurality of real images of the manufacturing process. The processing system is configured or otherwise programmed to generate, for each of the plurality of real images, a prediction by using the model to predict whether a defect is contained within a corresponding one of the plurality of real images; a non-transitory computer readable storage medium comprising an application program for use with a computer system that generates training datasets for neural networks, the application program comprising instructions that cause the computer system operate in the following manner. Loading a weld seam image and second image of a plate that is to be welded. Generating a preparation image that includes at least two instances of the plate separated by a gap. Generating a training image by sequentially applying a plurality of instances of the weld seam image over the gap, wherein each successive one of the plurality of instances overlaps a prior instance. The location at which each of the plurality of instances is applied over the gap may be varied based on a randomness factor. The generation of training images is repeated until a plurality of training images have been generated. This may be thousands or even millions of different images. The plurality of training images may be varied due to application of the randomness factor that is used in the creation of each one of the plurality of training images. Thus, the plurality of training images may be diverse and robust in nature and include many different examples of defective welds between the plates, including examples that would normally be quite rare).
Claims 17-20 recite system with elements corresponding to the method steps recited in Claims 1-2, 5 and 12 respectively. Therefore, the recited elements of the system claims 17-20 are mapped to the proposed combination in the same manner as the corresponding steps of Claims 1-2, 5 and 12 respectively. Additionally, the rationale and motivation to combine the Cho and Burkhardt references presented in rejection of Claim 1, apply to these claims.
Furthermore, the combination of Cho and Burkhardt further discloses A system for detecting abnormal transport of a defective electrode plate, the system comprising: one or more cameras configured to capture a plurality of images associated with transport of the defective electrode plate on a path of a secondary battery assembly process; and a detection module (Cho, Fig.1, discloses concept of the first embodiment according to the present invention. As shown in the drawing, the electrode plate transfer apparatus of the secondary battery manufacturing apparatus including the electrode plate inspection function according to the present invention includes a robot arm 200, a sensor unit, a control unit 600, and an alarm unit 700)
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
JP2021099913A (Abstract, provide a method for manufacturing an electrode plate capable of appropriately detecting an abnormal shape after the step including a step of cutting the electrode plate. SOLUTION: In manufacturing an electrode plate formed by forming an electrode layer on the surface of a current collector plate, the electrode layer in the electrode plate is formed while transporting the electrode plate after the formation of the electrode layer in the longitudinal direction. A slit process for cutting a position within a certain range to obtain an electrode plate divided in the width direction, a displacement acquisition process for acquiring the width direction displacement and the thickness direction displacement of the cut end of the electrode plate after the slit process, and displacement acquisition. At a position downstream from the execution position of the step, an imaging step of capturing an image of the cut end of the electrode plate is performed while manipulating the imaging target area based on the displacement information acquired in the displacement acquisition step)
US-20210387808-A1 ([0055] A method of fulfilling and delivering orders using order fulfillment and delivery system 10 and robotic system 14 will now be described. The desired items, for example, groceries may be packaged into order containers and/or storage containers and loaded within the storage area 34 of vehicle 12 at a warehouse or other order fulfillment center. The containers 36 may either be loaded into the truck by a warehouse worker or with assistance of gantry 38, a gantry 38′ (shown in FIG. 8), and/or other devices. Gantry 38′ may be constructed and operate similarly to gantry 38 with the exception that gantry 38′ is not secured within delivery truck 12. Instead, gantry 38′ may be provided at any loading/unloading dock and may be on rollers or another mechanism which allows the gantry to be slid into and out from the truck to assist with loading and/or unloading containers 36 from the truck in a more efficient manner than can be performed by a warehouse worker)
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PINALBEN V PATEL whose telephone number is (571)270-5872. The examiner can normally be reached M-F: 10am - 8pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Chineyere Wills-Burns can be reached at (303)297-4332. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/Pinalben Patel/Examiner, Art Unit 2673