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 .
DETAILED ACTION
This action is in response to the applicant's communication filed on 01/21/2026. In virtue of this communication, claims 1-9 filed on 01/21/2026 are currently pending in the instant application.
New claims 7-9 have been added.
Claims 1, 5, and 6 have been amended without adding a new subject matter.
Response to Arguments
Applicant's arguments filed 01/28/2026 have been fully considered but they are not persuasive.
Applicant has argued:
Independent claim 1 has been amended to provide further clarification. Support for the clarifying amendments is provided in paragraphs [0020] and [0021] of the specification, for example. As amended, claim 1 recites a method of virtually inspecting the quality of a product in a production environment using one or more classifier models. The method includes determining a label, via a virtual inspection, for the product using a first classifier model based on production information associated with the product. The method further includes comparing the label against inspection data associated with a physical inspection of the quality of the product. The cited art does not teach or suggest such a method. Specifically, Lee does not teach or suggest comparing a label (i.e., determined via virtual inspection) against inspection data associated with a physical inspection of the quality of the product. The Office has argued that paragraph [0032] of Lee discloses such a feature. (Office Action, p. 4.) Applicants respectfully disagree. Paragraph [0032] of Lee discloses a process in which a classifier performs a pattern recognition and generates a fault detection result.
Examiner respectfully disagrees, Examiner notes the virtual inspection would be interpreted as using software/program or computer code to perform the inspection, the prior art Lee paragraph [0023] discloses use of CNN, ML, and DNN, i.e. software program, which is considered as virtual inspection. Furthermore, paragraph [0023] discloses the use of ML to process raw trace data, and paragraph [0020] discloses the raw trace data being collected from various sensors such as temperature, time, density, etc. which are being interpreted as physical characteristics being captured by sensor of the quality of the product.
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 (i.e., changing from AIA to pre-AIA ) 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, 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.
Claim(s) 1-6 and 7 is/are rejected under 35 U.S.C. 103 as being unpatentable over Lee et al. (EP3693823), in view of Li (CN 110069607 A).
As per claim 1, Lee discloses “a method of virtually inspecting a quality of a product in a production environment using one or more classifier models, the method comprising:” (Lee, ¶[0001] discloses detecting fault in a manufacturing process using machine learning model, ¶[0005] using the convolutional neural network model; providing a fault indicator of the fault condition in the one or more manufacturing processes of the electronic device; and classifying whether the electronic device includes a fault based on the fault indicator).
“receiving production information associated with the product, wherein the production information is indicative of an operation performed on the product, in relation to a first process of the production environment;”(Lee ¶ [0020] discloses raw trace data maybe collection of data that reflect conditions and parameters of the manufacturing process, i.e. an operation performed on the product, ¶ [0022] discloses converting raw trace data to a data format that is adequate as an input to a data analytic tool, ¶ [0023] discloses the converted trace data may be fed, i.e. receiving product information associated with the product, to a standard data analytic model.)
“determining a label for the product using a first classifier model based on the production information associated with the product;”(Lee, ¶[0023] discloses classifying, i.e. determining a label, patterns to detect and predict fault and defect patterns of the converted trace data.)
“comparing the label against inspection data associated with an inspection of the product; the production information and the inspection data associated with the product” (Lee, ¶ [0032] discloses a fault indicator may then be provided of the fault condition in the one or more manufacturing processes of the electronic device. Then, classifying, i.e. comparing patterns of electronic device against the fault indicator, whether the electronic device includes a fault may be performed based on the fault indicator).
“and retraining the first classifier model using one or more samples present (Lee, ¶ [0077] discloses re-training the convolutional neural network model with the third set of data; and obtaining an updated convolutional neural network model),
“wherein each sample of the one or more samples comprises production information and corresponding inspection data associated with a corresponding product (Lee, ¶ [0078] discloses the second set of data may be a plurality of time-series sensor data obtained in one or more manufacturing processes of an electronic device, providing the third set of data having the first format of the 2D data array to the updated convolutional neural network model).
“and wherein the corresponding inspection data comprises a new label indicative of a condition of corresponding product determined during the inspection.” (Lee, ¶[0078] discloses identifying a pattern in the 2D data array that correlates to a fault condition using the updated convolutional neural network model; providing a fault indicator of the fault condition in the one or more manufacturing processes of the electronic device; and determining that the electronic device includes a fault based on the fault indicator.)
Lee does not discloses the following which would have been obvious in view of Li from similar field of endeavor “storing the information and the data associated in an extension buffer, based on the comparing of the label and the inspection data, for retraining the first classifier model, retraining the classifier model using one or more samples present in the extension buffer for adding one or more new labels upon detecting a predefined number of samples present in the extension buffer,” (Li, pages 34, lines 3-6 discloses tagging underdetermined new sample data, which includes comparing the new sample data against predefined labels. Furthermore, it discloses the new sample data is being stored in training buffer, i.e. extension buffer, Furthermore, page 34, lines 23-28 discloses whenever the number of new sample in the training buffer sample base is accumulated to a threshold value, i.e. predefined number of samples, updating and retraining using new samples).
Before the effective filing date of the claimed invention it would have been obvious to a person of ordinary skill in the art to combine Li technique of training buffer samples into Lee technique to provide the known and expected uses and benefits of Li technique over inspecting product technique of Lee. The proposed combination would have constituted a mere arrangement of old elements with each performing their known function, the combination yielding no more than one would expect from such an arrangement. Therefore, it would have been obvious to a person of ordinary skill in the art to incorporate Li to Lee in order to improve quality of service to client. (Refer to Li background.)
As per Claims 5 and 6, claims have been analyzed and are rejected for the reasons indicated in claim 1 above.
As per Claim 2, in view of claim 1, Lee as modified by Li discloses “determining a number of samples in the extension buffer” (Li, page 34, lines 23-28 discloses whenever the number of new sample in the training buffer sample base is accumulated to a threshold value, i.e. predefined number of samples, updating and retraining using new samples).
As per Claim 3, in view of Claim 1, Lee as modified by Li discloses “wherein the production environment includes one or more processes, and wherein each process of the one or more processes comprises one or more operations to be performed” (Lee, ¶ [0005] discloses one or more manufacturing processes of an electronic device).
As per Claim 4, in view of claim 1, Lee as modified by Li discloses “wherein the production information comprises visual data associated with the product subsequent to performance of the operation, process data of the first process associated with the product, or a combination thereof” (Lee ¶ [0020] discloses raw trace data maybe collection of data that reflect conditions and parameters of the manufacturing process, i.e. an operation performed on the product, ¶ [0022] discloses converting raw trace data to a data format that is adequate as an input to a data analytic tool, ¶ [0023] discloses the converted trace data may be fed, i.e. receiving product information associated with the product, to a standard data analytic model, ¶ [0030], disclose the raw trace data 101 is one-dimensional time-series data of multiple parameters, and the preprocessed trace data 115 is a two-dimensional (2D) image-like data. The 2D image-like data has a first dimension that corresponds to the parameters and a second dimension that corresponds to data samples in time. The first and second dimensions correspond to X and Y coordinates of the 2D image-like data.).
As per Claim 7, in view of claim 1, Lee as modified by Li discloses “wherein the first classifier model is trained to determine the label associated with the product based on the received production information” (Lee, ¶ [0023] discloses training the model using trace data, ¶ [0020] discloses the trace data being production information such as time, density, temperature, pressure, [0032] discloses using the model to determine various kind of faults, i.e. labels, such as etching process defect, a window defect, a Mura defect, etc.)
Claim(s) 8 is/are rejected under 35 U.S.C. 103 as being unpatentable over
Lee et al. (EP3693823), in view of Li (CN 110069607 A), further in view of Wang (US 2019/0069377).
As per Claim 8, in view of claim 1, Lee as modified by Li does not disclose the following which would have been obvious in view of Wang from similar field of endeavor “wherein, when the label does not match the inspection data, the first classifier model is identified as not working properly, and the production information and the inspection data is stored in the extension buffer for retraining the first classifier model” (Wang, ¶ [0025] discloses making predictions, ¶ [0027]-[0028] discloses self-validate and reconciling the result, which implies the misprediction, ¶[0030] discloses retraining the classifier using the updated results from the buffer.)
Before the effective filing date of the claimed invention it would have been obvious to a person of ordinary skill in the art to combine Wang technique of automated inspection into Lee as modified by Li technique to provide the known and expected uses and benefits of Wang technique over inspecting product technique of Lee as modified by Li. The proposed combination would have constituted a mere arrangement of old elements with each performing their known function, the combination yielding no more than one would expect from such an arrangement. Therefore, it would have been obvious to a person of ordinary skill in the art to incorporate Wang to Lee as modified by Li in order to provide safe inspection system. (Refer to Wang ¶[0002].)
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over
Lee et al. (EP3693823), in view of Li (CN 110069607 A), further in view of Miserendino et al. (US 2017/0032279), further in view of Faulhaber et al .( US 2019/0156247).
As per claim 9, The method of claim 1, further comprising: Lee as modified by Li does not explicitly disclose the following which would have been obvious in view of Miserendino from similar field of endeavor “testing the retrained first classifier model prior to replacement of a previous first classifier model,” (Miserendino, ¶[0025] discloses at the end of the evaluation period, in embodiments the user elects to either accept the new classifier and replace the current parent classifier or reject it and keep the parent classifier, block 13. In another aspect of this disclosure, the in-situ retraining system, without human intervention, may accept the new classifier and replace the current parent classifier or reject it and keep the parent classifier. In either case, the process may be repeated, e.g. at the user's discretion or by the in-situ retraining system, block 14. Once a new in-situ classifier is accepted it becomes the parent/base classifier for the next round of in-situ retraining 100. The user can further elect to deploy the retrained classifier model to all in-situ retraining systems throughout their enterprise thereby replacing each system's parent classifier with the new retrained classifier. )
Before the effective filing date of the claimed invention it would have been obvious to a person of ordinary skill in the art to combine Miserendino technique of classifier evaluation into Lee as modified by Li technique to provide the known and expected uses and benefits of Miserendino technique over inspecting product technique of Lee as modified by Li. The proposed combination would have constituted a mere arrangement of old elements with each performing their known function, the combination yielding no more than one would expect from such an arrangement. Therefore, it would have been obvious to a person of ordinary skill in the art to incorporate Miserendino to Lee as modified by Li in order to provide a parent classifier to be replaced which is not vulnerable. (Refer to Miserendino ¶[0008].)
However Lee as modified by Li as modified by Miserendino does not explicitly disclose the following which would have been obvious in view of Faulhaber from similar field of endeavor “wherein the retrained first classifier model is deployed and replaces the previous first classifier model when an accuracy of the retrained first classifier model exceeds a predefined threshold.”(¶[0036] discloses user may have a new model (e.g., an updated model, such as one trained using different hyperparameter values) that they seek to deploy. Before the user makes this model the public or “live” model (which would give results back to clients), the dynamic router 108 can apply both an old model (or models) and the new model for incoming requests that are actually serviced by an old model. the analytics engine 122 can interact with a ground truth collector 124 to obtain ground truth for a set of requests, and compare this obtained ground truth with the inference results generated by the ML model(s) 118A-118C under scrutiny to identify the true accuracy of these models. ¶[0037] Depending upon the particular use case, the analytics engine 122 can act in a variety of ways after such determinations, including but not limited to sending an update message 138 to the dynamic router 108 to cause the model selector 110 to switch over some or all traffic to a “new” model (e.g., if its performance meets or exceeds some threshold, such as having an accuracy value that is greater than the “old” model's corresponding accuracy value), sending analytic results 202 to a logging system or client, etc.)
Before the effective filing date of the claimed invention it would have been obvious to a person of ordinary skill in the art to combine Faulhaber technique of dynamic accuracy monitoring into Lee as modified by Li as modified by Miserendino technique to provide the known and expected uses and benefits of Faulhaber technique over inspecting product technique of Lee as modified by Li as modified by Miserendino. The proposed combination would have constituted a mere arrangement of old elements with each performing their known function, the combination yielding no more than one would expect from such an arrangement. Therefore, it would have been obvious to a person of ordinary skill in the art to incorporate Faulhaber to Lee as modified by Li as modified by Miserendino in order to provide models with less complexities. (Refer to Faulhaber ¶[0003].)
Conclusions
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHAGHAYEGH AZIMA whose telephone number is (571)272-1459. The examiner can normally be reached Monday-Friday, 9:30-6:30.
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, Vincent Rudolph can be reached at (571)272-8243. 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.
/SHAGHAYEGH AZIMA/Examiner, Art Unit 2671