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 .
Response to Amendment
The Amendment filed 10/28/2025 has been entered. Claims 21 and 23 have been canceled. Claims 1-20 and 22 are presented for examination.
Claim Rejections – 35 USC § 103
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 1-4, 7-9, 11-13 and 22 are rejected under 35 U.S.C. 103 as being unpatentable over Giering (US 2019/0147283) in view of Dasgupta (US 10,719,301).
Regarding claim 1, Giering teaches a system (Fig. 1, [0025-0027] describes a system), comprising:
a memory that stores computer executable components (Fig. 1, [0026] describes a memory with data and instructions); and
a processor that executes the computer executable components stored in the memory (Fig. 1, [0026] describes a processor), wherein the computer executable components comprise:
an active learning component that trains an analytics artificial intelligence model (Fig. 1, [0026] executable instructions/data; [0029], the system can train a model e.g., a neural network), wherein training the analytics artificial intelligence model comprises:
training the analytics artificial intelligence model using training data comprising training images ([0029], the system can update and train a model based on the annotated training images; this can be done as part of an iterative or incremental process; that is, such training iterations/phases are repeatedly used to update/train the model until a desired performance is achieved by the model);
annotating an image to generate an annotated image ([0029], a patch detector/labeler adds annotations/labels to training images);
retraining the analytics artificial intelligence model using the training data to generate a prediction for a new image associated with an engineering component ([0029], the system can update and train a model based on the annotated training images; this can be done as part of an iterative or incremental process; that is, such training iterations/phases are repeatedly used to update/train the model until a desired performance is achieved by the model i.e., retraining);,
wherein the prediction identifies whether the new image includes a defect by assigning a pass label or a fail label to the new image (Fig. 6, [0024, 0041], the model is used to generate a prediction of a defect or crack in a structure; the prediction/output can be in the form of a heatmap, which overlays image data to highlight the severity of a defect i.e., a pass or fail label)
wherein the prediction is provided to a user ([0039-0041], the result of the model can be presented visually for a user) and
wherein the analytics artificial intelligence model is updated based on feedback received regarding the prediction ([0029], the result of a model can be compared to a ground truth, and further iterations of the model can be run until a desired level of classification confidence is achieved and the model result/prediction is suitable; in other words, through the comparison and retraining, the system effectively provides/receives feedback regarding the model result/prediction and uses that to update the model).
However, Giering does not expressly disclose the annotating using the analytics artificial intelligence model; employing feedback from a plurality of human annotators regarding respective quality of one or more annotations in the annotated image; configuring, by the active learning component, a reviewing process for the one or more annotations including determining whether the annotated image should be annotated a second time by a human annotator, and modifying the reviewing process based on at least one of a confidence level for that human annotator or historical performance of that human annotator evaluated from prior annotations using feedback from another human annotator; updating the annotated image based on the reviewing process, as modified, to generate an updated annotated image; adding the updated annotated image to the training data; wherein the updated annotated image is provided to a user and wherein the analytics artificial intelligence model is updated based on feedback received regarding the updated annotated image.
In the same field of endeavor, Dasgupta teaches
the annotating using the analytics artificial intelligence model (col. 33, line 58 to col. 34, line 26, Dasgupta teaches an annotation system that can apply annotations/labels to training images; for example, a classifier can annotate some training images and present them to a user; the user can examine and correct the annotations/labels; the second set of annotations can be used to train the classifier; this training and annotation can be performed multiple times to train the classifier, which can then be used to make more annotations to training data; note that the claimed “artificial intelligence model” can be considered to be a larger model that includes the error detection model described in Giering and the annotation model described in Dasgupta, which cooperate to generate defect predictions as described in Giering; this larger model is used to provide annotations; additionally or alternatively, Dasgupta teaches a model that is iteratively trained to annotate training images; it would be obvious to incorporate such a model into Giering to support the Giering annotation process);
employing feedback from a plurality of human annotators regarding respective quality of one or more annotations in the annotated image (col. 33, line 58 to col. 34, line 26, Dasgupta teaches an annotation system that can apply annotations/labels to training images; for example, a classifier can annotate some training images and present them to a user; the user can examine and correct the annotations/labels; the second set of annotations can be used to train the classifier; col. 5, lines 43-52; col. 6, lines 21-37, Dasgupta also contemplates a system that enables multiple users and annotators to work together on the same task i.e., annotation of an image);
configuring, by the active learning component, a reviewing process for the one or more annotations including determining whether the annotated image should be annotated a second time by a human annotator (col. 33, line 58 to col. 34, line 26, Dasgupta teaches an annotation system that can apply annotations/labels to training images; for example, a classifier can annotate some training images and present them to a user; the user can examine and correct the annotations/labels; the second set of annotations can be used to train the classifier; see also col. 33, lines 12-25, 57-65, the system determines/selects a set of annotations to present to a reviewer/user, based on whether the annotations meet certain criteria e.g., a confidence metric), and
modifying the reviewing process based on at least one of a confidence level for that human annotator or historical performance of that human annotator evaluated from prior annotations using feedback from another human annotator (col. 33, lines 12-25, when training samples and annotations are presented to a user for review or annotation, the training samples are selectively presented to the user based on a confidence metric i.e., a "confidence level for a user"; see also col. 33, lines 57-65, the system may present training images and annotations/labels that the classifier finds particularly confusing or informative i.e., based on a confidence level used to select images for the classifier; col. 34, lines 6-26, the user may correct annotations made by the classifier, and such actions/corrections inherently indicate or represent a confidence level on the part of the user);
updating the annotated image based on the reviewing process, as modified,
to generate an updated annotated image (col. 33, line 58 to col. 34, line 26, the users may correct the annotations/labels on images by providing feedback via user controls on an interface; put another way, the system selects from the input/feedback provided by the users and thereby generates an image with corrected annotations/labels, which is used to train a model);
adding the updated annotated image to the training data (col. 33, line 58 to col. 34, line 26, Dasgupta teaches an annotation system that can apply annotations/labels to training images; for example, a classifier can annotate some training images and present them to a user; the user can examine and correct the annotations/labels; the second set of annotations can be used to train the classifier);
wherein the updated annotated image is provided to a user and wherein the analytics artificial intelligence model is updated based on feedback received regarding the updated annotated image (col. 33, line 58 to col. 34, line 26, Dasgupta teaches an annotation system that can apply annotations/labels to training images; for example, a classifier can annotate some training images and present them to a user; the user can examine and correct the annotations/labels; the second set of annotations can be used to train the classifier).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to have incorporated the annotating using the analytics artificial intelligence model; employing feedback from a plurality of human annotators regarding respective quality of one or more annotations in the annotated image; configuring, by the active learning component, a reviewing process for the one or more annotations including determining whether the annotated image should be annotated a second time by a human annotator, and modifying the reviewing process based on at least one of a confidence level for that human annotator or historical performance of that human annotator evaluated from prior annotations using feedback from another human annotator; updating the annotated image based on the reviewing process, as modified, to generate an updated annotated image; adding the updated annotated image to the training data; wherein the updated annotated image is provided to a user and wherein the analytics artificial intelligence model is updated based on feedback received regarding the updated annotated image as suggested in Dasgupta into Giering because Giering and Dasgupta pertain to analogous fields of technology. Both Giering and Dasgupta pertain to annotating training data, which is used to improve a model. In Dasgupta, a machine learning classifier is used to annotate the data, and that classifier is trained with annotations/feedback provided by a user. It would be desirable to incorporate this feature into Giering to improve the quality and speed of annotations e.g., see Dasgupta col. 14, lines 43-50; col. 19, line 43 to col. 20, line 12; col. 35, lines 11-28; Fig. 13, col. 31, lines 40-52.
Regarding claim 2, the combination of Giering and Dasgupta teaches the invention as claimed in claim 1. The combination of Giering and Dasgupta also teaches wherein the active learning component is in communication with a server component that stores the analytics artificial intelligence model (Giering [0027], cloud resources i.e., a server component, may provide the model training and any portion of the described process/system e.g., the model).
Regarding claim 3, the combination of Giering and Dasgupta teaches the invention as claimed in claim 1. The combination of Giering and Dasgupta also teaches wherein an annotator device comprises an annotation component that annotates the training data based on data associated with the analytics artificial intelligence model stored by a server component (Giering [0029], the system includes a component e.g., the patch detector, labeler etc., which annotates images as noted above; this component is part of a mechanism/device, such as a memory, system or instructions e.g., as described in Giering [0025-0027]; inherently, this is based on data e.g., at the very least, code or computer instructions that perform the annotation operation; or features in the images etc.; such data is associated with the model, at least because it is used to train the model; Giering Fig. 1, [0027] data sources e.g., the model, can be stored remotely in cloud computing resource 130 i.e., using or related to servers).
Regarding claim 4, the combination of Giering and Dasgupta teaches the invention as claimed in claim 1. The combination of Giering and Dasgupta also teaches wherein the active learning component is implemented on a server (Giering [0027], cloud resources i.e., a server, may provide the model training and any portion of the described process).
Regarding claim 7, the combination of Giering and Dasgupta teaches the invention as claimed in claim 1. The combination of Giering and Dasgupta also teaches wherein the computer executable components comprise: a display component that provides the training data to a display device to display information associated with the training data in a human-interpretable format (Dasgupta col. 33, line 57 to col. 34, line 26, the system provides annotated training images to a display for a user, so the user can review and correct the annotations; the above can be provided by a software component e.g., the model development environment of Dasgupta Fig. 1, the orchaestrator as described in Dasgupta col. 18, lines 35-44, instructions as described in Dasgupta col. 61, lines 8-17, etc.).
Regarding claim 8, the claim corresponds to claim 1 and is rejected for the reasons. (Claim 8 is largely a method claim performed by the system of claim 1). The combination of Giering and Dasgupta also teaches wherein the retraining is performed during an annotation process (Giering Fig. 2, [0028-0029] teaches a training 200; as indicated in the paragraphs and figure, the training includes repeated/iterative training i.e., retraining, and further indicates that part of the training is annotating training images; put another way, as defined in Giering, training or re-training can include various operations related to training and improving a model, such as annotating; naturally, if annotating is part of training/re-training, then the retraining is performed at least in part during the annotating).
Regarding claim 9, the combination of Giering and Dasgupta teaches the invention as claimed in claim 1. The combination of Giering and Dasgupta also teaches providing the artificial intelligence model to a system associated with the engineering component (Giering Fig. 1, [0024-0025, 0027] teaches a system that provides a machine learning model, which is used to analyze structures i.e., engineering components).
Regarding claim 11, the combination of Giering and Dasgupta teaches the invention as claimed in claim 8. The combination of Giering and Dasgupta also teaches receiving one or more annotations associated with the training data from a display device associated with a user identity (Dasgupta col. 33, line 57 to col. 34, line 26, the system may annotate training images and present them to a user; the user can examine the images and provide corrected annotations using controls on a displayed training interface; the corrected annotations are used to train the annotation classifier).
Regarding claim 12, the combination of Giering and Dasgupta teaches the invention as claimed in claim 8. Claim 12 also corresponds to claim 2 and is rejected for the same reasons.
Regarding claim 13, the combination of Giering and Dasgupta teaches the invention as claimed in claim 8. Claim 13 also corresponds to claim 3 and is rejected for the same reasons.
Regarding claim 22, the combination of Giering and Dasgupta teaches the invention as claimed in claim 8. The combination of Giering and Dasgupta also teaches assisting with pre-annotation (“assisting with pre-annotation” can be understood as any operation that helps with any operation occurring before annotation; Giering [0029-0030] teaches a variety of steps that occur before and/or help facilitate annotation; for example, Giering teaches preprocessing, receiving and selecting images to annotate, identifiying areas of interest in images, annotating and training a model prior to a particular round of annotations etc.; see also Dasgupta col. 33, lines 12-25; col. 34, lines 7-26, the system iterates through a process of having a classifier annotate images; the images are presented to a user; the user updates the annotations based on their review; and the classifier is trained using those annotations and generates further annotations; thus, before each iteration of this annotation process, the classifier and a user are both configured to provide a prediction as to how to annotate/label an image; see also Dasgupta col. 19, line 58 to col. 20, line 12; col. 31, line 32 to col. 32, line 15; col. 33, lines 12-25; col. 33, line 57 to col. 34, line 27, which describes various pre-annotation processes e.g., e.g., providing an interface to facilitate annotations, selecting images to be annotated, training a model prior to another round of annotations, processing to automatically generate an annotation, etc.).
Claims 5, 6, 14 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Giering and Dasgupta, as applied in claims 1 and 8, and further in view of Han (US 2020/0342322).
Regarding claim 5, Giering and Dasgupta teaches the invention as claimed in claim 1. However, the combination of Giering and Dasgupta does not expressly disclose wherein the active learning component incrementally updates the analytics artificial intelligence model via a first modeling process, and wherein the first modeling process simultaneously updates the analytics artificial intelligence model with respect to a second modeling process that incrementally updates the analytics artificial intelligence model.
In the same field of endeavor, Han teaches wherein the active learning component incrementally updates the analytics artificial intelligence model via a first modeling process, and wherein the first modeling process simultaneously updates the analytics artificial intelligence model with respect to a second modeling process that incrementally updates the analytics artificial intelligence model (Fig. 1, [0019-0023], a model can be split into sub-models, and then the sub-models can be trained concurrently).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to have incorporated wherein the active learning component incrementally updates the analytics artificial intelligence model via a first modeling process, and wherein the first modeling process simultaneously updates the analytics artificial intelligence model with respect to a second modeling process that incrementally updates the analytics artificial intelligence model as suggested in Han into Giering and Dasgupta because Giering and Han pertain to analogous fields of technology. Both Giering and Han pertain to the training of a machine learning model. In Han, a model can be split into sub-models, which can be trained concurrently. It would be desirable to incorporate this feature into Giering to accelerate the training process e.g., see Han [0023-0024].
Regarding claim 6, the combination of Giering, Dasgupta and Han teaches the invention as claimed in claim 5. The combination of Giering, Dasgupta and Han also teaches wherein the active learning component selects, for storage in a data store associated with a server, a first version of the analytics artificial intelligence model associated with the first modeling process or a second version of the analytics artificial intelligence model associated with the second modeling process (Han Fig. 1, [0019-0023], the system splits the model into sub-models e.g., first and second versions; inherently, such a sub-model is stored in some form of data store; [0024], the Han process may be performed at a server; see also Giering [0027], which notes that training can be supported by the cloud i.e., a server; inherently, when the system trains a sub-model, it selects that sub-model for such training).
Regarding claim 14, the combination of Giering and Dasgupta teaches the invention as claimed in claim 8. However, the combination of Giering and Dasgupta does not expressly disclose wherein the artificial intelligence model is incrementally updated by performing a first modeling process associated with a first version of the artificial intelligence model and performing a second modeling process associated with a second version of the artificial intelligence model.
In the same field of endeavor, Han teaches wherein the artificial intelligence model is incrementally updated by performing a first modeling process associated with a first version of the artificial intelligence model and performing a second modeling process associated with a second version of the artificial intelligence model (Fig. 1, [0019-0023], a model can be split into sub-models, and then the sub-models can be trained concurrently).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to have incorporated wherein the artificial intelligence model is incrementally updated by performing a first modeling process associated with a first version of the artificial intelligence model and performing a second modeling process associated with a second version of the artificial intelligence model as suggested in Han into Giering and Dasgupta because Giering and Han pertain to analogous fields of technology. Both Giering and Han pertain to the training of a machine learning model. In Han, a model can be split into sub-models, which can be trained concurrently. It would be desirable to incorporate this feature into Giering to accelerate the training process e.g., see Han [0023-0024].
Regarding claim 15, the combination of Giering, Dasgupta and Han teaches the invention as claimed in claim 14. The combination of Giering, Dasgupta and Han also teaches selecting the first version of the artificial intelligence model or the second version of the artificial intelligence model for storage by a server (Han Fig. 1, [0019-0023], the system splits the model into sub-models e.g., first and second versions; [0024], the Han process may be performed at a server; see also Giering [0027], which notes that training can be supported by the cloud i.e., a server; inherently, when the system trains a sub-model, it selects that sub-model for such training).
Claims 10 and 16-18 are rejected under 35 U.S.C. 103 as being unpatentable over Giering and Dasgupta, as applied in claim 8, and further in view of Niemi (US 2019/0294960).
Regarding claim 10, the combination of Giering and Dasgupta teaches the invention as claimed in claim 8. However, the combination of Giering and Dasgupta does not expressly disclose providing the artificial intelligence model to a display device to display information associated with the artificial intelligence model in a human-interpretable format.
In the same field of endeavor, Niemi teaches providing the artificial intelligence model to a display device to display information associated with the artificial intelligence model in a human-interpretable format (Figs. 1, 4, [0064, 0067, 0058, 0007], it is known to deploy a machine learning model to a client device, to be used in an application on the client device; it is further known that the client may display an application that uses the downloaded model; it should be further noted that the phrase “to display information” can be understood as an intended use, and thus lacks significant patentable weight).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to have incorporated providing the artificial intelligence model to a display device to display information associated with the artificial intelligence model in a human-interpretable format as suggested in Niemi into Giering and Dasgupta because Giering and Niemi pertain to analogous fields of technology. Both Giering and Niemi pertain to training and utilizing machine learning models. In Niemi, a machine learning model can be downloaded to a client and used locally. It would be desirable to incorporate this feature into Giering so that the trained model can be implemented at a local device e.g., see Niemi Figs. 1, 4, [0064, 0067, 0058, 0007].
Regarding claim 16, the claim corresponds to claim 10 and is rejected for the same reasons. The combination of Giering, Dasgupta and Niemi also teaches a non-transitory computer readable storage device comprising instructions that, in response to execution, cause a system comprising a processor to perform operations (Giering Fig. 1, [0025-0027] teaches a memory and a processor).
Regarding claim 17, the combination of Giering, Dasgupta and Niemi teaches the invention as claimed in claim 16. The combination of Giering, Dasgupta and Niemi also teaches wherein the operations further comprise communicating with a server that stores the artificial intelligence model (Giering [0027], cloud resources i.e., a server, may provide the model training and any portion of the described process e.g., the model itself; the system is in communication with the cloud resources; see also Niemi Fig. 1, [0012-0013, 0033], which notes that model data can be stored on a enterprise computing environment i.e., a server; a client is in communication with the server, and the server can respond to requests from the client).
Regarding claim 18, the combination of Giering, Dasgupta and Niemi teaches the invention as claimed in claim 16. The combination of Giering, Dasgupta and Niemi also teaches wherein the training data is annotated based on data associated with the artificial intelligence model stored by a server (Giering [0029], the system annotates images as noted above; inherently, this is based on data e.g., at the very least, code or computer instructions that perform the annotation operation; or features in the images etc.; such data is associated with the model, at least because it is used to train the model; Giering Fig. 1, [0027] data sources e.g., the model, can be stored remotely in cloud computing resource 130 i.e., using or related to servers).
Claims 19 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Giering, Dasgupta and Niemi, as applied in claim 16, and further in view of Han.
Regarding claim 19, the combination of Giering, Dasgupta and Niemi teaches the invention as claimed in claim 16. However, the combination of Giering, Dasgupta and Han does not expressly disclose wherein the artificial intelligence model is incrementally updated by performing a first modeling process associated with a first version of the artificial intelligence model and performing a second modeling process associated with a second version of the artificial intelligence model.
In the same field of endeavor, Han teaches wherein the artificial intelligence model is incrementally updated by performing a first modeling process associated with a first version of the artificial intelligence model and performing a second modeling process associated with a second version of the artificial intelligence model (Fig. 1, [0019-0023], a model can be split into sub-models, and then the sub-models can be trained concurrently).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to have incorporated wherein the artificial intelligence model is incrementally updated by performing a first modeling process associated with a first version of the artificial intelligence model and performing a second modeling process associated with a second version of the artificial intelligence model as suggested in Han into Giering, Dasgupta and Niemi because Giering and Han pertain to analogous fields of technology. Both Giering and Han pertain to the training of a machine learning model. In Han, a model can be split into sub-models, which can be trained concurrently. It would be desirable to incorporate this feature into Giering to accelerate the training process e.g., see Han [0023-0024].
Regarding claim 20, the combination of Giering, Dasgupta, Niemi and Han teaches the invention as claimed in claim 19. The combination of Giering, Dasgupta, Niemi and Han also teaches wherein the operations further comprise selecting the first version of the artificial intelligence model or the second version of the artificial intelligence model for storage by a server (Han Fig. 1, [0019-0023], the system splits the model into sub-models e.g., first and second versions; Han [0024], the Han process may be performed at a server; see also Giering [0027], which notes that training can be supported by the cloud i.e., a server; inherently, when the system trains a sub-model, it selects that sub-model for such training).
Response to Arguments
The Examiner acknowledges the Applicant's amendments to claims 1, 8 and 16.
Regarding independent claims 1, 8 and 16, Applicant alleges that the cited prior art does not teach the amended limitation of "configuring, by the active learning component, a reviewing process for the one or more annotations including determining whether the annotated image should be annotated a second time by a human annotator, and modifying the reviewing process based on at least one of a confidence level for that human annotator or historical performance of that human annotator evaluated from prior annotations using feedback from another human annotator; updating the annotated image based on the reviewing process, as modified, to generate an updated annotated image."
Examiner respectfully disagrees. Dasgupta teaches a system in which a classifier first annotates/labels training images. Afterward, the system allows a human reviewer to review the annotations and selectively correct them i.e., "determining whether the annotated image should be annotated a second time …" Additionally, the system determines a confidence metric or other considerations, to determine whether a particular annotation should be presented to the human reviewer i.e., "a confidence level for a user." For example, the system may select annotations for presentation to the human reviewer only if they meet confidence metric criteria, or if there is a finding that the annotation is particularly confusing or informative. It should be further noted that when the human reviewer provides input to correct the annotation, it is inherent that such input and corrections are based on or represent some level of confidence on the part of the human reviewer and the system controlled by the human reviewer. For at least the above reasons, Examiner respectfully submits that Dasgupta teaches the amended limitation, "determining whether the annotated image should be annotated a second time by a human annotator, and modifying the reviewing process based on … a confidence level for that human annotator."
Applicant further alleges that claims 2-7, 9-15, 17-20 and 22 are allowable in view of their dependency on claims 1, 8 and 16. Claims 2-7, 9-15, 17-20 and 22 are rejected as being taught by Giering, Dasgupta, Han and/or Niemi.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Lupien (US 10,482,167) teaches a system that passes an annotated image to a human reviewer based on a confidence level falling below a threshold e.g., see Lupien Abstract, claims 4 and 5.
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 extension fee 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 ERIC YOON whose telephone number is (408)918-7581. The examiner can normally be reached on 9 am to 5 pm ET Monday through Friday.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Scott Baderman, can be reached at telephone number 571-272-3644. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from Patent Center. Status information for published applications may be obtained from Patent Center. Status information for unpublished applications is available through Patent Center for authorized users only. Should you have questions about access to Patent Center, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
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) Form at https://www.uspto.gov/patents/uspto-automated- interview-request-air-form.
/ERIC J YOON/Primary Examiner, Art Unit 2118
Prosecution Insights