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 . The rejections and objection from the Office Action of 5/7/2026 are hereby withdrawn. New grounds for rejection are presented below.
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, 2, 5, 6, 7, 9, 10, 12, 13, and 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kral et al., Crack Propagation Analysis Using Acoustic Emission Sensors for Structural Health Monitoring Systems, The ScientificWorld Journal, 2013 [hereinafter “Kral”]; Janeliukstis et al., A Novel Separation Technique of Flexural Loading-Induced Acoustic Emission Sources in Railway Prestressed Concrete Sleepers, IEEE, 2019 [hereinafter “Janeliukstis”]; Xiang-jun et al., Application of Wavelet Analysis in Vibration Signal Processing of Bridge Structure, IEEE, 2010 [hereinafter “Xiang-jun”]; and Szegedy et al., Going deeper with convolutions, arXiv, 2014 [hereinafter “Szegedy”].
Regarding Claims 1 and 9, Kral discloses a real-time structural health monitoring computing system (and corresponding method) for predicting damage zone levels in a monitored structure using acoustic emission (AE) data [Page 10, second column – “Here the outputs of the network categorized each histogram into either crack growth or noise present. The datasets determined to be crack growth and not noise were then used in the severity network. This network then determined the size of the increment of crack growth over the time window. This experiment used two separate sensors. The data from each sensor were separated and run through the two neural networks. Figure 13(a) contains graphs of the results of the networks in terms of crack length. As time increased in the experiment, the total crack length increased. Finally an average of the two signals was taken to determine a net crack length value. This average crack growth length is illustrated again in Figure 13(b) along with the load history.”Page 1, second column – “The basic acoustic emission system was augmented with an artificial neural network analysis to provide near real-time analysis of acoustic emission data measured from aircraft structural components, during routine service operations.”], comprising: one or more processors; and one or more non-transitory computer-readable media that collectively store: machine-executable instructions for a machine-learned neural network architecture model comprising a multi-layer convolutional neural network (CNN) featuring dimension-reduction layers and multi-scale filters [See Fig. 5 and Page 9, second column – “The connections between the Kohonen layer and the output nodes were trained with the backpropagation, using the NeuralWorks software Delta rule, and used hyperbolic tangent activation functions. The purpose of this network was to filter out noise from the strain waves corresponding to actual crack growth.”], the model trained to use AE data from a singular sensor channel [Fig. 13(a), either of Ch 1 or Ch 2.] to predictively model Structural Health Maintenance (SHM) of the structure [Page 2, first column – “1.1. Acoustic Emission. As a crack propagates in a material, molecular bonds are broken, releasing small amounts of energy. The energy released spreads throughout the surrounding material in the form of strain waves. These waves are minute deformations in the material with wave frequencies in the ultrasonic range from 500 kHz to 3MHz. Generally all structural deformations transmit some form of energy into the material, resulting in waves similar to those of crack growth. The acoustic emission system of the study consisted of piezoelectric sensors, which were configured to receive waves, generated by other sources, such as crack extensions or impact events, within the structural component under investigation. … The recorded voltage time histories were broken down into characteristics of the waves, such as amplitude, rise time, and duration, using software provided by Physical Acoustics Corporation [4]. These characteristics of the waves were recorded with a network of sensors and analyzed via different software methods through MATLAB [5] and NeuralWorks [6] to determine if cracks were present and growing and whether the structural component should be replaced. A custom designed artificial neural network was used for the post-processing analysis of the detected waves.”Page 4, first column – “Because of the training process of neural networks, a complex relationship of inputs to outputs can be found quickly, accurately, and precisely if taught well. The advantages offered by the neural network when applied to a structural health monitoring system of ultrasonic sensors allow for quick assessment of the complex strain wave signals generated by the piezoelectric signals.”] without reliance on historical sensor data [Use of real-time sensor data, see Fig. 13(a), either of Ch 1 or Ch 2.]; and
instructions that, when executed by the one or more processors, configure the computing system to perform operations, the operations comprising: obtaining raw detected AE data from sensors used with an associated structure to be monitored [See Figs. 12 and 13 and Table 1. Fig. 13(a), either of Ch 1 or Ch 2.Page 10 – “3.1.2. Testing Datasets. Once the two artificial neural networks were created and fully trained, the next step was to use these in a situation, where datasets not previously presented to the networks were used. … The measured AE data shown in Figure 12 indicate that there was a great deal of noise and many strain waves detected after the crack extension was initiated. This dataset was evaluated using the neural networks, using the same time windows as those of the training sets.”];
filtering the raw detected AE data using predetermined decision tree-based rules; determining at least two respective different damage zones associated with the monitored structure using decision tree-based rules [Page 10, second column – “Histograms were made of ten bins each with the same range as before. This new dataset was first used in the yes-no network.”Page 9, first column – “Two artificial neural networks were created for two separate purposes. Both neural networks used the ten histogram bin values as input sets.The first network, named the “yes-no network,” was a self-organizing map. This network was used to classify each time window into two groups; “yes” crack growth was present or “no” crack growth present.”Page 9, second column – “With this first network being completely trained a second neural network, called the “severity network,” was constructed. This severity network used the histogram values to determine the crack growth extension in inches.”];
inputting the filtered AE data into the machine-learned neural network architecture model [Page 4, first column – “Because of the training process of neural networks, a complex relationship of inputs to outputs can be found quickly, accurately, and precisely if taught well. The advantages offered by the neural network when applied to a structural health monitoring system of ultrasonic sensors allow for quick assessment of the complex strain wave signals generated by the piezoelectric signals.”Page 1, second column – “The basic acoustic emission system was augmented with an artificial neural network analysis to provide near real-time analysis of acoustic emission data measured from aircraft structural components, during routine service operations.”Page 10, second column – “Histograms were made of ten bins each with the same range as before. This new dataset was first used in the yes-no network.”Page 9, first column – “Two artificial neural networks were created for two separate purposes. Both neural networks used the ten histogram bin values as input sets.The first network, named the “yes-no network,” was a self-organizing map. This network was used to classify each time window into two groups; “yes” crack growth was present or “no” crack growth present.”Page 9, second column – “With this first network being completely trained a second neural network, called the “severity network,” was constructed. This severity network used the histogram values to determine the crack growth extension in inches.”]; and
as an output of the machine-learned neural network architecture model, in real-time [Page 4, first column – “Because of the training process of neural networks, a complex relationship of inputs to outputs can be found quickly, accurately, and precisely if taught well. The advantages offered by the neural network when applied to a structural health monitoring system of ultrasonic sensors allow for quick assessment of the complex strain wave signals generated by the piezoelectric signals.”Page 1, second column – “The basic acoustic emission system was augmented with an artificial neural network analysis to provide near real-time analysis of acoustic emission data measured from aircraft structural components, during routine service operations.”], determining damage zone level predictions for the determined respective different damage zones of the associated monitored structure [Page 10, second column – “Here the outputs of the network categorized each histogram into either crack growth or noise present. The datasets determined to be crack growth and not noise were then used in the severity network. This network then determined the size of the increment of crack growth over the time window. This experiment used two separate sensors. The data from each sensor were separated and run through the two neural networks. Figure 13(a) contains graphs of the results of the networks in terms of crack length. As time increased in the experiment, the total crack length increased. Finally an average of the two signals was taken to determine a net crack length value. This average crack growth length is illustrated again in Figure 13(b) along with the load history.”See the damages zones established by the y-axis of Figs. 13(a)/(b).].
Kral fails to disclose transforming the raw detected AE data into a plurality of signal features comprising a peak frequency and a duration; and performing filtering to filter out anomalous signals using set thresholds comprising a peak frequency threshold and a duration threshold.
However, Janeliukstis discloses performing AE data filtering in such a manner [Page 51429, first column – “In addition, each sensor is connected to PAC 2/4/6 preamplifier operating in the frequency bandwidth of 20–1200 kHz. The amplification level of the pre-amplifiers (acoustic emission signal capturing threshold) was set to 50 dB prior to testing. An AE signal must surpass this threshold magnitude to filter out unwanted noise.”Page 51429, second column – “The sampling rate for recording AE signals was set to 1 MSample/s, peak definition time (PDT), hit definition time (HDT) and hit lockout time (HLT) were set to 600 µs, 600 µs and 1000 µs, respectively.”]. It would have been obvious to analyze and filter the AE data with regards to such features in order to reduce the amount of irrelevant data to be analyzed.
Kral also fails to disclose that the model is configured to process two-dimensional (2D) wavelet images; converting the filtered AE data into the 2D wavelet images for high-frequency AE processing by normalizing the filtered AE data and scaling wavelet coefficients; and that the 2D wavelet images are input to the neural network model.
However, Xiang-jun discloses the use of the wavelet transform on vibration data in performing structure monitoring [Abstract – “Data processing is important to the structure health monitoring system which produces large volumes of raw data containing the useful information and the noise. Wavelet analysis is a newly emerging theory in data processing field, which has good localization characteristics in both frequency and time domains compared to most traditional methods used for structural health monitoring. Wavelet can be used for discovering the local feature of a signal by selecting a proper basic wavelet. In addition, the feature components of a signal can be obtained by reconstructing the wavelet coefficients. Wavelet technique is adopted to process the vibration signals acquired from the bridge monitoring spot in this paper. Based on the wavelet analysis theory, an efficient signal processing approach to structure health monitoring has been developed. The results of analysis show that the method is not only feasible to signal de-noising, but also valuable and effective to detect the health status of bridge structure.”] through use of wavelet coefficient scaling [Page 672, first column – “The mother wavelet ψ(t) needs to satisfy certain admissibility condition in order to ensure existence of the inverse wavelet transform. The dilation parameter “a” and the translation parameter “b” are also referred as the scaling and shifting parameters respectively and play an important role in the wavelet analysis.”].
It would have been obvious to transform the AE data of Kral, prior to performing normalization [Page 8, second column of Kral – “A sliding time window for real-time monitoring was created that stepped through time at a step of 1.6 seconds. This procedure allowed for multiple readings from the same detection (see Figure 10). Within each time window, multiple detections could be observed. These were normalized into a histogram of the data within the time window, thus removing any time dependence. A histogram was made of 10 bins, grouping the values of the energy value of each strain wave between zero and a normalized maximum value of 100.”], using the wavelet transform and to use such data in monitoring a structure because doing so would have been useful in denoising the data and would have made the structural monitoring more accurate.
Kral fails to disclose that the convolutional neural network (CNN) features pooling layers and multiple convolutional kernels.
However, Szegedy discloses the use of a GoogLeNet convolutional neural network (CNN) as a high performing classifier [See Abstract.See Fig. 3 – “GoogLeNet network with all the bells and whistles”See Table 2.]. It would have been obvious to use such a neural network as the neural network because it was known to have good performance. Szegedy discloses the GoogLeNet convolutional neural network (CNN) features pooling layers [Page 6 – “The network is 22 layers deep when counting only layers with parameters (or 27 layers if we also count pooling).”] and multiple convolutional kernels [Page 5 – “That is, 1×1 convolutions are used to compute reductions before the expensive 3×3 and 5×5 convolutions.”Page 8 – “A1×1convolution with 128 filters for dimension reduction and rectified linear activation”].
Regarding Claim 2, Kral discloses that the one or more processors are further configured so that the determining operations include determining damage zone level predictions separately for each of a plurality greater than two of respective different damage zones of the associated monitored structure [See Fig. 13, the discrete determined crack lengths].
Regarding Claim 3, Kral discloses that the one or more processors are further configured so that the machine-learned Al-enabled technology neural network architecture model learns to predict damage level of respective damage zones directly from raw AE data signals [Page 8 – “3.1.1. Training Datasets. … This process allowed for acoustic emission detections for a series of finite increments of crack growth, which could then be used for a training set for an artificial neural network to identify a crack extension event.”], for estimating in real-time the damage level of respective damage zones [Page 10, second column – “Here the outputs of the network categorized each histogram into either crack growth or noise present. The datasets determined to be crack growth and not noise were then used in the severity network. This network then determined the size of the increment of crack growth over the time window. This experiment used two separate sensors. The data from each sensor were separated and run through the two neural networks. Figure 13(a) contains graphs of the results of the networks in terms of crack length. As time increased in the experiment, the total crack length increased. Finally an average of the two signals was taken to determine a net crack length value. This average crack growth length is illustrated again in Figure 13(b) along with the load history.”Page 1, second column – “The basic acoustic emission system was augmented with an artificial neural network analysis to provide near real-time analysis of acoustic emission data measured from aircraft structural components, during routine service operations.”].
Regarding Claim 5, Kral discloses that the one or more processors are further configured so that the machine-learned Al-enabled technology neural network architecture model predicts damage level separately for each of four respective different damage zones [See Fig. 13, the discrete determined crack lengths].
Regarding Claim 6, Kral discloses that the monitored structure is equipped with a plurality of AE sensors configured for remote sensing [See Fig. 7].
Regarding Claim 7, Kral fails to disclose that the machine-learned Al-enabled technology neural network architecture model comprises a GoogLeNet convolutional neural network (CNN).
However, Szegedy discloses the use of a GoogLeNet convolutional neural network (CNN) as a high performing classifier [See Abstract.See Fig. 3 – “GoogLeNet network with all the bells and whistles”See Table 2.]. It would have been obvious to use such a neural network as the neural network because it was known to have good performance.
Regarding Claim 10, Kral fails to disclose determining maintenance activities for the monitored structure based on determined damage zone level predictions for the determined respective different damage zones of the associated monitored structure.
However, Kral contemplates the structural health monitoring system could be used in the scheduling of maintenance [Page 1, first column – “Maintenance cost might be reduced since an SHM system could be embedded into the aircraft structure, thereby reducing or eliminating the need to remove the aircraft from service to scan for damage during the ground inspection.”]. It would have been obvious to schedule maintenance in the event that a crack is indicated in order to perform an appropriate repair.
Regarding Claim 12, Kral fails to disclose that the machine-learned neural network architecture model is pre-trained using a subset of images from a preexisting database of images.
However, Szegedy discloses training a neural network in such a manner [Page 8 – “The ILSVRC 2014 classification challenge involves the task of classifying the image into one of 1000 leaf-node categories in the Imagenet hierarchy. There are about 1.2 million images for training, 50,000 for validation and 100,000 images for testing. Each image is associated with one ground truth category, and performance is measured based on the highest scoring classifier predictions.”]. It would have been obvious to use such a database (in this case containing images in the form of datasets per Fig. 9 of Kral) in order to facilitate appropriate training.
Regarding Claim 13, Kral discloses that the associated structure is continuously monitored in real-time, and the predictions for the determined respective different damage zones of the associated monitored structure are continuously produced in real-time [Page 10, second column – “Here the outputs of the network categorized each histogram into either crack growth or noise present. The datasets determined to be crack growth and not noise were then used in the severity network. This network then determined the size of the increment of crack growth over the time window. This experiment used two separate sensors. The data from each sensor were separated and run through the two neural networks. Figure 13(a) contains graphs of the results of the networks in terms of crack length. As time increased in the experiment, the total crack length increased. Finally an average of the two signals was taken to determine a net crack length value. This average crack growth length is illustrated again in Figure 13(b) along with the load history.”Page 1, second column – “The basic acoustic emission system was augmented with an artificial neural network analysis to provide near real-time analysis of acoustic emission data measured from aircraft structural components, during routine service operations.”].
Regarding Claim 14, Kral discloses that the decision tree-based rules are determined as a set of rules [Page 10, second column – “Histograms were made of ten bins each with the same range as before. This new dataset was first used in the yes-no network. Here the outputs of the network categorized each histogram into either crack growth or noise present. The datasets determined to be crack growth and not noise were then used in the severity network.”] that optimally fits the training data of the machine-learned neural network architecture model to four respective structural damage zones [See Fig. 13, the discrete determined crack lengths (there are considerably more than 4)].
Response to Arguments
Applicant argues:
PNG
media_image1.png
125
797
media_image1.png
Greyscale
Examiner’s Response:
The Examiner agrees and the corresponding rejections under 35 USC 112 are hereby withdrawn.
Applicant argues:
PNG
media_image2.png
261
774
media_image2.png
Greyscale
Examiner’s Response:
The Examiner respectfully disagrees. Xiang-jun discloses the use of the wavelet transform on vibration data in performing structure monitoring [Abstract – “Data processing is important to the structure health monitoring system which produces large volumes of raw data containing the useful information and the noise. Wavelet analysis is a newly emerging theory in data processing field, which has good localization characteristics in both frequency and time domains compared to most traditional methods used for structural health monitoring. Wavelet can be used for discovering the local feature of a signal by selecting a proper basic wavelet. In addition, the feature components of a signal can be obtained by reconstructing the wavelet coefficients. Wavelet technique is adopted to process the vibration signals acquired from the bridge monitoring spot in this paper. Based on the wavelet analysis theory, an efficient signal processing approach to structure health monitoring has been developed. The results of analysis show that the method is not only feasible to signal de-noising, but also valuable and effective to detect the health status of bridge structure.”] through use of wavelet coefficient scaling [Page 672, first column – “The mother wavelet ψ(t) needs to satisfy certain admissibility condition in order to ensure existence of the inverse wavelet transform. The dilation parameter “a” and the translation parameter “b” are also referred as the scaling and shifting parameters respectively and play an important role in the wavelet analysis.”].
It would have been obvious to transform the AE data of Kral, prior to performing normalization [Page 8, second column of Kral – “A sliding time window for real-time monitoring was created that stepped through time at a step of 1.6 seconds. This procedure allowed for multiple readings from the same detection (see Figure 10). Within each time window, multiple detections could be observed. These were normalized into a histogram of the data within the time window, thus removing any time dependence. A histogram was made of 10 bins, grouping the values of the energy value of each strain wave between zero and a normalized maximum value of 100.”], using the wavelet transform and to use such data in monitoring a structure because doing so would have been useful in denoising the data and would have made the structural monitoring more accurate.
Applicant argues:
PNG
media_image3.png
35
747
media_image3.png
Greyscale
PNG
media_image4.png
165
756
media_image4.png
Greyscale
Examiner’s Response:
The Examiner agrees that Kral does not disclose utilizing 2D wavelet images or scaling wavelet coefficients. However, Kral does disclose performing data normalization [Page 8, second column – “A sliding time window for real-time monitoring was created that stepped through time at a step of 1.6 seconds. This procedure allowed for multiple readings from the same detection (see Figure 10). Within each time window, multiple detections could be observed. These were normalized into a histogram of the data within the time window, thus removing any time dependence. A histogram was made of 10 bins, grouping the values of the energy value of each strain wave between zero and a normalized maximum value of 100.”].
Applicant argues:
PNG
media_image5.png
305
777
media_image5.png
Greyscale
PNG
media_image6.png
29
777
media_image6.png
Greyscale
Examiner’s Response:
The Examiner respectfully disagrees. Xiang-jun discloses the use of wavelet coefficient scaling [Page 672, first column – “The mother wavelet ψ(t) needs to satisfy certain admissibility condition in order to ensure existence of the inverse wavelet transform. The dilation parameter “a” and the translation parameter “b” are also referred as the scaling and shifting parameters respectively and play an important role in the wavelet analysis.”] in producing 2D wavelet images [See Figs. 3-5].
Applicant argues:
PNG
media_image7.png
169
781
media_image7.png
Greyscale
Examiner’s Response:
Janeliukstis is not relied on as disclosing such.
Applicant argues:
PNG
media_image8.png
167
772
media_image8.png
Greyscale
Examiner’s Response:
Szegedy is not relied on as disclosing such.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Ahmad et al., Supervised Learning Methods for Modeling Concrete Compressive Strength Prediction at High Temperature, MDPI, 2021
Ahmed et al., Advancements in fiber-reinforced polymer composite materials damage detection methods: Towards achieving energy-efficient SHM systems, Elsevier, 2021
Alghamdi, Classifying High Strength Concrete Mix Design Methods Using Decision Trees, MDPI, 3.6.2022
Avendano et al., Application of Statistical Machine Learning Algorithms for Classification of Bridge Deformation Data Sets, IEEE, 2021
Ji et al., The Research of Acoustic Emission Techniques for Non Destructive Testing and Health Monitoring on Civil Engineering Structures, IEEE, 2007
Bianchi et al., Life-Cycle Assessment of Deteriorating RC Bridges Using Artificial Neural Networks, ASCE, 2.7.2022
Chao et al., Research on material loss based on multi-channel acoustic emission and LSTM, IEEE, 6.17.2022
de Oliveira et al., A New Structural Health Monitoring Strategy Based on PZT Sensors and Convolutional Neural Network, MDPI, 2018
Levy et al., Crack growth detection and estimation of depth by monitoring acoustic emission activity, ASME, 1994
Nguyen et al., A Scheme with Acoustic Emission Hit Removal for the Remaining Useful Life Prediction of Concrete Structures, MDPI, 2021
Ullah et al., Nondestructive Inspection of Reinforced Concrete Utility Poles with ISOMAP and Random Forest, MDPI, 2018
US 20230013626 A1 – AUTOMATED MONITORING DIAGNOSTIC USING AUGMENTED STREAMING DECISION TREE
US 20210388950 A1 – SYSTEM AND METHOD FOR MECHANICAL FAILURE CLASSIFICATION, CONDITION ASSESSMENT AND REMEDIATION RECOMMENDATION
US 20210231515 A1 – MONITORING BOLT TIGHTNESS USING PERCUSSION AND MACHINE LEARNING
US 20170168024 A1 – MONITORING SYSTEMS AND METHODS FOR ELECTRICAL MACHINES
US 20140320298 A1 – Method And Apparatus For Detection Of Structural Failure
US 20080075352 A1 – DEFECT CLASSIFICATION METHOD AND APPARATUS, AND DEFECT INSPECTION APPARATUS
US 20030009300 A1 – In-situ Structural Health Monitoring, Diagnostics And Prognostics System Utilizing Thin Piezoelectric Sensors
US 20010047691 A1 – Hybrid Transient-parametric Method And System To Distinguish And Analyze Sources Of Acoustic Emission For Nondestructive Inspection And Structural Health Monitoring
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 KYLE ROBERT QUIGLEY whose telephone number is (313)446-4879. The examiner can normally be reached 9AM-5PM EST.
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, Arleen Vazquez can be reached at (571) 272-2619. 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.
/KYLE R QUIGLEY/Primary Examiner, Art Unit 2857
Prosecution Insights