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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Claims 3-7, 10-20 are cancelled. Claims 1, 2, 8, 9, 21-36 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, 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, 2, 8, 9, and 21-36 are rejected under 35 U.S.C. 103 as being unpatentable over Turkelson et al. (20210004589 A1), in view of HALATA (DE 102013002554 A), in further view of Voegele et al. (US 20220332504 A1).
Re-claim 1, Turkelson et al. teach a system for [multi-]object detection with improved accuracy, the system comprising:
--an autonomous robot [configured to migrate throughout a storage area in a retail facility;]
(see e.g. [0110] In some embodiments, the object recognition model and context classification model may form a loop for dynamically analyzing captured video or images in real-time, and making adjustments based on the continuously evolving analysis. For example, a mobile robot, autonomous vehicle, drone, mobile manipulator, assistive robots, and the like, may ingest video or images in real-time, determine a context of the image (e.g., a scene), determine objects within the image based on the determined context and the image, and then return the determined object and initially determined context to update, if necessary, the context. Furthermore, the continuous real-time analysis of the object(s) within the image and the context(s) of the image may further refine the sub-class of object and sub-class of context of the image.)
***Note: Turkelson anticipate the robot migrate throughout a storage area in a retail facility based on the following teaching:
[0010] the image is captured by a mobile computing device
[0052] In some embodiments, a context classification model, such as a scene classification model, may be a unified ontology of retail, where retail can encompass various items, products, or services that are searchable and capable of being obtained (e.g., purchased).
[0108] For example, the geographical location of an image may be used to determine an approximate location of the user within a retail store (e.g., the image was determined to be taken in a Sporting Goods section of a store).
-an image capture device integrated into the autonomous robot, the image capture device configured to capture an image at a recognized location within [[a]] the storage area of the retail facility; and
(see e.g. [0110] In some embodiments, the object recognition model and context classification model may form a loop for dynamically analyzing captured video or images in real-time, and making adjustments based on the continuously evolving analysis. For example, a mobile robot, autonomous vehicle, drone, mobile manipulator, assistive robots, and the like, may ingest video or images in real-time, determine a context of the image (e.g., a scene), determine objects within the image based on the determined context and the image, and then return the determined object and initially determined context to update, if necessary, the context. Furthermore, the continuous real-time analysis of the object(s) within the image and the context(s) of the image may further refine the sub-class of object and sub-class of context of the image.
[0236] Each instance of mobile computing devices 104 may include an image capturing component, such as a camera, however some instances of mobile computing devices 104 may be communicatively coupled to an image capturing component. For example, a mobile computing device 104 may be wirelessly connected (e.g., via a Bluetooth connection) to a camera, and images captured by the camera may be viewable, stored, edited, shared, or a combination thereof, on mobile computing device 104.)
-a computer-readable medium storing programming instructions for execution by one or more processors, the programming instructions, upon execution by the one or more processors, causing the system to perform the following operations :
-analyzing, by a [ multi-]object detection model, the image to identify objects of interest within the image, (see e.g. [0110] In some embodiments, the object recognition model and context classification model may form a loop for dynamically analyzing captured video or images in real-time, and making adjustments based on the continuously evolving analysis.
[0199] For example, the images from image set 302B may be analyzed using a pre-trained object recognition model (e.g., AlexNet, GoogLeNet, MobileNet v2, etc.), and features may be extracted from each image.
[0103] In some embodiments, the object recognition model may be trained using a training data set including a set of images depicting different objects, either of a same type (e.g., all depicting dogs) or of different types (e.g., some depicting dogs, some depicting cats, some depicting houses, etc.)
[0137] In some embodiments, the plurality of images included within the training data set used to train the object recognition model may each be labeled with an object identifier of an object depicted within that image. For example, an image depicting a drill may be labeled with an object identifier of the drill for performing supervised learning. In some embodiments, each object identifier may correspond to an object from an object ontology. The object ontology may include a plurality of objects, which may differ from one another or which may be similarly. For example, the object ontology may include images depicting a plurality of different objects, such as drills, baseballs, coats, etc.)
--the [multi-]object detection model being trained to classify the objects of interest as object types
(see e.g. [0103] . Each object recognition model may be trained for a particular object (e.g., a specific object recognition model, such as an object recognition model configured to recognize dogs, logos, hardware, etc., within an image) or trained for general object recognition (e.g., capable of recognizing various different objects).
[0200] upon receipt of a new image to be analyzed, the trained computer-vision object recognition model may be retrieved and used to classify and locate objects that may be depicted within the new image.)
-classifying, by the [multi-]object detection model, the pallet as the pallet-related object type, the pallet storage structure as the location-related object type, and the void as the space- related object type and
(see e.g. [0174] As an example, an R-CNN may take each input image, extract region proposals, and compute features for each proposed region using a CNN. The features of each region may then be classified using a class-specific SVM, identifying the location of any objects within an image, as well as classifying those images to a class of objects. )
-labeling the image with indicators identifying the pallet as the pallet-related object type, the pallet storage structure as the location-related object type, and the void as the space-related object type; and
(see e.g. [0143] a determination may be made that the first object (e.g., first object 702A) depicted in the image corresponds to an object from the object ontology labeled with a first object identifier, while the second object (e.g., second object 704A) depicted in the image corresponds to another object from the object ontology labeled with a second object identifier.
[0047] In some cases, the object recognition model may be trained on a training data set in which both objects depicted are labeled and scenes are labeled according to the context (e.g., scene) ontology or taxonomy, such that the object recognition model is responsive to both pixel values and context classifications when recognizing objects.
[0174] As an example, an R-CNN may take each input image, extract region proposals, and compute features for each proposed region using a CNN. The features of each region may then be classified using a class-specific SVM, identifying the location of any objects within an image, as well as classifying those images to a class of objects.
-wherein the indicators include a first bounding box encircling the pallet
[0298] In some embodiments, mobile computing device 104C may include a mini-classifier configured to generate and display a bounding box 404C surrounding any object detected within a displayed image. In some embodiments, bounding box 404C may be displayed on display screen 400C regardless of whether a candidate video or image is being captured.)
--presenting the item-to-location mapping table to a user via a user interface device
(see e.g. [0297] FIG. 19 illustrates an example user interface of an image-capture task displayed on a mobile computing device, in accordance with various embodiments
[0298] In some embodiments, mobile computing device 104C may include a mini-classifier configured to generate and display a bounding box 404C surrounding any object detected within a displayed image. In some embodiments, bounding box 404C may be displayed on display screen 400C regardless of whether a candidate video or image is being captured.)
Turkelson et al. do not teach the following limitations.
However, HALATA teaches the multi-object model (see e.g. support vector machine).
--analyzing, by a multi-object detection model, the image to identify objects of interest within the image, the objects of interest including a pallet, a pallet storage structure housing the pallet, and a void on the pallet storage structure,
(see e.g. In one embodiment, a support vector machine is used to classify. A Support Vector Machine is a computer program used to classify objects based on a machine learning process. Each object to be classified is represented by a vector in a vector space. In the invention, the objects to be classified are the segments, each of which is described by a vector. The vector contains all segment attributes assigned to the relevant segment. To "teach in" the Support Vector Machine training objects are used that belong to given classes. For this purpose, for example, in a data record acquired with the 3D camera, the known objects can be manually assigned to the intended classes. Based on such training datasets, the Support Vector Machine learns the correct classification of the segments. The result is a reliable classification with relatively little computational effort and a continuous optimization of the classes used, without the need for extensive theoretical considerations to the individual segment attributes of the intended classes would have to be made.
--In one embodiment, one or more of the following classes are taken into account when classifying the segments: vertical shelf supports, horizontal shelf supports, signs, transport boxes, walls, pallets. In particular, all of these classes can be taken into account and thus the entirety of the objects present in a typical warehouse can be largely completely classified.)
--the multi-object detection model being trained to classify the objects of interest as object types that include a pallet-related object type including pallet bases, a location-related object type including structural bars for supporting the pallet bases, and a space-related object type including empty spaces for the pallet bases;
(see e.g. Subsequently, segment attributes are calculated for each segment, which then form the basis for a classification of the segments. The recognition of an object of the warehouse then takes place on the basis of the classification.
In one embodiment, a support vector machine is used to classify. A Support Vector Machine is a computer program used to classify objects based on a machine learning process. Each object to be classified is represented by a vector in a vector space. In the invention, the objects to be classified are the segments, each of which is described by a vector. The vector contains all segment attributes assigned to the relevant segment. To "teach in" the Support Vector Machine training objects are used that belong to given classes. For this purpose, for example, in a data record acquired with the 3D camera, the known objects can be manually assigned to the intended classes. Based on such training datasets, the Support Vector Machine learns the correct classification of the segments. The result is a reliable classification with relatively little computational effort and a continuous optimization of the classes used, without the need for extensive theoretical considerations to the individual segment attributes of the intended classes would have to be made.)
HALATA also teaches---classifying, by the multi-object detection model, the pallet as the pallet-related object type, the pallet storage structure as the location-related object type, and the void as the space- related object type and -labeling the image with indicators identifying the pallet as the pallet-related object type, the pallet storage structure as the location-related object type, and the void as the space-related object type; and
(see e.g. In one embodiment, a segment is recognized as a palette if the following criteria are met: • the segment has been classified as a pallet with a minimum probability and / or A midpoint of the segment is located within a predetermined height range above a horizontal shelf support and within a predetermined range of distances behind its leading edge and / or • a comparison of the front of the segment with a predefined pattern of a palette gives a minimum match.
After calculating the segment attributes for each segment, the segments are classified using a Support Vector Machine. The result of this classification is in the 16 in which the identified classes are illustrated by different gray levels. The darkest grayscale 42 corresponds to the class of vertical shelf supports. The classified as such segments form in the 16 a vertical strip 56 ,
The second darkest grayscale 44 illustrates the class of horizontal shelf supports. The classified as such segments are located in two in the 16 (taking into account the perspective view) horizontally extending stripes 58 ,
The third darkest grayscale 46 corresponds to the class sign. As such, segments recognized 60 are above all above the lower horizontal shelf support and right of the vertical shelf support. You are at the front of other segments 62 , which have been classified as transport boxes, represented by the next lighter gray level 48 ,
Another, in the 16 shown gray scale 50 corresponds to the class wall surface. Corresponding segments 64 are particularly recognizable within the trained between the individual shelf supports shelves.
An even lighter shade of gray 52 stands for the class pallets and is found in the 16 predominantly below the segments classified as transport boxes 62 , The legend includes another grayscale 54 , which stands for a class not further explained here.)
The Examiner notes that Turkelson et al., in view of HALATA teach the claimed limitations including “a multi-object detection model identifying objects of interest. The object of interest claimed (pallet, a pallet storage structure housing the pallet, and a void) are considered obvious variation of HALATA since HALATA teaches identifying different types of segments in an image and classifying the segments as (i.e.) pallets, free storage spaces, wall surface.
Under KSR, a claim would have been obvious if the claimed elements were known in the prior art and one skilled in the art could have substituted one known element for another to obtain predictable results with no change in their respective functions, and the substitution would have yielded nothing more than predictable results to one of ordinary skill in the art at the time of the invention.
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Turkelson et al., and include the steps cited above, as taught by HALATA, in order to detect an object of the warehouse based on the classification.
Turkelson et al., in view of HALATA, do not teach the following limitations.
However, Voegele et al. teach updating an item-to-location mapping table based on the labeled image by analyzing a portion of the image within the first bounding box to identify a pallet tag associated with the pallet, performing optical character recognition (OCR) on the pallet tag to identify a pallet identifier (ID), identifying a set of items associated with the pallet ID, and mapping the set of items to the recognized location within the item-to-location mapping table; and presenting the item-to-location mapping table labeled image to a user via a user interface device.
(see e.g. [0178] The image analysis module 1004 can be configured to receive image data that is captured by the cameras 1016. The module 1004 can perform image analysis techniques on such image data to identify information about the pallet. As described herein, during a profiling process, the module 1004 can at least detect a product label or other unique identifier from the image data. The module 1004 can be trained using one or more machine learning models to identify different types of labels from the image data. The module 1004 can also be configured to perform object detection techniques and optical character recognition (OCR) to not only detect the label but also to determine information associated with the label. Using similar techniques, the module 1004 can determine text on the pallet and any other identifying characteristics about the pallet. For example, as described herein, the module 1004 can use machine learning trained models to detect different features of the pallet from the image data.
-- [0008] The computing system can receive image data from cameras and, using image analysis techniques and/or machine learning models, identify information about the pallet. For example, the computing system can identify one or more labels or other unique identifiers associated with the pallet and/or its component parts (i.e., cases on pallet). In another example, the computing system can determine the dimensions of the pallet (e.g., height, width, depth), types of items included in the pallet, quantities of those items, information on those specific items (e.g., date, origin, expiration), and/or other details.
[0010] In some implementations, instead of sending the pallet to a temporary storage location, additional images of the pallet can be captured (e.g., rotating the turntable again and taking images at different orientations) and analyzed. The cameras can capture images of the pallet, which can be processed by the computing system to identify the pallet label. If the label is unidentifiable, the pallet can be routed to the off conveyor belt for temporary storage. If the label is identified, the pallet can be routed to the off conveyor belt for organized storage.
[0021] The notification can cause the warehouse management system to route the pallet into different storage locations based on identification of the unique identifier.
[0165] The WMS 154 can receive the pallet information in 910. The WMS 154 can determine an optimal storage location for the pallet based on the pallet information (912). The pallet information can include owner, item type/name, and the label. The WMS 154 can also determine the optimal organized storage location based on other information. As an example, the computing system 152 can identify the pallet by its owner by analyzing the clear capture of the label. The owner name/identifier can be communicated to the WMS 154. The WMS 154 can look up the owner in a data store that can store information about warehouse customers. The stored information about the owner can indicate where pallets for that owner are stored and/or what storage conditions must be satisfied. Using such information, the WMS 154 can determine what organized storage location the pallet should be routed to.
Voegele et al. also teach --presenting the item-to-location mapping table to a user via a user interface device.
(see e.g. [0159] F or example, as described herein, the computing system 152 and/or the WMS 154 can provide the image data of the pallet to a user device (e.g., the user device 156) of the warehouse worker. The user device can output the image data at a graphical user interface (GUI) display.
[0160] The identified information can then be stored, by the computing system 152 and/or the WMS 154 in association with the placeholder label. The placeholder label, therefore, can become the unique identifier of the pallet in the warehouse. Information collected or determined about the pallet can be linked or otherwise associated to the placeholder label.)
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the bounding box of Turkelson et al., with a tag, in view of HALATA, and include the steps cited above, as taught by Voegele et al., in order to route the pallet into different storage locations based on identification of the unique identifier (see e.g. [0012]).
Re-claim 2, Turkelson et al. teach the system of claim 1, wherein the instructions are further operative to: train the multi-object detection model using labeled training data comprising labeled objects of interest associated with the plurality of object types.
(see e.g. [0200] After providing images 314B and 324B to computer-vision object recognition model 310B, feature vectors 332B and 334B may be obtained. Furthermore, providing images 314B and 324B to computer-vision object recognition model 310B may cause computer-vision object recognition model 310B to be trained to recognize objects within images. A trained instance of computer-vision object recognition model 310B may be stored in model database 136B, and upon receipt of a new image to be analyzed, the trained computer-vision object recognition model may be retrieved and used to classify and locate objects that may be depicted within the new image.
[0239] In some embodiments, image-capture task subsystem 112C determines an object or a set of objects that an image-capture task will be directed towards. To determine the objects, image-capture task subsystem 112C may access training data database 138C. Training data database 138C may include training data sets, where each training data set is associated to a particular object or category of objects with which an object recognition model uses or will use to train that object recognition model for recognizing the object within an input image. For example, training data database 138C may include a training data set including a plurality of images depicting a table, and this training data set may be used to train an object recognition model to recognize whether an input image depicts a table.)
(see e.g. [0293] At step 308C, a number of images related to a different object included by the training data that the object recognition model is to be trained to recognize is determined. Process 300C may then return to step 306C to determine if the number of images of the different object is less than the threshold value. In some embodiments, if the object recognition model is specific and is only used to recognize one object or one type of object, then process 300C may end after step 306C. However, if the object recognition model is generic, capable of recognizing at least two different objects or two different types of objects, then process 300C may proceed to step 308C.)
Claims 8, 31 recite similar limitations as claim 1 and are therefore rejected under the same arts and rationale.
Claims 9, 32 recite similar limitations as claim 2 and are therefore rejected under the same arts and rationale.
Re-claim 21,Turkelson et al. clearly anticipate the limitation ( see e.g. [0169], [0143], [0047], [0102], [0106].
Turkelson et al., in view of HALATA, do not explicitly teach the following limitations.
However, Voegele et al. teach ---the method of claim 8, wherein the indicators further include a text-based label identifying at least one of the pallet as the pallet-related object type, the pallet storage structure as the location-related object type, or the void as the space-related object type.
(see e.g. [0178] Using similar techniques, the module 1004 can determine text on the pallet and any other identifying characteristics about the pallet. For example, as described herein, the module 1004 can use machine learning trained models to detect different features of the pallet from the image data. One or more of such models can be used for detecting damage to contents, lean, height, weight, dimensions, contents, damage to the pallet, temperature of contents, text, etc.
[0165] The WMS 154 can receive the pallet information in 910. The WMS 154 can determine an optimal storage location for the pallet based on the pallet information (912). The pallet information can include owner, item type/name, and the label. The WMS 154 can also determine the optimal organized storage location based on other information. As an example, the computing system 152 can identify the pallet by its owner by analyzing the clear capture of the label.)
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Turkelson et al., in view of HALATA, and include the steps cited above, as taught by Voegele et al., in order to identify more information about the pallet. (see e.g. [00178]).
Re-claim 22, Turkelson et al. teach the method of claim 8, wherein the indicators further include color-coded indicators such that the pallet-related object type is identified with a different color indicator than the location- related object type.
(see e.g. [0250] For instance, one or more filters may be applied to a frame to increase clarity, a region of interest may be identified and a bounding box overlaid on the image representing the region of interest, color enhancement, noise removal, de-blurring, or any other image enhancement technique, or any combination thereof.
[0067] In some embodiments, after placing the bounding box around a detected object, computational enhancement techniques may be applied to improve the quality of the portion of the image including the bounding box (e.g., contours, color schemes).
Re-claim 23, Turkelson et al. teach the method of claim 8. With respect to “a second bounding box encircling the void. “ Turkelson teaches bounding boxes for each object. Therefore, a second bounding box is anticipated.
Turkelson et al., do not teach the void. However, HALATA teaches --After calculating the segment attributes for each segment, the segments are classified using a Support Vector Machine. The result of this classification is in the 16 in which the identified classes are illustrated by different gray levels
--In one embodiment, a free storage space is recognized if no pixels or segments are present in a defined spatial area above an object that was recognized or classified as a horizontal shelf carrier.)
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Turkelson et al., and include the steps cited above, as taught by HALATA, so that the entirety of the objects present in a typical warehouse can be largely completely classified.
Re-claim 24, Turkelson et al. teach the method of claim 23, wherein the first bounding box and the second bounding box are color-coded based on colors that identify the pallet-related object type and the space- related object type, respectively.
(see e.g. [0250] For instance, one or more filters may be applied to a frame to increase clarity, a region of interest may be identified and a bounding box overlaid on the image representing the region of interest, color enhancement, noise removal, de-blurring, or any other image enhancement technique, or any combination thereof.
[0067] In some embodiments, after placing the bounding box around a detected object, computational enhancement techniques may be applied to improve the quality of the portion of the image including the bounding box (e.g., contours, color schemes).
Re-claim 25, Turkelson et al., in view of HALATA, do not teach the following limitations.
However, Voegele et al. teach -- The method of claim 8, further comprising: generating an item-to-location mapping table that maps the pallet to the recognized location, wherein the item-to-location mapping table is presented to the user via the user interface device.
(see e.g. [0165] The WMS 154 can receive the pallet information in 910. The WMS 154 can determine an optimal storage location for the pallet based on the pallet information (912). The pallet information can include owner, item type/name, and the label. The WMS 154 can also determine the optimal organized storage location based on other information.
[0159] F or example, as described herein, the computing system 152 and/or the WMS 154 can provide the image data of the pallet to a user device (e.g., the user device 156) of the warehouse worker. The user device can output the image data at a graphical user interface (GUI) display.)
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Turkelson et al., in view of HALATA, and include the steps cited above, as taught by Voegele et al., in order to determine a storage location (e.g., organized storage location) for the pallet (see e.g. [00169]).
Claims 26, 33 recite similar limitations as claim 21 and are therefore rejected under the same arts and rationale.
Claims 27, 34 recite similar limitations as claim 22 and are therefore rejected under the same arts and rationale.
Claims 28, 35 recite similar limitations as claim 23 and are therefore rejected under the same arts and rationale.
Claims 29, 36 recite similar limitations as claim 24 and are therefore rejected under the same arts and rationale.
Claim 30 recites similar limitations as claim 25 and is therefore rejected under the same arts and rationale.
Response to Arguments
Applicant’s arguments with respect to the office action have been considered but are moot.
Applicant’s remark:
Turkelson and Ganapathi do not disclose the above-recited features of amended claim 1 because Turkelson and Ganapathi are silent on labeling an image with a bounding box that encircles a pallet (depicted in the image). By extension, neither Turkelson nor Ganapathi suggests analyzing a portion of the labeled image within the bounding box (i.e., identifying a pallet tag, performing OCR on the pallet tag, etc.) for the purpose of updating an item-to- location mapping table (e.g., to map items associated with the pallet ID to a location of the pallet).
Examiner’s response:
The argument regarding Ganapathi is now moot since the reference is no longer relied upon to teach the limitations as claimed.
However, Turkelson et al. teach the bounding box that encircles a pallet, and Voegele et al. teach analyzing labeled image on a pallet by performing OCR for the purpose of properly identifying a storage location for the pallet and displaying the information on a user interface.
The Examiner notes that the storage location is then stored in the system (updating an item-to-location in the system).
Conclusion
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.
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, Florian Zeender can be reached at 571 272-6790. 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.
/LUNA CHAMPAGNE/Primary Examiner, Art Unit 3627 April 17, 2026
Prosecution Insights