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 .
Status of Claims
This communication is a first office action, non-final rejection on the merits. Claims 1-29 as filed, are currently pending and have been considered below.
Specification
The disclosure is objected to because of the following informalities:
0144: Typographic error: Suggest changing
“jointly quantize the positions and velocities all of the time steps of the trajectory” to “jointly quantize the positions and velocities of all the time steps of the trajectory”
Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 26-27 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 26 recites the limitation "the token prediction neural network " in line 1. There is insufficient antecedent basis for this limitation in the claim.
Dependent Claim 27 is also rejected under 112(b) as depending from Claim 26.
Examiner suggests changing to, and will further evaluate the claims in light of the art if as written as “the token processing neural network.”
Claim Rejections - 35 USC § 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-21, and 23-29 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Pronovost (US Patent application publication 20240101157 A1 hereinafter “Pronovost”).
Regarding Claim 1, Pronovost discloses
A method performed by one or more computers, the method comprising:
receiving input data that characterizes a driving environment, wherein the input data comprises a respective input for each of a plurality of data modalities characterizing the driving environment;
Pronovost pertains to techniques for predicting an object trajectory or scene information for a vehicle computing device for consideration during vehicle planning and discloses the use of a plurality of data inputs to characterize the driving environment by detailing “For example, a computing device can receive sensor data, log data, map data, and so on, as input and determine feature vectors representing an object, a vehicle, and/or an environment” (0025). Pronovost further discloses
generating an input multimodal token sequence of input tokens that represents the inputs for each of the plurality of data modalities; and
by describing the tokenization of input data (there characteristics) as “In this way, the first machine learned model can be trained, or conditioned, to output a token sequence, a set of tokens, or a cluster of tokens based at least in part on specific characteristics of the object, the environment, or a vehicle having a vehicle computing device to implement these techniques” (0023).
Pronovost further discloses
processing the input multimodal token sequence using a token processing neural network to generate an output token sequence representing a prediction about the driving environment
by describing the processing of the token sequence to make predictions (there trajectory) as “inputting the set of tokens into a second machine learned model; determining, by the second machine learned model and based at least in part on the set of tokens, an object trajectory for the object to follow in the environment; and causing the vehicle to be controlled in the environment based at least in part on the object trajectory” (0192). Pronovost further specifies that the machine learning model is a neural network by detailing “and the second machine learned model comprises at least one of a Generative Adversarial Network (GAN), a Graph Neural Network (GNN), a Convolutional Neural Network (CNN), a Recurrent Neural Network (RNN), or another transformer model” (0193).
Regarding Claim 2, Pronovost discloses all the limitations of claim 1 and further discloses in figure 3 wherein the input data (302) characterizes the driving environment for a vehicle within the driving environment specifying “input data 302 representing object trajectories associated with one or more objects, object state data, and scene data can be input into an encoder 304” (0062).
Regarding Claim 3, Pronovost discloses all the limitations of claim 2 and further discloses wherein the plurality of data modalities includes data obtained by each of one or more sensor types of the vehicle by specifying “In some examples, the vehicle computing device may detect the objects, based on sensor data received from one or more sensors. In some examples, the sensors may include sensors mounted on the vehicle 102, and include, without limitation, ultrasonic sensors, radar sensors, light detection and ranging (lidar) sensors, cameras, microphones, inertial sensors (e.g., inertial measurement units, accelerometers, gyros, etc.), global positioning satellite (GPS) sensors, and the like” (0040).
Regarding Claim 4, Pronovost discloses all the limitations of claim 3 and further discloses wherein the plurality of data modalities includes image data obtained by a camera of the vehicle by specifying “the sensor system(s) 806 may include lidar sensors, radar sensors, …, cameras (e.g., RGB, IR, intensity, depth, time of flight, etc.), … As another example, the camera sensors may include multiple cameras disposed at various locations about the exterior and/or interior of the vehicle 802” (0124).
Regarding Claim 5, Pronovost discloses all the limitations of claim 3 and further discloses wherein the plurality of data modalities includes point-cloud data obtained by a LIDAR sensor of the vehicle by specifying “In some instances, the localization component 820 may utilize SLAM (simultaneous localization and mapping), CLAMS (calibration, localization and mapping, simultaneously), relative SLAM, bundle adjustment, non-linear least squares optimization, or the like to receive image data, lidar data, radar data, IMU data, GPS data, wheel encoder data, and the like to accurately determine a location of the autonomous vehicle” (0105). It is understood by those of ordinary skill in the art of robotic control that SLAM via LIDAR comprises point-cloud data.
Regarding Claim 6, Pronovost discloses all the limitations of claim 3 and further discloses wherein the plurality of data modalities includes data obtained by a RADAR sensor of the vehicle by specifying “In at least one example, the sensor system(s) 806 may include lidar sensors, radar sensors, …. The sensor system(s) 806 may include multiple instances of each of these or other types of sensors. … The sensor system(s) 806 may provide input to the vehicle computing device 804” (0124).
Regarding Claim 7, Pronovost discloses all the limitations of claim 1 and further discloses wherein the plurality of data modalities includes data characterizing a road graph for the environment that characterizes roadways in the environment by detailing “For example, a computing device can receive sensor data, log data, map data, and so on, as input and determine feature vectors representing an object, a vehicle, and/or an environment” (0025). Pronovost further specifies that features of the environment includes roadways by stating “For instance, the vector representation 322 can comprise vectors to represent features of the environment including roadway boundary vectors 328 and roadway centerline vectors 330, among others” (0070).
Regarding Claim 8, Pronovost discloses all the limitations of claim 2 and further discloses wherein the plurality of data modalities includes structured navigational data generated by a navigation system of the vehicle. It is understood to those of ordinary skill in the art of robotic control that input data to a vehicle that is generated by the vehicle is feedback. Pronovost details navigational data feedback in figure 2 where the vehicle behavior token 202 (navigation data) is feed back as input into the transformer model 204.
Regarding Claim 9, Pronovost discloses all the limitations of claim 8 and further discloses wherein the structured navigational data generated by the navigation system of the vehicle comprises data characterizing states of one or more objects within the driving environment by detailing “a codebook 202 can store tokens that represent object behavior, vehicle behavior, and/or environment features. For example, a first token can represent an object state, a second token can represent a vehicle state, and a third token can represent environment features (e.g., a traffic signals, crosswalk, weather, etc.). The object state can indicate a position, orientation, velocity, acceleration, yaw, etc. of an object” (0049).
Regarding Claim 10, Pronovost discloses all the limitations of claim 9 and further discloses wherein the data characterizing the states of one or more objects comprises data generated based on sensor data obtained by one or more sensors of the vehicle. Pronovost refers to the state of objects in terms of the presence and proximity of an object, specifying “In some instances, the perception component 822 may include functionality to perform object detection, segmentation, and/or classification. In some examples, the perception component 822 may provide processed sensor data that indicates a presence of an object (e.g., entity) that is proximate to the vehicle 802 and/or a classification of the object as an object type (e.g., car, pedestrian, cyclist, animal, building, tree, road surface, curb, sidewalk, unknown, etc.)” (0106).
Regarding Claim 11, Pronovost discloses all the limitations of claim 10 and further discloses wherein the data characterizing the states of one or more objects comprises object data generated by processing the sensor data obtained by one or more sensors of the vehicle using an object detection system. Pronovost refers to the object detection system as the perception component 822, specifying “the perception component 822 may include functionality to perform object detection, segmentation, and/or classification. In some examples, the perception component 822 may provide processed sensor data that indicates a presence of an object (e.g., entity) that is proximate to the vehicle 802” (0106).
Regarding Claim 12, Pronovost discloses all the limitations of claim 1 and further discloses wherein the plurality of data modalities includes text data by specifying “The diffusion model 1302 can receive condition data 1306 for use during different diffusion steps to condition the input data, as discussed herein. For example, the condition data 1306 can represent one or more of: a semantic label, text, an image, an object representation, an object behavior, a vehicle representation, historical information associated with an object and/or the vehicle, a scene label indicating a level of difficulty to associate with a simulation, an environment attribute, a control policy, or object interactions, just to name a few” (0166).
Regarding Claim 13, Pronovost discloses all the limitations of claim 12 and further discloses wherein:
the input data comprises a request to perform a prediction task; and
processing the input multimodal token sequence using the token processing neural network to generate the output token sequence representing the prediction about the driving environment comprises:
processing the input multimodal token sequence using the token processing neural network to generate the output token sequence representing a prediction about the driving environment for the driving task.
Pronovost details the request to perform a prediction about the driving environment for the driving task in figure 9. Data, such as a request, is inputted into the transformer model 204 at step 902. The output token sequence representing the prediction about the environment is generated at 906. This token sequence is used for the driving task in step 912.
Regarding Claim 14, Pronovost discloses all the limitations of claim 13 and further discloses wherein the input multimodal token sequence comprises one or more multimodal tokens specifying the request to perform the prediction task. Pronovost refers to the prediction task in terms of a simulation to generate a trajectory and discloses the tokenization of the request by detailing “receiving, by a transformer model, a request to generate a simulated environment that includes a vehicle and an object; accessing, by the transformer model and based at least in part on the request, tokens from a codebook, at least one token in the codebook representing a behavior of the object; arranging, by the transformer model, the tokens into a sequence of tokens; inputting the sequence of tokens into a machine learned model; generating, by the machine learned model, the simulated environment that includes an object trajectory for the object; and causing the vehicle to be controlled in a real-world environment based at least in part on the object trajectory” (0187).
Regarding Claim 15, Pronovost discloses all the limitations of claim 13 and further discloses wherein the request to perform the prediction task comprises a request to generate a description of the driving environment. Pronovost refers to the prediction task in terms of a simulation to generate a trajectory and discloses the output of the transformer model to be a description of the environment, specifying “The prediction component 104 can also or instead predict scene data that describes a simulated environment. For instance, the prediction component 104 can output one or more scenes usable in a simulation (also referred to as a scenario or estimated states) to determine a response by the vehicle 102 to a simulated object” (0042).
Regarding Claim 16, Pronovost discloses all the limitations of claim 13 and further discloses wherein the request to perform the prediction task comprises a request to generate a description of an attribute of the driving environment. Pronovost refers to the prediction task in terms of a simulation to generate a trajectory and discloses the output of the transformer model to be a description of an attribute or behavior of the environment, particularly objects in the environment, specifying “The output data 106 may also or instead be used to perform a simulation by setting up conditions (e.g., an intersection, a number of objects, a likelihood for the object to exhibit abnormal behavior, etc.) for use during the simulation” (0046).
Regarding Claim 17, Pronovost discloses all the limitations of claim 13 and further discloses wherein the request to perform the prediction task comprises a request to predict trajectories for one or more objects in the environment. Pronovost refers to the prediction task in terms of a simulation to generate a trajectory and discloses the output of the transformer model to be a predicted object trajectory, specifying “In some examples, the token sequence 206 can be input into a machine learned model 208 (e.g., a CNN, a GNN, etc.) configured to generate output data 210. In some examples, the output data 210 can include one or more: an object trajectory, a heatmap, or scene data. For instance, the output data 210 can include a scene 212 which can further include the object trajectory 116” (0056).
Regarding Claim 18, Pronovost discloses all the limitations of claim 13 and further discloses wherein the request to perform the prediction task comprises a request to generate a planned trajectory for the vehicle in the environment. Pronovost refers to the prediction task in terms of a simulation to generate a trajectory and discloses the output of the transformer model to be a planned vehicle trajectory, specifying “The output data 106 from the prediction component 104 can be used by a vehicle computing device in a variety of ways. For instance, information about the object trajectories and/or the scene data can be used by a planning component of the vehicle computing device to control the vehicle 102 in the environment 100 (e.g., determine a vehicle trajectory 118 and/or control a propulsion system, a braking system, or a steering system). (0046).
Regarding Claim 19, Pronovost discloses all the limitations of claim 13 and further discloses wherein the request to perform the prediction task comprises a request to predict sensor data for one or more sensors of the vehicle. Pronovost refers to the prediction task in terms of a simulation to generate a trajectory and discloses the output of the transformer model to be object state data, specifying “the second machine learned model can output data representing one or more of: an object trajectory, a heatmap showing a likelihood of occupancy by an object(s), object state data, or scene data usable in simulation, just to name a few” (0021). Pronovost details that the vehicle sensors record object state data by stating “The data may include sensor data, such as data regarding the objects detected in the environment 100” (0040).
Regarding Claim 20, Pronovost discloses all the limitations of claim 1 and further discloses
wherein generating an input multimodal token sequence of input tokens that represents the inputs for each of the plurality of data modalities comprises, for each of the plurality of data modalities:
generating a token sequence of input tokens that represents the input of the data modality
Pronovost details the tokenization of input data in figure 2, addressing the different modalities by specifying “As illustrated, a codebook 202 can store tokens that represent object behavior, vehicle behavior, and/or environment features. For example, a first token can represent an object state, a second token can represent a vehicle state, and a third token can represent environment features (e.g., a traffic signals, crosswalk, weather, etc.)” (0049).
Regarding Claim 21, Pronovost discloses all the limitations of claim 1 and further discloses
wherein generating the input multimodal token sequence of input tokens that represents the inputs for each of the plurality of data modalities comprises, for each input:
selecting one or more input tokens representing numerical values for the input
Pronovost details the tokenization of input data in figure 3, and refers to the values assigned to the tokens as hyperparameters, specifying “Over time, the token sequence 316 can be determined to include a number of tokens based at least in part on a hyperparameter indicating a length, a size, or a total number of tokens to include in the token sequence” (0065).
Regarding Claim 23, Pronovost discloses all the limitations of claim 21 and further discloses
wherein, for each of the inputs, selecting the one or more input tokens representing the numerical values for the input, comprises:
quantizing each of the numerical values for the input; and
selecting one or more input tokens representing the quantized numerical values for the input
Pronovost details the quantization of input data in step 310 of figure 3, and details the selecting of tokens by stating “Feature vectors output from a machine learned model (e.g., the GNN or other model) can be input into a quantizer to generate discretized feature vectors for sending to the codebook (or other model or component of the computing device). In various examples, the codebook can receive the discretized feature vectors output by the quantizer and determine which token(s) of available tokens in the codebook are similar (e.g., information or characteristics associated with a feature vector (or object associated therewith) is similar to information or characteristics associated with a token (or object associated therewith), etc.)” (0026).
Regarding Claim 24, Pronovost discloses all the limitations of claim 23 and further discloses
wherein, for one or more of the inputs, quantizing each of the numerical values for the input comprises jointly quantizing a plurality of the numerical values for the input. Pronovost refers to a plurality of inputs as feature vectors, and details the quantization of feature vectors by stating “Feature vectors output from a machine learned model (e.g., the GNN or other model) can be input into a quantizer to generate discretized feature vectors for sending to the codebook (or other model or component of the computing device)” (0026).
Regarding Claim 25, Pronovost discloses all the limitations of claim 21 and further discloses
wherein, for one or more of the inputs, the one or more input tokens representing the numerical values for the input comprise input tokens characterizing text representations of the numerical values. It is understood by those of ordinary skill in the art of robotic control that the tokenization of a numerical value entails characterizing text representations of the number. Additionally, Pronovost details the text representation of numerical values in figure 13 by stating “FIG. 13 illustrates an example block diagram 1300 of an example diffusion model implemented by a computing device to generate latent variable data. For example, a computing device (e.g., the vehicle computing device(s) 804) can implement the diffusion model 1102 to generate the condition data 1106 for use by a machine learned model such as the variable autoencoder 1110 of FIG. 11” (0165), whereas “the condition data 1306 can represent one or more of: a semantic label, text, an image, an object representation, an object behavior, a vehicle representation (0166).
Regarding Claim 26, Pronovost discloses all the limitations of claim 1 and further discloses
wherein the token prediction neural network has been trained using a machine learning technique to generate predictions about the driving environment, the training comprising:
Pronovost refers to the last phase of the token processing neural network as the decoder 318, which is a machine learned Generative Adversarial Network (GAN), specifying “In some examples, the GAN 402 can be implemented for training the codebook 308, the machine learned model 314, and/or the decoder 318” (0076). Pronovost further discloses
obtaining a plurality of training examples, wherein each training example is associated with a processing task for the training example and comprises (i) example input data that characterizes a driving environment for the training example and (ii) a target prediction about the driving environment for the prediction task for the training example; and
Pronovost refers to the driving environment as vehicle state data and map data, and refers to the prediction task as the associated prediction, inference, or classification value, detailing “In some instances, the training component 848 may be executed by the processor(s) 836 to train a machine learning model based on training data. The training data may include a wide variety of data, such as sensor data, audio data, image data, map data, inertia data, vehicle state data, historical data (log data), or a combination thereof, that is associated with a value (e.g., a desired classification, inference, prediction, etc.)” (0135). Pronovost further discloses
updating the token prediction neural network to optimize an objective function that measures an error between the target predictions for the training examples and corresponding predictions generated for the training examples using the token prediction neural network
Pronovost refers to the error between the training examples and the token processing neural network in terms of the difference between ground truth and the discretized feature vectors, specifying “In some examples, the training component 702 can train the quantizer 310 by comparing the discretized feature vectors 312 with ground truth data and adjust parameters associated with the quantizer 310 to minimize a rounding error between the feature vectors of the ground truth and the discretized feature vectors 312. In some examples, the training component 702 can determine a setting for a lambda hyperparameter, or other parameter, to balance the errors associated with training different entities (e.g., determine a rate to improve a first loss of a first component relative to also improving a second loss of a second component that is interdependent on the first loss)” (0097).
Regarding Claim 27, Pronovost discloses all the limitations of claim 26 and further discloses
wherein the plurality of training examples comprises, for each of a plurality of processing tasks, one or more training examples associated with the processing task It is understood by those of ordinary skill in the art of robotic control that in order to train a processing task, the training example is associated with the task. Additionally, Pronovost details that the training examples are associated with the processing task by stating ”In some instances, the training component 848 can include functionality to train a machine learning model to output classification values. For example, the training component 848 can receive data that represents labelled collision data (e.g. publicly available data, sensor data, and/or a combination thereof). At least a portion of the data can be used as an input to train the machine learning model. Thus, by providing data where the vehicle traverses an environment, the training component 848 can be trained to output occluded value(s) associated with objects and/or occluded region(s), as discussed herein” (0136).
Regarding Claim 28, Pronovost discloses in figure 8
A system (800) comprising:
one or more computers (804); and
one or more storage devices (818) communicatively coupled to the one or more computers, wherein the one or more storage devices store instructions that, when executed by the one or more computers, cause the one or more computers to perform operations comprising:
Pronovost details “The vehicle computing device 804 may include one or more processors 816 and memory 818 communicatively coupled with the one or more processors 816” (0102), further specifying “The memory 818 and memory 838 may store an operating system and one or more software applications, instructions, programs, and/or data to implement the methods described herein and the functions attributed to the various systems” (0139). The remainder of claim 28 is rejected in a similar manner as claim 1.
Regarding Claim 29, Pronovost discloses One or more non-transitory computer storage media storing instructions that when executed by one or more computers cause the one or more computers to perform operations comprising: by detailing “Memory 818 and memory 838 are examples of non-transitory computer-readable media. The memory 818 and memory 838 may store an operating system and one or more software applications, instructions, programs, and/or data to implement the methods described herein and the functions attributed to the various systems” (0139). The remainder of claim 29 is rejected in a similar manner as claim 1.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim 22 is rejected under 35 U.S.C. 103 as being unpatentable over Pronovost in view of Ankit et al (French Patent Application FR3152908 A1 hereinafter “Ankit”).
Regarding Claim 22, Pronovost discloses all the limitations of claim 21, and Pronovost further discloses wherein selecting one or more input tokens representing numerical values for the input comprises selecting the one or more input tokens representing numerical values for the input using encoding is known to those of ordinary skill in the art of robotic control. As addressed with the rejection of claim 25 above, Pronovost teaches the tokenization of the numerical input values, but is silent on the specific method used. However, Ankit teaches the use of byte pair encoding. Ankit pertains to a method and apparatus for processing images of an environment of a vehicle and details the use of byte pair encoding by stating “A transformer model uses an encoder/decoder architecture. The encoder includes encoding layers that iteratively process input tokens from one layer to the next. The input tokens are obtained by parsing the input data (e.g., speech data) using a byte-pair encoding jetter reader, and each token can be converted into a vector by looking up its value in an integration table” (0040). Therefore, it would have been known to one of ordinary skill in the art of robotic control to use byte pair encoding to tokenize input data as taught by Ankit to provide the details of the specific method of input data tokenization as taught by Pronovost.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Nathan Daniel Neckel whose telephone number is (571)272-9537. The examiner can normally be reached M-F, 7-3.
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, Wade Miles can be reached at 571-270-7777. 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.
/NATHAN DANIEL NECKEL/Examiner, Art Unit 3656
/WADE MILES/Supervisory Patent Examiner, Art Unit 3656
Prosecution Insights