DETAILED ACTION
Claims 1-21 are presented for examination. Claims 8, 9, 12, and 13 stand currently amended.
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Finality of Office Action
The following is a brief summary description of new ground(s) of rejection (if any) and the reason why those new ground(s) are made necessary by this amendment:
No new grounds of rejection are presented herein.
Response to Arguments
Applicant's remarks filed 30 January 2026 have been fully considered and Examiner’s response is as follows:
Regarding §101:
The USPTO has issued a Memorandum “Advance notice of change to the MPEP in light of Ex Parte Desjardins” (5 December 2025) available from <https://www.uspto.gov/sites/default/files/documents/memo-desjardins.pdf> [herein “Desjardins Memo”].
Because Ex Parte Desjardins (precedential) relates to machine learning it is relevant for consideration of the instant claims which involve training a GNNM. Based on the updated office guidance in the Desjardins Memo, Examiner finds the instant claims are directed toward improvements in reservoir modeling and hydrocarbon extraction.
Applicant’s §101 remarks do not discuss the Desjardins Memo and are considered moot in light of the above withdrawal of the §101 rejection.
Regarding §102/103:
Applicant remarks page 22 argues:
Claim 1 recites, inter alia, "generating, based on the geologic reservoir information and the initial spatial pressure profile, a computational sector model for a selected sector comprising a plurality of modelling cells forming a grid or mesh." Maucec fails to teach or fairly suggest this feature of claim 1. Maucec proposes a "reservoir grid model', which serves as the structural basis for creating a reservoir graph network suitable for processing using neural networks (Column 2, lines 60-62). …. Thus, while the generation step appears to use geologic reservoir information like "the well log data, the core sample data [or] seismic data,” this data does not seem to comprise pressure data, let alone an "initial spatial pressure profile" as required by Claim 1. Instead, pressure data is used only later in connection with the "reservoir graph" (Block 610) and the "neural network" (Block 620). More specifically, "embodiments are contemplated that may include pressure data, pore volume data, connate water saturation data, oil-water relative permeability data, grid cell indices, etc. as inputs to the graph neural network" (Column 8, lines 27-31).
This argument is unpersuasive.
Maucec column 9 lines 1-4 discloses “Dynamic response data may include transmissibility values or similar data based on one or more full-physics reservoir simulations performed for the geological region of interest.” Maucec column 9 lines 12-15 discloses “The simulator device may use grid model data to determine dynamic response data as well as updated property data for multiple reservoir property realizations.” A full-physics reservoir simulation performed corresponds to generating a computational sector model. The grid model corresponds with a plurality of modelling cells to form a grid or mesh.
Maucec column 11 lines 26-27 disclose “The simulator device may also use dynamic data, such as fluid rates and pressure data.” See further Maucec figure 9 (“pressure data E” (925)). Accordingly, pressure data is input as a part of the grid model.
This is not to be confused with the subsequently “predicted pressure data Z (563)).” See Maucec column 8 line 23. Thus, Maucec includes teachings of both inputting pressure data (e.g. Maucec column 11 lines 26-27) and outputting different predicted pressure data (e.g. Maucec column 8 line 23). Thus, while Examiner agrees that Maucec teaches that “pressure data is used” later in connection with the reservoir graph as Applicant argues, Examiner notes this subsequent teaching is not mutually exclusive to also using different pressure data as a part of the input data as required by the instant claim limitation.
Furthermore, it is unclear how citing Maucec column 8 lines 27-31 helps Applicant’s argument. Maucec column 8 lines 27-31 as cited by Applicant above underlines the teachings showing that “pressure data” is explicitly considered “as inputs to the graph neural network.” Examiner finds that data which is used as an “input” corresponds with generating respective outputs based on the respective input, e.g. input pressure data.
Overall, claim 1 recites using pressure data in two places. First, claim 1 clause 2 “generating, based on … the initial spatial pressure profile, a computational sector model….” Second, claim 1 clause 4 “initializing the GNNM … using … and the initial spatial pressure profile.” In both recitations the claim terms “based on” and “using” allow for indirect usage of the respective pressure data. However, Examiner does not need to rely upon the breadth of reasonable claim interpretation to anticipate the claim here as Maucec column 11 lines 26-27 explicitly teach the simulator device “may also use … pressure data.” This corresponds with the first recitation of claim 1 clause 2. Applicant’s citation to Maucec column 8 lines 27-31 corresponds more closely to the second recitation of claim 1 clause 4. The fact that Maucec teaches pressure data of claim 1 clause 4 is not a teaching away from where Maucec also teaches claim 1 clause 2.
Applicant remarks page 22 further argues:
Further, "for network encoding of a reservoir graph network, a network encoding may be applied to grid cells with the following reservoir grid attributes: grid cells, oil volume (OV) water volume (WV). pressure (P) ... " (Column 13, lines 29-34). Thus, Maucec does not teach or fairly suggest that the generation of the reservoir grid model is based on geological information and pressure data.
This argument is unpersuasive.
Maucec column 13 lines 29-34 relate to the graph network and thus correspond more closely with claim 1 clause 4 than claim 1 clause 2. Furthermore, Examiner has cited Maucec column 11 lines 26-27 as clarifying that the simulator device itself also using pressure data as input. The fact that Maucec teaches pressure data of claim 1 clause 4 is not a teaching away from where Maucec also teaches claim 1 clause 2. Examiner further notes that the teaching of “reservoir grid attributes: … pressure (P)” teaches that pressure data is a part of a grid model. Thus, the cited teaching indicates the preceding existence of the pressure data as a part of “modelling cells forming a grid.” Compare with Claim 1 clause 2 “a plurality of modeling cells forming a grid or mesh.” This is further evidence that Maucec column 11 lines 26-27 teaching of including “pressure data” with the “simulator device” results in the pressure data being incorporated into the “grid model data” of Maucec column 9 lines 12-15.
Applicant remarks page 22 further argues:
Maucec does not disclose obtaining an initial spatial pressure profile as a distinct dataset corresponding to an initial time (e.g., t = 0), as required by claim 1. The claim language requires that the initial spatial pressure profile serve as a foundational input for initializing the GNNM. In contrast, Maucec treats pressure as one of several reservoir attributes that may be included during graph encoding or parameterization, without any disclosure that such pressure data corresponds to an initial reservoir state or is singled out for model initialization.
This argument is unpersuasive.
A “initial time” is a somewhat arbitrary temporal reference point. Any time preceding a subsequent time may be considered an ‘initial’ time before the subsequent time. Here, without loss of generality, the “acquired” pressure data of Maucec figure 9 is considered an ‘initial’ pressure profile. See further Maucec column 11 line 58 “acquired data (i.e., acquired pressure data (815))” and Maucec figure 8.
Applicant remarks page 23 further argues:
Maucec further fails to teach or fairly suggest "generating, using the computational sector model, a training set of simulated spatiotemporal pressure profiles for the selected sector, each spatiotemporal pressure profile being generated for a corresponding training injection production plan (TIPP)." Although Maucec performs full-physics reservoir simulations to generate synthetic dynamic data, including pressure responses over time, Maucec does not disclose a training set having a one-to-one correspondence between individual spatiotemporal pressure profiles and predefined injection-production plans for training purposes.
This argument is unpersuasive.
The claim does not recite a one-to-one correspondence. The claim merely recites “a corresponding” which encompasses correspondences other than one-to-one correspondence. Accordingly, the claim language as currently recited merely requires any correspondence.
Maucec column 13 lines 36-40 disclose “For encoding production wells, well oil production rate (WOPR) and well water production rate (WWPR) may be encoded to effect relation information. For injection wells, well water injection rate (WWIR) may be the encoded effect for the relation information.” The well water injection rate and well water production rates relation information correspond with associated training injection production plain information.
Accordingly, including the well water injection rate and water production rate relation information as a part of “network encoding of a reservoir graph network” (Maucec column 13 line 29) is including corresponding TIPP in the training. Maucec column 13 line 32 explicitly includes “pressure (P)” as well. Because Maucec column 13 lines 29-40 cites both pressure (P) and the respective TIPP of the injection and production rate information together as a part of the network encoding, Maucec is teaching training with pressure and corresponding TIPP. That is, being used together in the same network encoding process is a correspondence.
Applicant remarks page 23 further argues:
Moreover, Maucec's disclosure of training data generation is generic and does not include the specific sequence of steps recited in claim 1, namely: (i) generating a computational sector model based on geologic reservoir information and an initial spatial pressure profile, and (ii) generating, using that computational sector model, a training set of simulated spatiotemporal pressure profiles, each tied to a corresponding TIPP. Maucec merely discloses that training data may include acquired, synthetic, or augmented dynamic data, without teaching the claimed dependency between sector level modeling, initial pressure-based initialization, and injection-production-plan-specific training profiles.
Applicant appears to be arguing the overall arrangement of steps of claim 1. This argument is unpersuasive.
Maucec figure 9 shows:
PNG
media_image1.png
200
400
media_image1.png
Greyscale
The data generated by the reservoir simulation (960) clearly corresponds with the cited training set of simulated [data]. This data is clearly then input into the “graph neural network generation process” (990). Thus, Maucec figure 9 shows the overall relationship of how the simulation data is generated, and that these training sets of data are used for “training the GNNM using the training set” as cited by Examiner.
The detailed rejection below cites respective details of respective claim clauses while Maucec figure 9 shows the overall relationship corresponding with the arrangement of limitations of claim 1.
Applicant remarks page 24 restates the above arguments of claim 1 for independent claim 14. Examiner’s response is the same as above.
Applicant remarks page 25 further argues:
[N]either Maucec nor Chaki teaches or fairly suggests generalizing a sector-trained GNNM to a reservoir-wide GNNM based at least in part on a measured pressure profile of the hydrocarbon reservoir, as expressly required by claim 8. Maucec discloses generating and training a graph neural network for a region of interest and updating that network using acquired data. But Maucec does not disclose training a GNN at a sector level and then generalizing that trained sector model to a reservoir-scale model using measured reservoir pressure profiles. Rather, the neural networks in Maucec are generated directly for the region of interest and updated via misfit with acquired data, without any disclosure of a hierarchical or staged generalization from sector to reservoir.
This argument is unpersuasive.
The claim language “generalizing” is interpreted according to its plain meaning. See MPEP §2111.01. Nowhere in the disclosure is a specific hierarchical or staged generalization. Specification [0049] states “The sector GNNM is then generalized 520 to a large-scale reservoir GNNM 525 which can be 15 used to predict 530 - as discussed above - time evolution of the pressure profile 535 of the whole hydrocarbon reservoir.” Specification [0048] states “the trained network parameters … of the GNNM can be used to generate a large scale GNNM.” Accordingly, the most specific that the disclosure gets is a general teaching of using the network parameters for generating the large scale GNNM. Furthermore, claim 8 is not limited by the embodiments of Specification [0048]-[0049].
Maucec column 8 lines 66-67 disclose “In Block 600, dynamic response data is obtained regarding a geological region of interest.” Maucec column 9 lines 18-23 disclose:
the geological region of interest may correspond to a portion of a geological region desired or selected for running simulations and further analysis. For example, the geological region of interest may be similar to geological region (200) or reservoir region (230)
A portion of the geologic region is a sector of the reservoir. The entire reservoir region (230) is a reservoir GNNM. Repeating the graph neural network on the reservoir region is generalizing the modeling to a reservoir GNNM for the hydrocarbon reservoir.
Maucec column 8 lines 27-31 disclose “While some types of inputs are shown in FIG. 5, other embodiments are contemplated that may include pressure data, pore volume data, connate water saturation data, oil-water relative permeability data, grid cell indices, etc. as inputs to the graph neural network.” Inputting pressure data for respective grid cell indices corresponds to obtaining a pressure profile for the reservoir.
Examiner does not rely upon Chaki for curing this alleged deficiency.
Applicant remarks pages 26-28 repeat above arguments and argue respective dependent claims are allowable based on their respective dependencies. Applicant pages 26-28 does not present arguments substantively different from the above discussed arguments.
Specification
The Specification has been appropriately corrected for the noted apparent typographic errors. Accordingly, Examiner's objection(s) to the specification is withdrawn.
Claim Rejections - 35 USC § 112
Claims 8, 9, 12, and 13 have been appropriately corrected. Accordingly, Examiner's rejection of claims 8-13 and 21 under § 112 is withdrawn.
Claim Rejections - 35 USC § 101
Claim 22 was canceled. Accordingly, the §101 software per se rejection is withdrawn.
The USPTO has issued a Memorandum “Advance notice of change to the MPEP in light of Ex Parte Desjardins” (5 December 2025) available from <https://www.uspto.gov/sites/default/files/documents/memo-desjardins.pdf> [herein “Desjardins Memo”].
Because Ex Parte Desjardins (precedential) relates to machine learning it is relevant for consideration of the instant claims which involve training a GNNM. Based on the updated office guidance in the Desjardins Memo, Examiner finds the instant claims are directed toward improvements in reservoir modeling and hydrocarbon extraction. In Ex parte Desjardins, the ARP states:
We are persuaded that constitutes an improvement to how the machine learning model itself operates, and not, for example, the identified mathematical calculation.
Appeals Review Panel (ARP) decision in Ex Parte Desjardins, Appeal No. 2024-000567 at page 9, first paragraph.
By reason of analogy, if the claims in Desjardins are considered an improvement to machine learning, Examiner finds the instant claims are at least as much an improvement to reservoir modeling and hydrocarbon extraction. Accordingly, the Examiner’s §101 abstract idea rejection of claims 1-22 is withdrawn.
Claim Rejections - 35 USC § 102
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)(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 1-4, 7, 14, 15-17, and 20 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by US patent 11,409,015 B2 Maucec, et al. [herein “Maucec”].
Claim 1 recites “1. A computer-implemented method for generating a trained, graph-based neural network model (GNNM) for use in prediction of a time-evolution of a pressure profile of a hydrocarbon reservoir of a hydrocarbon field.” Maucec column 3 lines 1-6 disclose:
the reservoir graph network may be transformed into a machine-learning model, such as a graph convolutional neural network (GCNN) or simply a graph neural network. Accordingly, the machine-learning model may be trained in order to predict data that may otherwise require full-field physics simulations.
Maucec column 13 lines 53-54 disclose “the outputs of the graph neural network include oil volume, water volume, and pressure data.” Output pressure data corresponds with a predicted pressure profile.
Claim 1 further recites “comprising: obtaining geologic reservoir information and an initial spatial pressure profile for one or more reservoirs of the hydrocarbon field.” Maucec column 8 lines 66-67 disclose “In Block 600, dynamic response data is obtained regarding a geological region of interest.” Maucec column 8 lines 27-31 disclose “While some types of inputs are shown in FIG. 5, other embodiments are contemplated that may include pressure data, pore volume data, connate water saturation data, oil-water relative permeability data, grid cell indices, etc. as inputs to the graph neural network.” Inputting pressure data for respective grid cell indices corresponds to obtaining a pressure profile for the reservoir.
Claim 1 further recites “generating, based on the geologic reservoir information and the initial spatial pressure profile, a computational sector model for a selected sector comprising a plurality of modelling cells forming a grid or mesh.” Maucec column 9 lines 1-4 discloses “Dynamic response data may include transmissibility values or similar data based on one or more full-physics reservoir simulations performed for the geological region of interest.” Maucec column 9 lines 12-15 discloses “The simulator device may use grid model data to determine dynamic response data as well as updated property data for multiple reservoir property realizations.” A full-physics reservoir simulation performed corresponds to generating a computational sector model. The grid model corresponds with a plurality of modelling cells to form a grid or mesh.
Maucec column 11 lines 26-27 disclose “The simulator device may also use dynamic data, such as fluid rates and pressure data.” See further Maucec figure 9 (“pressure data E” (925)). Accordingly, pressure data is input as a part of the grid model.
Claim 1 further recites “generating, using the computational sector model, a training set of simulated spatiotemporal pressure profiles for the selected sector, each spatiotemporal pressure profile being generated for a corresponding training injection production plan (TIPP).” Maucec column 9 lines 1-4 discloses “Dynamic response data may include transmissibility values or similar data based on one or more full-physics reservoir simulations performed for the geological region of interest.” The respective dynamic response data based on the full-physics simulation is generated spatiotemporal pressure profiles.
Maucec column 13 lines 36-40 disclose “For encoding production wells, well oil production rate (WOPR) and well water production rate (WWPR) may be encoded to effect relation information. For injection wells, well water injection rate (WWIR) may be the encoded effect for the relation information.” The well water injection rate and well water production rates relation information correspond with associated training injection production plain information.
Claim 1 further recites “initializing the GNNM for the selected sector using the geologic reservoir information and the initial spatial pressure profile for the one or more reservoirs of the hydrocarbon field.” Maucec column 13 lines 55-56 disclose “an initial graph neural network.” The initial graph neural network corresponds with an initialized GNMM for the reservoir.
Claim 1 further recites “and training the GNNM using the training set of simulated spatiotemporal pressure profiles for the selected sector.” Maucec column 14 lines 7-13 disclose:
a graph neural network is trained using a loss function in accordance with one or more embodiments. For example, the loss function may be a mean square error (MSE) function that determines the difference between the training data and the predicted data for a reservoir simulation.
Training the graph neural network is a training of the GNNM using respective training data. See further Maucec figure 9.
Claim 2 further recites “2. The method of claim 1, further comprising obtaining spatiotemporal pressure profiles from a history-matched reservoir simulation model of the hydrocarbon field; and wherein the initial spatial pressure profile for the selected sector is obtained from the spatiotemporal pressure profiles obtained from the history-matched reservoir simulation model of the hydrocarbon field.” Maucec column 17 lines 37-40 disclose “an optimization problem solved by a DRL algorithm may be a dynamic model calibration criterion and/or a history matching criterion.” A model which uses a history matching criterion corresponds with a history matched model.
Maucec column 8 lines 66-67 disclose “In Block 600, dynamic response data is obtained regarding a geological region of interest.” Maucec column 8 lines 27-31 disclose “While some types of inputs are shown in FIG. 5, other embodiments are contemplated that may include pressure data, pore volume data, connate water saturation data, oil-water relative permeability data, grid cell indices, etc. as inputs to the graph neural network.” Inputting pressure data for respective grid cell indices corresponds to obtaining a pressure profile for the reservoir.
Claim 3 further recites “3. The method of claim 1, wherein the GNNM comprises one or more graph processing units configured to process input graphs to obtain output graphs, each having a graph structure having a plurality of nodes and edges.” Maucec column 7 lines 50-54 disclose “In particular, individual graph nodes in a reservoir graph network may correspond to respective grid cells in a reservoir grid model. Likewise, graph edges in a reservoir graph network may represent reservoir connectivity values between graph nodes.”
Claim 3 further recites “wherein an input graph represents global reservoir parameters and the spatiotemporal pressure profile of the selected sector at a time t and the output graph represents the global reservoir parameters and the spatiotemporal pressure profile of the selected sector at a time
t
+
∆
t
.” Maucec column 15 lines 12-19 disclose:
In Block 1050, predicted data is determined for a selected timestep using a graph neural network in accordance with one or more embodiments. For example, at a selected timestep, a graph neural network may predict oil volume data, water volume data, and/or pressure data for the current reservoir simulation. Thus, the predicted data may form a portion of a time-series set of data for analyzing the accuracy of the graph neural network within the current epoch.
A plurality of timesteps of a time-series corresponds to the data predictions being at a time t. The timestep corresponds with a
∆
t
.
Claim 4 further recites “4. The method of claim 3, wherein graph nodes are associated with a local reservoir pressure for a corresponding cell of the computational sector model.” Maucec column 13 lines 29-34 disclose “For network encoding of a reservoir graph network, a network encoding may be applied to grid cells with the following reservoir grid attributes: grid cells, oil volume (OV) water volume (WV), pressure (P), pore-volume (PORV) and well connate water (CW).” A pressure grid cell attribute corresponds with a reservoir pressure for the respective cell graph node.
Claim 4 further recites “and wherein graph edges are associated with a fluid permeability or a fluid transmissivity associated with a connection between adjacent cells of the computational sector model.” From the above list of alternatives the Examiner is selecting “a fluid transmissivity.”
Maucec column 7 lines 50-54 disclose “In particular, individual graph nodes in a reservoir graph network may correspond to respective grid cells in a reservoir grid model. Likewise, graph edges in a reservoir graph network may represent reservoir connectivity values between graph nodes.”
Maucec column 7 lines 58-63 discloses “More specifically, graph edges between nodes may correspond to transmissibility values within a reservoir region, where a transmissibility value may be a measure of the ability of a reservoir region to produce a particular fluid based on rock properties and fluid properties within the reservoir region.” The transmissibility values for the edges correspond with a fluid transmissivity.
Claim 7 further recites “7. The method of claim 1, wherein the training set comprises more than 103 or more than 104 spatiotemporal pressure profiles for the selected sector, wherein each spatiotemporal pressure profile corresponds to a different TIPP and wherein each spatiotemporal pressure profile comprises a time series of more than 103 or 104 spatial pressure profiles for the selected sector; and / or wherein a history-matched reservoir simulation model provides geologic characteristics of the hydrocarbon reservoir and allows to extract a time series of three-dimensional reservoir pressure profiles for each injection-production plan.” From the above list of alternatives the Examiner is selecting:
wherein a history-matched reservoir simulation model provides geologic characteristics of the hydrocarbon reservoir and allows to extract a time series of three-dimensional reservoir pressure profiles for each injection-production plan.
Maucec column 17 lines 37-40 disclose “an optimization problem solved by a DRL algorithm may be a dynamic model calibration criterion and/or a history matching criterion.” A model which uses a history matching criterion corresponds with a history matched model.
Maucec column 3 lines 7-10 disclose “Moreover, in simulations of massive hydrocarbon reservoirs, simulation grid sizes routinely exceed hundreds of millions grid cells with numerous possible scenarios and realizations.” Each possible scenario and realization corresponds with a different injection-production plans. See further Maucec column 11 lines 34-37 regarding time steps.
Maucec column 15 lines 12-19 disclose:
In Block 1050, predicted data is determined for a selected timestep using a graph neural network in accordance with one or more embodiments. For example, at a selected timestep, a graph neural network may predict oil volume data, water volume data, and/or pressure data for the current reservoir simulation. Thus, the predicted data may form a portion of a time-series set of data for analyzing the accuracy of the graph neural network within the current epoch.
The time-series data corresponds with an extracted time series.
Claim 14 recites “14. Computing system for generating a trained, graph-based neural network model (GNNM) for use in prediction of a time-evolution of a pressure profile of a hydrocarbon reservoir of a hydrocarbon field.” Maucec column 3 lines 1-6 disclose:
the reservoir graph network may be transformed into a machine-learning model, such as a graph convolutional neural network (GCNN) or simply a graph neural network. Accordingly, the machine-learning model may be trained in order to predict data that may otherwise require full-field physics simulations.
Maucec column 13 lines 53-54 disclose “the outputs of the graph neural network include oil volume, water volume, and pressure data.” Output pressure data corresponds with a predicted pressure profile.
Claim 14 further recites “comprising: an interface subsystem or circuitry configured for obtaining geologic reservoir information and an initial spatial pressure profile for one or more reservoirs of the hydrocarbon fields; a processing subsystem or circuitry coupled to a memory subsystem or circuitry.” Maucec column 18 lines 15-22 disclose “the computing system (1400) may include one or more computer processors (1402), non-persistent storage (1404) (e.g., volatile memory, such as random access memory (RAM), cache memory), persistent storage (1406) (e.g., a hard disk, an optical drive such as a compact disk (CD) drive or digital versatile disk (DVD) drive, a flash memory, etc.), a communication interface (1412).” The communication interface is an interface subsystem or circuitry. The computing system and/or processors are a processing subsystem or circuitry.
Maucec column 8 lines 66-67 disclose “In Block 600, dynamic response data is obtained regarding a geological region of interest.” Maucec column 8 lines 27-31 disclose “While some types of inputs are shown in FIG. 5, other embodiments are contemplated that may include pressure data, pore volume data, connate water saturation data, oil-water relative permeability data, grid cell indices, etc. as inputs to the graph neural network.” Inputting pressure data for respective grid cell indices corresponds to obtaining a pressure profile for the reservoir.
Claim 14 further recites “and configured for: generating, based on the geologic reservoir information and the initial spatial pressure profile, a computational sector model for a selected sector comprising a plurality of modelling cells forming a grid or mesh.” Maucec column 9 lines 1-4 discloses “Dynamic response data may include transmissibility values or similar data based on one or more full-physics reservoir simulations performed for the geological region of interest.” Maucec column 9 lines 12-15 discloses “The simulator device may use grid model data to determine dynamic response data as well as updated property data for multiple reservoir property realizations.” A full-physics reservoir simulation performed corresponds to generating a computational sector model. The grid model corresponds with a plurality of modelling cells to form a grid or mesh.
Maucec column 11 lines 26-27 disclose “The simulator device may also use dynamic data, such as fluid rates and pressure data.” See further Maucec figure 9 (“pressure data E” (925)). Accordingly, pressure data is input as a part of the grid model.
Claim 14 further recites “generating, using the computational sector model, a training set of simulated spatiotemporal pressure profiles for the selected sector, each spatiotemporal pressure profile being generated for a corresponding training injection production plan (TIPP).” Maucec column 9 lines 1-4 discloses “Dynamic response data may include transmissibility values or similar data based on one or more full-physics reservoir simulations performed for the geological region of interest.” The respective dynamic response data based on the full-physics simulation is generated spatiotemporal pressure profiles.
Maucec column 13 lines 36-40 disclose “For encoding production wells, well oil production rate (WOPR) and well water production rate (WWPR) may be encoded to effect relation information. For injection wells, well water injection rate (WWIR) may be the encoded effect for the relation information.” The well water injection rate and well water production rates relation information correspond with associated training injection production plain information.
Claim 14 further recites “initializing the GNNM for the selected sector using the geologic reservoir information and the initial spatial pressure profile for the one or more reservoirs of the hydrocarbon field.” Maucec column 13 lines 55-56 disclose “an initial graph neural network.” The initial graph neural network corresponds with an initialized GNMM for the reservoir.
Claim 14 further recites “and training the GNNM using the training set of simulated spatiotemporal pressure profiles for the selected sector.” Maucec column 14 lines 7-13 disclose:
a graph neural network is trained using a loss function in accordance with one or more embodiments. For example, the loss function may be a mean square error (MSE) function that determines the difference between the training data and the predicted data for a reservoir simulation.
Training the graph neural network is a training of the GNNM using respective training data. See further Maucec figure 9.
Dependent claims 15-17 and 20 are substantially similar to claims 2-4 and 7 above and are rejected for the same reasons.
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.
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.
Claims 8, 9, 11, and 21
Claims 8, 9, 11, and 21 are rejected under 35 U.S.C. 103 as being unpatentable over US patent 11,409,015 B2 Maucec, et al. [herein “Maucec”] in view of US patent 11,846,175 B2 Chaki, et al. [herein “Chaki”].
Claim 8 recites “8. A computer-implemented method for controlling reservoir pressure of a hydrocarbon reservoir of a hydrocarbon field.” Maucec column 3 lines 43-47 disclose “The control system (126) may control various operations of the well system (106), such as well production operations, well completion operations, well maintenance operations, and reservoir monitoring, assessment and development operations.”
Maucec does not explicitly disclose controlling well injection operations; however, in analogous art of machine learning used with hydrocarbon reservoirs, Chaki column 9 lines 51-53 teach “apply the trained ANN to the set of static geological data, the dynamic outputs, and subterranean grid information to determine well injection rates.” Chaki column 9 lines 45-47 teach “The well production rates may be used for controlling a wellbore operation such as operating existing wellbores.” Controlling a wellbore operation corresponds with controlling reservoir pressure.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Maucec and Chaki. One having ordinary skill in the art would have found motivation to use controlling well injection with the system of a Graph Neural Network for reservoir grid models for the advantageous purpose to be “more accurate and improve overall profitability of reservoir development.” See Chaki column 2 lines 54-56.
Claim 8 further recites “comprising: obtaining a sector graph-based neural network model (sector GNNM), trained for predicting a spatiotemporal pressure profile of a selected sector of the hydrocarbon reservoir.” Maucec column 8 lines 66-67 disclose “In Block 600, dynamic response data is obtained regarding a geological region of interest.” Maucec column 9 lines 18-23 disclose:
the geological region of interest may correspond to a portion of a geological region desired or selected for running simulations and further analysis. For example, the geological region of interest may be similar to geological region (200) or reservoir region (230)
A portion of the geologic region is a sector of the reservoir. The entire reservoir region (230) is a reservoir GNNM. Repeating the graph neural network on the reservoir region is generalizing the modeling to a reservoir GNNM for the hydrocarbon reservoir.
Maucec column 8 lines 27-31 disclose “While some types of inputs are shown in FIG. 5, other embodiments are contemplated that may include pressure data, pore volume data, connate water saturation data, oil-water relative permeability data, grid cell indices, etc. as inputs to the graph neural network.” Inputting pressure data for respective grid cell indices corresponds to obtaining a pressure profile for the reservoir.
Claim 8 further recites “generalizing, based at least in part on a measured pressure profile of the hydrocarbon reservoir, the sector GNNM to a reservoir GNNM for the hydrocarbon reservoir.” Maucec column 8 lines 66-67 disclose “In Block 600, dynamic response data is obtained regarding a geological region of interest.” Maucec column 8 lines 27-31 disclose “While some types of inputs are shown in FIG. 5, other embodiments are contemplated that may include pressure data, pore volume data, connate water saturation data, oil-water relative permeability data, grid cell indices, etc. as inputs to the graph neural network.” Inputting pressure data for respective grid cell indices corresponds to obtaining a pressure profile for the reservoir. Grid indices covering the reservoir correspond with generalizing the sectors (cells) to the overall reservoir grid.
Claim 8 further recites “obtaining a current injection-production plan, CIPP, associated with the hydrocarbon reservoir; predicting, using the reservoir GNNM and the CIPP, a pressure profile of the hydrocarbon reservoir; and adjusting, based on the predicted pressure profile of the hydrocarbon reservoir, the CIPP to optimize hydrocarbon extraction from the hydrocarbon reservoir.” Maucec column 9 lines 1-4 discloses “Dynamic response data may include transmissibility values or similar data based on one or more full-physics reservoir simulations performed for the geological region of interest.” The respective dynamic response data based on the full-physics simulation is generated spatiotemporal pressure profiles.
Maucec column 13 lines 36-40 disclose “For encoding production wells, well oil production rate (WOPR) and well water production rate (WWPR) may be encoded to effect relation information. For injection wells, well water injection rate (WWIR) may be the encoded effect for the relation information.” The well water injection rate and well water production rates relation information correspond with associated training injection production plain information.
But Maucec does not explicitly disclose adjusting a current injection-production plan; however, in analogous art of machine learning used with hydrocarbon reservoirs, Chaki column 2 lines 45-56 teaches:
The proxy-flow model can approximate the solution of a full-physics numerical reservoir simulator with enhanced computational efficiency and can be used for different tasks such as history matching, field development optimization, uncertainty quantification, and the like. A large number of simulations can be run and results can be analyzed for assimilating more information into the subterranean reservoir model and, thus, reduce uncertainty associated with the model or assess various development scenarios. In tum, the developed subterranean reservoir model can be more accurate and improve overall profitability of reservoir development.
Field development optimization is an adjustment(s) to optimize respective outputs.
Chaki column 3 lines 31-43 teach:
Gridded dynamic properties, such as pressure, fluid-phase saturations, phase compositions, etc., can be projected from time-invariant static properties like porosity and permeability and from gridded dynamic properties from previous time duction. states using a deep-learning model such as the CNN. An additional deep learning algorithm such as the DNN or the RNN, can be applied to portions of an output of the CNN to develop a proxy-flow model for the well production and injection rates. In this case, input data may be static, gridded data, such as porosity and permeability, and dynamic, gridded data, such as saturation and pressure, that are generated by the CNN. Output data can include well production rates at certain locations along the wellbore.
The gridded pressure property corresponds with a pressure profile. The well production and injection rates correspond with a injection-production plan. An optimization of a well production rate output corresponds to optimizing a hydrocarbon extraction from the reservoir.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Maucec and Chaki. One having ordinary skill in the art would have found motivation to use controlling well injection with the system of a Graph Neural Network for reservoir grid models for the advantageous purpose to be “more accurate and improve overall profitability of reservoir development.” See Chaki column 2 lines 54-56.
Claim 9 further recites “9. A computer-implemented method for controlling reservoir pressure of a hydrocarbon reservoir of a hydrocarbon field.” Maucec column 3 lines 43-47 disclose “The control system (126) may control various operations of the well system (106), such as well production operations, well completion operations, well maintenance operations, and reservoir monitoring, assessment and development operations.”
Maucec does not explicitly disclose controlling well injection operations; however, in analogous art of machine learning used with hydrocarbon reservoirs, Chaki column 9 lines 51-53 teach “apply the trained ANN to the set of static geological data, the dynamic outputs, and subterranean grid information to determine well injection rates.” Chaki column 9 lines 45-47 teach “The well production rates may be used for controlling a wellbore operation such as operating existing wellbores.” Controlling a wellbore operation corresponds with controlling reservoir pressure.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Maucec and Chaki. One having ordinary skill in the art would have found motivation to use controlling well injection with the system of a Graph Neural Network for reservoir grid models for the advantageous purpose to be “more accurate and improve overall profitability of reservoir development.” See Chaki column 2 lines 54-56.
Claim 9 further recites “comprising: obtaining a sector graph-based neural network model (sector GNNM), trained for predicting a spatiotemporal pressure profile of a selected sector of the hydrocarbon reservoir.” Maucec column 3 lines 1-6 disclose:
the reservoir graph network may be transformed into a machine-learning model, such as a graph convolutional neural network (GCNN) or simply a graph neural network. Accordingly, the machine-learning model may be trained in order to predict data that may otherwise require full-field physics simulations.
Maucec column 13 lines 53-54 disclose “the outputs of the graph neural network include oil volume, water volume, and pressure data.” Output pressure data corresponds with a predicted pressure profile.
Claim 9 further recites “generalizing, based at least in part on a measured pressure profile of the hydrocarbon reservoir, the sector GNNM to a reservoir GNNM for the hydrocarbon reservoir.” Maucec column 8 lines 66-67 disclose “In Block 600, dynamic response data is obtained regarding a geological region of interest.” Maucec column 8 lines 27-31 disclose “While some types of inputs are shown in FIG. 5, other embodiments are contemplated that may include pressure data, pore volume data, connate water saturation data, oil-water relative permeability data, grid cell indices, etc. as inputs to the graph neural network.” Inputting pressure data for respective grid cell indices corresponds to obtaining a pressure profile for the reservoir. Grid indices covering the reservoir correspond with generalizing the sectors (cells) to the overall reservoir grid.
Claim 9 further recites “obtaining a current injection-production plan, CIPP, associated with the hydrocarbon reservoir; predicting, using the reservoir GNNM and the CIPP, a pressure profile of the hydrocarbon reservoir; and adjusting, based on the predicted pressure profile of the hydrocarbon reservoir, the CIPP to optimize hydrocarbon extraction from the hydrocarbon reservoir, wherein the trained sector GNNM is obtained via the method of claim 1.” Maucec column 9 lines 1-4 discloses “Dynamic response data may include transmissibility values or similar data based on one or more full-physics reservoir simulations performed for the geological region of interest.” The respective dynamic response data based on the full-physics simulation is generated spatiotemporal pressure profiles.
Maucec column 13 lines 36-40 disclose “For encoding production wells, well oil production rate (WOPR) and well water production rate (WWPR) may be encoded to effect relation information. For injection wells, well water injection rate (WWIR) may be the encoded effect for the relation information.” The well water injection rate and well water production rates relation information correspond with associated training injection production plain information.
But Maucec does not explicitly disclose adjusting a current injection-production plan; however, in analogous art of machine learning used with hydrocarbon reservoirs, Chaki column 2 lines 45-56 teaches:
The proxy-flow model can approximate the solution of a full-physics numerical reservoir simulator with enhanced computational efficiency and can be used for different tasks such as history matching, field development optimization, uncertainty quantification, and the like. A large number of simulations can be run and results can be analyzed for assimilating more information into the subterranean reservoir model and, thus, reduce uncertainty associated with the model or assess various development scenarios. In tum, the developed subterranean reservoir model can be more accurate and improve overall profitability of reservoir development.
Field development optimization is an adjustment(s) to optimize respective outputs.
Chaki column 3 lines 31-43 teach:
Gridded dynamic properties, such as pressure, fluid-phase saturations, phase compositions, etc., can be projected from time-invariant static properties like porosity and permeability and from gridded dynamic properties from previous time duction. states using a deep-learning model such as the CNN. An additional deep learning algorithm such as the DNN or the RNN, can be applied to portions of an output of the CNN to develop a proxy-flow model for the well production and injection rates. In this case, input data may be static, gridded data, such as porosity and permeability, and dynamic, gridded data, such as saturation and pressure, that are generated by the CNN. Output data can include well production rates at certain locations along the wellbore.
The gridded pressure property corresponds with a pressure profile. The well production and injection rates correspond with a injection-production plan. An optimization of a well production rate output corresponds to optimizing a hydrocarbon extraction from the reservoir.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Maucec and Chaki. One having ordinary skill in the art would have found motivation to use controlling well injection with the system of a Graph Neural Network for reservoir grid models for the advantageous purpose to be “more accurate and improve overall profitability of reservoir development.” See Chaki column 2 lines 54-56.
Claim 11 further recites “11. The method of claim 8, the method further comprising: obtaining a second reservoir GNNM comprising different network parameters than the first reservoir GNNM.” Maucec column 10 lines 65-67 disclose “In Block 710, various reservoir property realizations are determined using one or more geostatistical processes and grid model data.” Each realization corresponds with a different reservoir with corresponding different network parameters.
Claim 11 further recites “calculating, based at least in part on a set of measured spatiotemporal reservoir pressure profiles of the hydrocarbon reservoir, a reliability metric for the first and second reservoir GNNM.” Maucec column 14 lines 32-34 disclose “that satisfy a predetermined criterion, i.e., achieve a predetermined level of prediction accuracy for the model.” A level of prediction accuracy corresponds with a reliability metric. See further Maucec column 11 lines 38-46 regarding “uncertainty quantification.”
Claim 11 further recites “and selecting, based on the reliability metric, the first or the second reservoir GNNM for predicting the reservoir pressure profile of the hydrocarbon reservoir.” Maucec column 15 lines 1-3 disclose “In some embodiments, multiple trained graph neural networks are compared and the best trained model is selected accordingly.” Selecting a best model, i.e. with a best prediction accuracy, corresponds with selecting a trained graph neural network according to a reliability metric.
Claim 21 further recites “21. Computing system comprising processing and interface circuitry coupled to memory storing instructions for carrying out the method of claim 8.” Maucec column 18 lines 15-22 disclose “the computing system (1400) may include one or more computer processors (1402), non-persistent storage (1404) (e.g., volatile memory, such as random access memory (RAM), cache memory), persistent storage (1406) (e.g., a hard disk, an optical drive such as a compact disk (CD) drive or digital versatile disk (DVD) drive, a flash memory, etc.), a communication interface (1412).” The communication interface is an interface subsystem or circuitry. The computing system and/or processors are a processing subsystem or circuitry.
Dependent Claim 10
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over Maucec and Chaki as applied to claim 9 above, and further in view of Razak, S., et al. “Transfer Learning with Recurrent Neural Networks for Long-term Production Forecasting in Unconventional Reservoirs” Unconventional Resources Tech. Conf. (2021) [herein “Razak”].
Claim 10 further recites “10. The method of claim 9, wherein the sector GNNM and the reservoir GNNM comprises a mesh-graph neural network model using an encoder-processor-decoder architecture.” Maucec nor Chaki explicitly disclose an encoder decode architecture; however, in analogous art of machine learning for reservoir forecast, Razak page 5 “Forecast model” section teaches The neural network architecture of the proposed forecast model is composed of a pair of LSTM cells denoted as
E
n
c
θ
and
D
e
c
γ
.” Razak page 6 first paragraph further teaches “a dynamic encoding of relevant information from the past data” and “These encodings are concatenated and fed into the LSTM decoder.” This corresponds with an encode and decode architecture for the neural network model.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Maucec, Chaki, and Razak. One having ordinary skill in the art would have found motivation to use an LSTM encoder-decoder architecture into the system of Graph Neural Network for reservoir grid models for the advantageous purpose of obtaining long-term production forecasts. See Razak title and abstract.
Dependent Claims 12 and 13
Claims 12 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Maucec and Chaki as applied to claim 8 above, and further in view of US patent 8,271,247 B2 Davidson [herein “Davidson”].
Claim 12 further recites “12. The method of claim 8, wherein adjusting the CIPP further comprises: calculating, based at least in part on the predicted pressure profile of the hydrocarbon reservoir, an average reservoir pressure, ARP, for one or more sectors of the hydrocarbon reservoir.” Maucec column 16 lines 13-17 disclose “well workover design and optimization of pressure maintenance, and/or injection strategies. Thus, a trained model may determine a maximized field recovery that satisfies target rates and maximizes a production plateau length.”
Maucec does not explicitly disclose calculating an average reservoir pressure; however, in analogous art of reservoir pressure management, Davidson column 16 lines 58-60 disclose “for the pressure maintenance strategy, a reservoir engineer may specify a target average pressure for the geobody.” Davidson column 17 lines 1-8 disclose “a target VRR
V
R
R
t
a
r
g
e
t
is dynamically calculated using the following equation (Eq6): …
E
p
is the error in the target pressure minus the average pressure
P
t
a
r
g
e
t
-
P
a
v
e
r
a
g
e
.” The average pressure corresponds with a calculated average reservoir pressure for at least one sector.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Maucec, Chaki, and Davidson. One having ordinary skill in the art would have found motivation to use a pressure maintenance strategy into the system of Graph Neural Network for reservoir grid models for the advantageous purpose of “use of process control theory to set well rates for the reservoir simulation, incorporation of material balance with built-in corrections to numerical errors for voidage replacement and pressure maintenance strategies, and development of well management strategies based on reservoir fluid flows across reservoir boundaries.” See Davidson column 10 lines 27-33.
Claim 12 further recites “and comparing the ARP for the one or more sectors with a pressure maintenance requirement, PMR, for the one or more sectors corresponding to a reservoir exploitation plan.” Davidson column 16 lines 58-60 disclose “for the pressure maintenance strategy, a reservoir engineer may specify a target average pressure for the geobody.” Davidson column 17 lines 1-8 disclose “a target VRR
V
R
R
t
a
r
g
e
t
is dynamically calculated using the following equation (Eq6): …
E
p
is the error in the target pressure minus the average pressure
P
t
a
r
g
e
t
-
P
a
v
e
r
a
g
e
.” The target pressure of the pressure maintenance strategy corresponds with a pressure maintenance requirement for at least one sector.
Claim 12 further recites “and adjusting operational parameters of the CIPP if the difference between the ARP and the PMR is larger than a threshold value; or maintaining operational parameters of the CIPP if the difference between the ARP and the PMR is smaller or equal than the threshold value.” Davidson column 16 lines 58-60 disclose “for the pressure maintenance strategy, a reservoir engineer may specify a target average pressure for the geobody.” Davidson column 17 lines 1-8 disclose “a target VRR
V
R
R
t
a
r
g
e
t
is dynamically calculated using the following equation (Eq6): …
E
p
is the error in the target pressure minus the average pressure
P
t
a
r
g
e
t
-
P
a
v
e
r
a
g
e
.” Dynamically calculating the VRR target is an adjustment of respective operation parameters of the injection allocation and production plan.
Davidson column 17 lines 43-45 teach “The MBG pressure maintenance algorithm correctly deviated the VRR away from one so as to return the average reservoir pressure to the target pressure.” The deviation of the VRR is an adjustment of respective operational parameters.
Claim 13 further recites “13. The method of claim 12, wherein adjusting operational parameters of the CIPP further comprises: developing a reward based artificial intelligence technique using the CIPP as an initial input.” Maucec column 17 lines 18-21 disclose “The DRL algorithm architecture (1300) may further includes a reward (r) that defines the feedback by which the success or failure of an agent's actions is measured.” A Deep Reinforcement Learning (DRL) architecture is an artificial intelligence technique. The reward function is a developed reward.
Maucec column 16 lines 40-49 disclose “As shown in FIG. 12A, the neural network parameterization C (1210) includes object data (1240) that defines various grid-cell objects, various producer objects, and an injector object, … oil-well oil production rate (WOPR), well water production rate (WWPR), and well water injection rate (WWIR).” The neural network parameterizing the oil production rate, water production rate, and water injection rate
Claim 13 further recites “and using the reinforcement model to optimize the operational parameters of the CIPP such that the difference between the ARP and the PMR becomes smaller than the threshold value.” Maucec does not explicitly disclose calculating an average reservoir pressure; however, in analogous art of reservoir pressure management, Davidson column 16 lines 58-60 disclose “for the pressure maintenance strategy, a reservoir engineer may specify a target average pressure for the geobody.” Davidson column 17 lines 1-8 disclose “a target VRR
V
R
R
t
a
r
g
e
t
is dynamically calculated using the following equation (Eq6): …
E
p
is the error in the target pressure minus the average pressure
P
t
a
r
g
e
t
-
P
a
v
e
r
a
g
e
.” Dynamically calculating the VRR target is an adjustment of respective operation parameters of the injection allocation and production plan.
Davidson column 17 lines 43-45 teach “The MBG pressure maintenance algorithm correctly deviated the VRR away from one so as to return the average reservoir pressure to the target pressure.” The deviation of the VRR is an adjustment of respective operational parameters. Davidson is teaching returning the average reservoir pressure to the target pressure. This corresponds with a minimization of the difference between ARP and PMR.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Maucec, Chaki, and Davidson. One having ordinary skill in the art would have found motivation to use a pressure maintenance strategy into the system of Graph Neural Network for reservoir grid models for the advantageous purpose of “use of process control theory to set well rates for the reservoir simulation, incorporation of material balance with built-in corrections to numerical errors for voidage replacement and pressure maintenance strategies, and development of well management strategies based on reservoir fluid flows across reservoir boundaries.” See Davidson column 10 lines 27-33.
Allowable Subject Matter
Claims 5, 6, 18, and 19 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
A statement of reasons for indication of allowable subject matter was previously presented in the office action dated 30 July 2025.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Jay B Hann whose telephone number is (571)272-3330. The examiner can normally be reached M-F 10am-7pm EDT.
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, Renee Chavez can be reached at (571) 270-1104. 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.
/Jay Hann/Primary Examiner, Art Unit 2186 18 February 2026