DETAILED ACTIONS
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 12/10/2025 has been entered.
Response to Amendment
This office action is in response to the amendments/arguments submitted by the Applicant(s) on 07/14/2025.
Status of the Claims
Claims 1-6, 8, 10-12, 14-17, 19-20 are pending.
Claims 1,12,16 and 20 are amended.
Response to Arguments
Rejections Under 35 U.S.C. 103
Applicant's arguments, see remarks pages 8-14, filed 07/14/2025.
with respect to the rejection(s) of Claims under 35 U.S.C. 103 has been considered, and are moot because the amendment has necessitated a new ground of rejections. The new rejections are set forth below.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-6,8,10, 12, 14-17,and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Li et al. (US 20220099855 A1, hereinafter Li) and in view of and in further view of Vasudevan et al. (US 20120179635 A1, hereinafter Vasudevan, previously cited).
Regarding claim 1, Li teaches,
A method, comprising:
obtaining a first surface seismic dataset (Li, Figure 3, [0065], “acquire seismic data (see, e.g., data 360)”) of a first subterranean region (Li, Figure 3, a geologic environment 341”), by a first plurality of surface seismic receivers disposed at a first surface location on a surface of the earth, (Li, Figure 3, [0066], energy source (e.g., a transmitter) 342, acquisition equipment 392) wherein the first subterranean region comprises a first wellbore (Li,, Figure 3, a bore 343 [0066],;
obtaining a measured seismic velocity profile (Li, Figure 3, [0076], “As an example, seismic data may be processed to obtain an elastic model pertaining to elastic properties of a geologic subsurface region. For example, compression velocity (vP)-to-shear velocity (vs). Figure 8, [0152], As explained, geological and/or geophysical interpretation of a seismic data cube can involve (…) velocity model building, stratigraphic analysis, etc”” through the first subterranean region by a plurality of wellbore seismic receivers (one or more sensors (e.g., receivers) 344[0068]); disposed in the first wellbore (a bore 343) of the first subterranean region (Li, Figure 3, 341)
determining a set of seismic attributes for the first subterranean region from the first surface seismic dataset; (Li, Figure 1, [0034] In the example of FIG. 1, “the simulation component 120 may process information to conform to one or more
attributes specified by the attribute component 130”).
training, using the measured seismic velocity profile (Li, Figures 9, Step 910, “receive seismic data” NOTE: velocity profile is one of the seismic data)and the set of seismic attributes relating to the first subterranean region (Li, Figures 1, Framework 170, [0039] As an example, seismic data may be processed using a framework [0041], a framework may include an input component for receipt of information from interpretation of seismic data, one or more attributes based at least in part on seismic data, log data, image data, etc. NOTE: “velocity profile” is one of the seismic data).
a machine-learning network to predict a predicted seismic velocity profile from the set of seismic attributes (Li, Figures 9, Step 910-960, steps 920 Interpret Seismic Date using Trained ML Model to Generate Processed Seismic Data 920;),
wherein the predicted seismic velocity profile is an estimate of the measured seismic velocity profile (Li, Figure 9,11 [0165] As shown in the example of FIG. 9, the supervised portion 904 includes a label block 930 for labeling a portion of the processed seismic data, a training block 940 for training a ML model to generate a trained ML model. the method 900 can include an output block 960 that output results from the supervised portion 904”)., and
obtaining a second surface seismic dataset of a second subterranean region by a second plurality of surface seismic receivers disposed at a second surface location on the surface of the earth, wherein the second subterranean region does not comprise a wellbore; (Li, Figure 1, 156, [0043] Other equipment 156 may be located remote from a well site and include sensing, detecting, emitting or other circuitry.
Such equipment may include storage and communication circuitry to store and to communicate data, instructions, etc. As an example, one or more satellites may be provided for purposes of communications, data acquisition, etc. For
example, FIG. 1 shows a satellite in communication with the network 155 that may be configured for communications, noting that the satellite may additionally or alternatively include circuitry for imagery (e.g., spatial, spectral, temporal,
radiometric, etc.”).
determining a set of seismic attribute volumes for the second subterranean region from the second surface seismic dataset predicting a seismic velocity model (Li, Figure 19, steps 1910-19-30, [0263] “a second trained machine model can predict a stratigraphic Earth model of a geologic region. As an example, such a model may be sufficient to model seismic energy using a velocity model that models velocities of seismic energy in various layers of a stratigraphic sequence”) for the second subterranean region, using the trained machine-learning network and the set of seismic attribute volumes; forming a depth-domain seismic image of the second subterranean region based, at least in part, on the seismic velocity model (LI, Figure 19, step 1930, [0243] a build block 1930 for building a second trained machine model, as initialized from the first trained machine model, via supervised learning using the received labels, where the second trained machine model predicts stratigraphy of a geologic region from seismic image data of the geologic region. [0263] As an example, such a model may be sufficient to model seismic energy using a velocity model that models velocities of seismic energy in various layers of a stratigraphic sequence” NOTE: second ML model is generated for a second target region based on first target region trained ML model.)); and
determining, using a seismic interpretation workstation, a location of a hydrocarbon reservoir in the second subterranean region based, (Li, [0150] “once the neural network has been trained, it can inherit the domain knowledge that has been put into an interpretation by an interpretation expert or experts. The inherited domain knowledge can then be applied to new seismic data, automatically, by a user (Figure 7, desktop), etc. As interpretation results may be generated in a lesser amount of time for an area of interest through use of ML (e.g., a trained machine, etc.), various processes may be improved. For example, seismic survey parameters may be adjusted during a seismic survey (e.g., land, marine, etc.), one or more field operations may be adjusted, optionally during a seismic survey, etc.)
at least in part, on the depth-domain seismic image (LI, Figure 7, 8, [0075] “seismic data may be processed in a technique called "depth imaging" to form an image (e.g.,a depth image) of reflection amplitudes in a depth domain for a particular target structure (e.g., a geologic subsurface region of interest)”).
Li is silent on wherein the machine-learning network comprises Gaussian process regression.
However, Vasudevan teaches a wherein the machine-learning network comprises Gaussian process regression. (Vasudevan, Figure 1, [0010]-[0011], [0010] To generate a Gaussian process model the method may include using the first dataset to learn optimized kernel hyperparameters and a first noise hyperparameter, and using the learnt kernel hyperparameters and second dataset to learn an optimized second noise hyperparameter. Gaussian process
regression may then be performed using the first and second datasets, first and second noise hyperparameters and kernel hyperparameters. [0011] The quantity of interest may be geographical elevation, and the domain of interest may be a mining pit).
It would have been obvious to a person having ordinary skill in the
art before the effective filing date to modify Li’s method for predicting optimized velocity profile model to incorporate a machine learning network with the Gaussian process regression as taught by Vasudevan and obtain an accurate terrain model by incorporating uncertainty and correlation of data and generate geographic map (Vasudevan, [0002]-[0005]). It would have been obvious to a person of ordinary skill to include the well-known Gaussian Process Regression (GPR) along with the other machine learning network, in order to yield the predicted results of generating accurate seismic image or geographic map, yet with higher accuracy (KSR).
Regarding claim 2, Combination of Li and Vasudevan teaches the method of claim 1,
Li further teaches wherein the measured seismic velocity profile comprises a check-shot survey velocity profile (Li, Figure 7,[0131] For example, consider utilizing
the FDMOD for generating synthetic shot gathers by using full 3D”).
Regarding Claim 3, Combination of Li and Vasudevan teaches the method of
claim 1,
Li further teaches wherein the set of seismic attribute volumes comprises a dip volume. (Li, [0202], “training of a neural network system (NNS) is performed on a relatively small portion of a single seismic cube (volumetric seismic data) to generate
a trained NNS and where the trained NNS is utilized for prediction on one or more other portions of the single seismic cube “)
Regarding Claim 4, Combination of Li and Vasudevan teaches the method of
claim 1,
Li further teaches wherein the second surface seismic dataset comprises the first surface seismic dataset. (Li, Figure 19, step1910-1920, [0150] “once the neural network has been trained, it can inherit the domain knowledge that has been put into an interpretation by an interpretation expert or experts. The inherited domain knowledge can then be applied to new seismic data, automatically, by a user” NOTE: first seismic data is used to trained the model.).
Regarding Claim 5, Combination of Li and Vasudevan teaches the method of
claim 1,
Li is silent on wherein the Gaussian process regression is a stochastic process comprising at least one mean function, at least one kernel, and at least one hyper-parameter.
However, Vasudevan teaches wherein the Gaussian process regression is a stochastic process comprising at least one mean function, at least one kernel, and at least one hyper-parameter. (Vasudevan, [0005] Gaussian process-based terrain modeling provides an approach that enables a multi-resolution representation of
space, incorporates and manages uncertainty in a statistically sound way, and can handle spatially correlated data in an appropriate manner. Gaussian Processes (GPs) are stochastic processes based on the normal (Gaussian) distribution and can be used to good effect as a powerful nonparametric learning
technique for spatial modeling. [0010] To generate a Gaussian process model the method may include using the first dataset to learn optimized kernel hyperparameters and a first noise hyperparameter, and using the learnt kernel hyperparameters and second dataset to learn an optimized second noise hyperparameter).
It would have been obvious to a person having ordinary skill in the
art before the effective filing date to modify Li’s method for predicting optimized velocity profile model to incorporate a machine learning network with the Gaussian process regression as taught by Vasudevan and obtain an accurate terrain model by incorporating uncertainty and correlation of data and generate geographic map (Vasudevan, [0002]-[0005]). It would have been obvious to a person of ordinary skill to include the well-known Gaussian Process Regression (GPR) along with the other machine learning network, in order to yield the predicted results of generating accurate seismic image or geographic map, yet with higher accuracy (KSR).
Regarding Claim 6, Combination of Li and Vasudevan teaches the method of
claim 1,
Li is silent on wherein the seismic velocity model comprises an uncertainty estimate provided by a posterior covariance matrix.
However, Vasudevan wherein the seismic velocity model comprises an uncertainty estimate provided by a posterior covariance matrix (Vasudevan, Figure 9-10, [0095], “The resulting output (225) is a fused estimate and uncertainty based on the multiple datasets and sensor characteristic [0071], matrix of covariances evaluated at all pairs of training and test points”.
It would have been obvious to a person having ordinary skill in the
art before the effective filing date to modify Li’s method for predicting optimized velocity profile model to incorporate a machine learning network with the Gaussian process regression as taught by Vasudevan and obtain an accurate terrain model by incorporating uncertainty and correlation of data and generate geographic map (Vasudevan, [0002]-[0005]). It would have been obvious to a person of ordinary skill to include the well-known Gaussian Process Regression (GPR) along with the other machine learning network, in order to yield the predicted results of generating accurate seismic image or geographic map, yet with higher accuracy (KSR).
Regarding Claim 8, Combination of Li and Vasudevan teaches the method of
claim 6,
Li further teaches wherein forming the depth-domain seismic image (LI, Figure 7, 8, [0075] “seismic data may be processed in a technique called "depth imaging" to form an image (e.g.,a depth image) of reflection amplitudes in a depth domain for a particular target structure (e.g., a geologic subsurface region of interest)”) comprises determining a confidence interval, based on the uncertainty estimate, for at least one depth in the depth-domain seismic image, and wherein the confidence interval indicates a reliability of the depth-domain seismic image at the at least one depth and is based on a mean prediction bounded by a variance (Li, [0152],by one or more ML processes. For example, consider a first ML process that is trained for a first pass through a seismic cube to generate ML interpretation results followed by a second ML process where an interpreter further refines a portion of the ML interpretations as part of the second ML process where the refinement includes labeling the portion for to train a ML model that can then be applied to one or more other portions of the seismic cube, optionally as the seismic cube prior to the first pass or after the first pass. In such an approach, the interpreter's effort can be limited to a particular portion or portions of a seismic cube as pre-processed with some level of confidence whereby the interpreter's effort improves that level of confidence for applying to one or more other portions of the seismic cube”).
Regarding Claim 10, Combination of Li and Vasudevan teaches the method of
claim 6,
Li further teaches further comprising planning, using a wellbore planning system, a wellbore path to intersect the hydrocarbon reservoir (Li, Figures 1-2, [0054], [0054] As to the convention 240 for dip, as shown, the three-dimensional orientation of a plane can be defined by its dip and strike. Dip is the angle of slope of a plane from a horizontal plane (e.g., an imaginary plane) measured in a vertical plane in a specific direction. Dip may be defined by magnitude (e.g., also known as angle or amount) and azimuth (e.g., also known as direction). As shown in the convention 240 of FIG. 2,”).
Regarding Claim 11, Combination of Li and Vasudevan teaches the method of
claim 10,
Li further teaches further comprising drilling, using a wellbore drilling system, a second wellbore guided by the planned wellbore path (Li, Figure 1-3, the equipment
157 and/or 158 may include components, a system, systems, etc. for fracturing, seismic sensing, analysis of seismic data, assessment of one or more fractures, etc.).
Regarding Claim 12, Li teaches,
A non-transitory computer readable medium storing a set of instructions (Li, Figure 2, Computer 254),
executable by a computer processor (Li, Figure 2, Processor 256), the set of instructions (Li, Figure 2, Instructions 270), comprising functionality for:
receiving a first surface seismic dataset (Li, Figure 3, [0065], “acquire seismic data (see, e.g., data 360)”) of a first subterranean region (Li, Figure 3, a geologic environment 341”),
wherein the first surface seismic dataset (see, e.g., data 360)”) was obtained by a first plurality of surface seismic receivers disposed at a first surface location on a surface of the earth (Li, Figure 3, [0066], energy source (e.g., a transmitter) 342, acquisition equipment 392), wherein the first subterranean region comprises a first wellbore (Li, Figure 3, a bore 343 [0066]);
receiving a measured seismic velocity profile (Li, Figure 3, [0076], “As an example, seismic data may be processed to obtain an elastic model pertaining to elastic properties of a geologic subsurface region. For example, compression velocity (vP)-to-shear velocity (vs). Figure 8, [0152], As explained, geological and/or geophysical interpretation of a seismic data cube can involve (…) velocity model building, stratigraphic analysis, etc”) through the first subterranean region, wherein the measured seismic velocity profile was obtained by a plurality of wellbore seismic receivers (one or more sensors (e.g., receivers) 344[0068]) disposed in the first wellbore (a bore 343) of the first subterranean region, Figure 3, 341);
determining a set of seismic attributes for the first subterranean region from the first surface seismic dataset (Li, Figure 1, [0034] In the example of FIG. 1, “the simulation component 120 may process information to conform to one or more
attributes specified by the attribute component 130”);
training, using the measured seismic velocity profile (Li, Figures 9, Step 910, “receive seismic data” NOTE: velocity profile is one of the seismic data)and the set of seismic attributes relating to the first subterranean region (Li, Figures 1, Framework 170, [0039] As an example, seismic data may be processed using a framework [0041], a framework may include an input component for receipt of information from interpretation of seismic data, one or more attributes based at least in part on seismic data, log data, image data, etc. NOTE: “velocity profile” is one of the seismic data).
a machine-learning network to predict a predicted seismic velocity profile from the set of seismic attributes (Li, Figures 9, Step 910-960, steps 920 Interpret Seismic Date using Trained ML Model to Generate Processed Seismic Data 920;),
wherein the predicted seismic velocity profile is an estimate of the measured seismic velocity profile (Li, Figure 9,11 [0165] As shown in the example of FIG. 9, the supervised portion 904 includes a label block 930 for labeling a portion of the processed seismic data, a training block 940 for training a ML model to generate a trained ML model. the method 900 can include an output block 960 that output results from the supervised portion 904”)., and
obtaining a second surface seismic dataset of a second subterranean region by a second plurality of surface seismic receivers disposed at a second surface location on the surface of the earth, wherein the second subterranean region does not comprise a wellbore; (Li, Figure 1, 156, [0043] Other equipment 156 may be located remote from a well site and include sensing, detecting, emitting or other circuitry.
Such equipment may include storage and communication circuitry to store and to communicate data, instructions, etc. As an example, one or more satellites may be provided for purposes of communications, data acquisition, etc. For
example, FIG. 1 shows a satellite in communication with the network 155 that may be configured for communications, noting that the satellite may additionally or alternatively include circuitry for imagery (e.g., spatial, spectral, temporal,
radiometric, etc.”).
determining a set of seismic attribute volumes for the second subterranean region from the second surface seismic dataset predicting a seismic velocity model (Li, Figure 19, steps 1910-19-30, [0263] “a second trained machine model can predict a stratigraphic Earth model of a geologic region. As an example, such a model may be sufficient to model seismic energy using a velocity model that models velocities of seismic energy in various layers of a stratigraphic sequence”) for the second subterranean region, using the trained machine-learning network and the set of seismic attribute volumes; forming a depth-domain seismic image of the second subterranean region based, at least in part, on the seismic velocity model (LI, Figure 19, step 1930, [0243] a build block 1930 for building a second trained machine model, as initialized from the first trained machine model, via supervised learning using the received labels, where the second trained machine model predicts stratigraphy of a geologic region from seismic image data of the geologic region. [0263] As an example, such a model may be sufficient to model seismic energy using a velocity model that models velocities of seismic energy in various layers of a stratigraphic sequence” NOTE: second ML model is generated for a second target region based on first target region trained ML model.)); and
determining, using a seismic interpretation workstation, a location of a hydrocarbon reservoir in the second subterranean region based, (Li, [0150] “once the neural network has been trained, it can inherit the domain knowledge that has been put into an interpretation by an interpretation expert or experts. The inherited domain knowledge can then be applied to new seismic data, automatically, by a user (Figure 7, desktop), etc. As interpretation results may be generated in a lesser amount of time for an area of interest through use of ML (e.g., a trained machine, etc.), various processes may be improved. For example, seismic survey parameters may be adjusted during a seismic survey (e.g., land, marine, etc.), one or more field operations may be adjusted, optionally during a seismic survey, etc.)
at least in part, on the depth-domain seismic image (LI, Figure 7, 8, [0075] “seismic data may be processed in a technique called "depth imaging" to form an image (e.g. a depth image) of reflection amplitudes in a depth domain for a particular target structure (e.g., a geologic subsurface region of interest)”).
Li is silent on wherein the machine-learning network comprises Gaussian process regression.
However, Vasudevan teaches a wherein the machine-learning network comprises Gaussian process regression. (Vasudevan, Figure 1, [0010]-[0011], [0010] To generate a Gaussian process model the method may include using the first dataset to learn optimized kernel hyperparameters and a first noise hyperparameter, and using the learnt kernel hyperparameters and second dataset to learn an optimized second noise hyperparameter. Gaussian process
regression may then be performed using the first and second datasets, first and second noise hyperparameters and kernel hyperparameters. [0011] The quantity of interest may be geographical elevation, and the domain of interest may be a mining pit).
It would have been obvious to a person having ordinary skill in the
art before the effective filing date to modify Li’s method for predicting optimized velocity profile model to incorporate a machine learning network with the Gaussian process regression as taught by Vasudevan and obtain an accurate terrain model by incorporating uncertainty and correlation of data and generate geographic map (Vasudevan, [0002]-[0005]). It would have been obvious to a person of ordinary skill to include the well-known Gaussian Process Regression (GPR) along with the other machine learning network, in order to yield the predicted results of generating accurate seismic image or geographic map, yet with higher accuracy (KSR).
Regarding Claim 14, Combination of Li and Vasudevan teaches the non-transitory computer readable medium of claim 12,
Li further teaches wherein forming the depth-domain seismic image (LI, Figure 7, 8, [0075] “seismic data may be processed in a technique called "depth imaging" to form an image (e.g.,a depth image) of reflection amplitudes in a depth domain for a particular target structure (e.g., a geologic subsurface region of interest)”) comprises determining a confidence interval, based on the uncertainty estimate, for at least one depth in the depth-domain seismic image, and wherein the confidence interval indicates a reliability of the depth-domain seismic image at the at least one depth and is based on a mean prediction bounded by a variance (Li, [0152],by one or more ML processes. For example, consider a first ML process that is trained for a first pass through a seismic cube to generate ML interpretation results followed by a second ML process where an interpreter further refines a portion of the ML interpretations as part of the second ML process where the refinement includes labeling the portion for to train a ML model that can then be applied to one or more other portions of the seismic cube, optionally as the seismic cube prior to the first pass or after the first pass. In such an approach, the interpreter's effort can be limited to a particular portion or portions of a seismic cube as pre-processed with some level of confidence whereby the interpreter's effort improves that level of confidence for applying to one or more other portions of the seismic cube”).
Regarding Claim 15, Combination of Li and Vasudevan teaches the non-transitory computer readable medium of claim 12,
Li further teaches comprising planning, using a wellbore planning system, a wellbore path to intersect the hydrocarbon reservoir. (Li, Figures 1-2, [0054], [0054] As to the convention 240 for dip, as shown, the three-dimensional orientation of a plane can be defined by its dip and strike. Dip is the angle of slope of a plane from a horizontal plane (e.g., an imaginary plane) measured in a vertical plane in a specific direction. Dip may be defined by magnitude (e.g., also known as angle or amount) and azimuth (e.g., also known as direction). As shown in the convention 240 of FIG. 2,”).
Regarding Claim 16, Li teaches,
A system, (Li, Figure 1,100 system) system comprising:
a seismic acquisition system Li, Figure 1-3) configured to:
obtaining a first surface seismic dataset (Li, Figure 3, [0065], “acquire seismic data (see, e.g., data 360)”) of a first subterranean region (Li, Figure 3, a geologic environment 341”), by a first plurality of surface seismic receivers disposed at a first surface location on a surface of the earth, (Li, Figure 3, [0066], energy source (e.g., a transmitter) 342, acquisition equipment 392) wherein the first subterranean region comprises a first wellbore (Li,, Figure 3, a bore 343 [0066],;
obtaining a measured seismic velocity profile (Li, Figure 3, [0076], “As an example, seismic data may be processed to obtain an elastic model pertaining to elastic properties of a geologic subsurface region. For example, compression velocity (vP)-to-shear velocity (vs). Figure 8, [0152], As explained, geological and/or geophysical interpretation of a seismic data cube can involve (…) velocity model building, stratigraphic analysis, etc”” through the first subterranean region by a plurality of wellbore seismic receivers (one or more sensors (e.g., receivers) 344[0068]); disposed in the first wellbore (a bore 343) of the first subterranean region (Li, Figure 3, 341)
obtaining a second surface seismic dataset of a second subterranean region by a second plurality of surface seismic receivers disposed at a second surface location on the surface of the earth, wherein the second subterranean region does not comprise a wellbore; (Li, Figure 1, 156, [0043] Other equipment 156 may be located remote from a well site and include sensing, detecting, emitting or other circuitry.
Such equipment may include storage and communication circuitry to store and to communicate data, instructions, etc. As an example, one or more satellites may be provided for purposes of communications, data acquisition, etc. For
example, FIG. 1 shows a satellite in communication with the network 155 that may be configured for communications, noting that the satellite may additionally or alternatively include circuitry for imagery (e.g., spatial, spectral, temporal,
radiometric, etc.”).;
a computer Li, Figure 2, Computer 254), system configured to:
receive the first surface seismic dataset (Li, Figure 3, [0065], “acquire seismic data (see, e.g., data 360)”),
receive the measured seismic velocity profile, (Li, Figure 3, [0076], “As an example, seismic data may be processed to obtain an elastic model pertaining to elastic properties of a geologic subsurface region. For example, compression velocity (vP)-to-shear velocity (vs). Figure 8, [0152], As explained, geological and/or geophysical interpretation of a seismic data cube can involve (…) velocity model building, stratigraphic analysis, etc”)
determine a set of seismic attributes for the first subterranean region from the first surface seismic dataset, (Li, Figure 1, [0034] In the example of FIG. 1, “the simulation component 120 may process information to conform to one or more
attributes specified by the attribute component 130”);
training, using the measured seismic velocity profile (Li, Figures 9, Step 910, “receive seismic data” NOTE: velocity profile is one of the seismic data)and the set of seismic attributes relating to the first subterranean region (Li, Figures 1, Framework 170, [0039] As an example, seismic data may be processed using a framework [0041], a framework may include an input component for receipt of information from interpretation of seismic data, one or more attributes based at least in part on seismic data, log data, image data, etc. NOTE: “velocity profile” is one of the seismic data).
a machine-learning network to predict a predicted seismic velocity profile from the set of seismic attributes (Li, Figures 9, Step 910-960, steps 920 Interpret Seismic Date using Trained ML Model to Generate Processed Seismic Data 920;),
wherein the predicted seismic velocity profile is an estimate of the measured seismic velocity profile (Li, Figure 9,11 [0165] As shown in the example of FIG. 9, the supervised portion 904 includes a label block 930 for labeling a portion of the processed seismic data, a training block 940 for training a ML model to generate a trained ML model. the method 900 can include an output block 960 that output results from the supervised portion 904”). and
determining a set of seismic attribute volumes for the second subterranean region from the second surface seismic dataset predicting a seismic velocity model (Li, Figure 19, steps 1910-19-30, [0263] “a second trained machine model can predict a stratigraphic Earth model of a geologic region. As an example, such a model may be sufficient to model seismic energy using a velocity model that models velocities of seismic energy in various layers of a stratigraphic sequence”) for the second subterranean region, using the trained machine-learning network and the set of seismic attribute volumes; forming a depth-domain seismic image of the second subterranean region based, at least in part, on the seismic velocity model (LI, Figure 19, step 1930, [0243] a build block 1930 for building a second trained machine model, as initialized from the first trained machine model, via supervised learning using the received labels, where the second trained machine model predicts stratigraphy of a geologic region from seismic image data of the geologic region. [0263] As an example, such a model may be sufficient to model seismic energy using a velocity model that models velocities of seismic energy in various layers of a stratigraphic sequence” NOTE: second ML model is generated for a second target region based on first target region trained ML model); forming a depth-domain seismic image of the second subterranean region based, at least in part, on the seismic velocity model (LI, Figure 7, 8, [0075] “seismic data may be processed in a technique called "depth imaging" to form an image (e.g. a depth image) of reflection amplitudes in a depth domain for a particular target structure (e.g., a geologic subsurface region of interest)”).and
a seismic interpretation workstation configured to determine a location of a hydrocarbon reservoir in the second subterranean region based, at least in part, on the depth-domain seismic image. determining, using a seismic interpretation workstation, a location of a hydrocarbon reservoir in the second subterranean region based, (Li, [0150] “once the neural network has been trained, it can inherit the domain knowledge that has been put into an interpretation by an interpretation expert or experts. The inherited domain knowledge can then be applied to new seismic data, automatically, by a user (Figure 7, desktop), etc. As interpretation results may be generated in a lesser amount of time for an area of interest through use of ML (e.g., a trained machine, etc.), various processes may be improved. For example, seismic survey parameters may be adjusted during a seismic survey (e.g., land, marine, etc.), one or more field operations may be adjusted, optionally during a seismic survey, etc.)
receive the second surface seismic dataset, Li is silent on wherein the machine-learning network comprises Gaussian process regression.
However, Vasudevan teaches a wherein the machine-learning network comprises Gaussian process regression. (Vasudevan, Figure 1, [0010]-[0011], [0010] To generate a Gaussian process model the method may include using the first dataset to learn optimized kernel hyperparameters and a first noise hyperparameter, and using the learnt kernel hyperparameters and second dataset to learn an optimized second noise hyperparameter. Gaussian process
regression may then be performed using the first and second datasets, first and second noise hyperparameters and kernel hyperparameters. [0011] The quantity of interest may be geographical elevation, and the domain of interest may be a mining pit).
It would have been obvious to a person having ordinary skill in the
art before the effective filing date to modify Li’s method for predicting optimized velocity profile model to incorporate a machine learning network with the Gaussian process regression as taught by Vasudevan and obtain an accurate terrain model by incorporating uncertainty and correlation of data and generate geographic map (Vasudevan, [0002]-[0005]). It would have been obvious to a person of ordinary skill to include the well-known Gaussian Process Regression (GPR) along with the other machine learning network, in order to yield the predicted results of generating accurate seismic image or geographic map, yet with higher accuracy (KSR).
Regarding Claim 17, Combination of Li and Vasudevan teaches the system of claim 16,
Li is silent on wherein the Gaussian process regression is a stochastic process comprising at least one mean function, at least one kernel, and at least one hyper-parameter.
However, Vasudevan teaches wherein the Gaussian process regression is a stochastic process comprising at least one mean function, at least one kernel, and at least one hyper-parameter. (Vasudevan, [0005] Gaussian process-based terrain modeling provides an approach that enables a multi-resolution representation of
space, incorporates and manages uncertainty in a statistically sound way, and can handle spatially correlated data in an appropriate manner. Gaussian Processes (GPs) are stochastic processes based on the normal (Gaussian) distribution and can be used to good effect as a powerful nonparametric learning
technique for spatial modeling. [0010] To generate a Gaussian process model the method may include using the first dataset to learn optimized kernel hyperparameters and a first noise hyperparameter, and using the learnt kernel hyperparameters and second dataset to learn an optimized second noise hyperparameter).
It would have been obvious to a person having ordinary skill in the
art before the effective filing date to modify Li’s method for predicting optimized velocity profile model to incorporate a machine learning network with the Gaussian process regression as taught by Vasudevan and obtain an accurate terrain model by incorporating uncertainty and correlation of data and generate geographic map (Vasudevan, [0002]-[0005]). It would have been obvious to a person of ordinary skill to include the well-known Gaussian Process Regression (GPR) along with the other machine learning network, in order to yield the predicted results of generating accurate seismic image or geographic map, yet with higher accuracy (KSR).
Regarding Claim 19, Combination of Li and Vasudevan teaches the system of claim 16,
Li further teaches wherein forming the depth-domain seismic image (LI, Figure 7, 8, [0075] “seismic data may be processed in a technique called "depth imaging" to form an image (e.g.,a depth image) of reflection amplitudes in a depth domain for a particular target structure (e.g., a geologic subsurface region of interest)”) comprises determining a confidence interval, based on the uncertainty estimate, for at least one depth in the depth-domain seismic image, and wherein the confidence interval indicates a reliability of the depth-domain seismic image at the at least one depth and is based on a mean prediction bounded by a variance (Li, [0152],by one or more ML processes. For example, consider a first ML process that is trained for a first pass through a seismic cube to generate ML interpretation results followed by a second ML process where an interpreter further refines a portion of the ML interpretations as part of the second ML process where the refinement includes labeling the portion for to train a ML model that can then be applied to one or more other portions of the seismic cube, optionally as the seismic cube prior to the first pass or after the first pass. In such an approach, the interpreter's effort can be limited to a particular portion or portions of a seismic cube as pre-processed with some level of confidence whereby the interpreter's effort improves that level of confidence for applying to one or more other portions of the seismic cube”).
Regarding Claim 20, Combination of Li and Vasudevan teaches the system of claim 16,
Li further teaches further comprising:
a wellbore planning system configured to plan a wellbore path to intersect the hydrocarbon reservoir (Li, Figures 1-2, [0054], [0054] As to the convention 240 for dip, as shown, the three-dimensional orientation of a plane can be defined by its dip and strike. Dip is the angle of slope of a plane from a horizontal plane (e.g., an imaginary plane) measured in a vertical plane in a specific direction. Dip may be defined by magnitude (e.g., also known as angle or amount) and azimuth (e.g., also known as direction). As shown in the convention 240 of FIG. 2,”).
; and a wellbore drilling system configured to drill a second wellbore guided by the planned wellbore path. (Li, Figure 1, a workflow component 144. [0032] output from the simulation component 120 may be input to one or more other workflows, as indicated by a workflow component 144. See Figure 19)
Conclusions
Citation of Pertinent Prior Art
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Roy et al. (US 2020/0088897 A1) recites “Systems and methods are disclosed that include generating reservoir property profiles corresponding to reservoir properties for pseudo wells based on reservoir data, generating seismic attributes for the pseudo wells, and training a machine learning model by comparing the reservoir property profiles against the seismic attributes. In this manner, the machine learning model may be used to predict reservoir properties for use with seismic exploration above a region of a subsurface that contains structural or stratigraphic features conducive to a presence, migration, or accumulation of hydrocarbons” (abstract).
Salman et al. (US 2019/0383965 A1) recites “A method can include selecting a type of geophysical data; selecting a type of algorithm; generating synthetic geophysical data based at least in part on the algorithm; training a deep learning framework based at least in part on the synthetic geophysical data to generate a trained deep learning framework; receiving acquired geophysical data for a geologic environment; implementing the trained deep learning framework to generate interpretation results for the acquired geophysical data; and outputting the interpretation results” (abstract).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to DILARA SULTANA whose telephone number is (571)272-3861. The examiner can normally be reached Mon-Fri, 9:00AM-6 PM.
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, EMAN ALKAFAWI can be reached on (571) 272-4448. 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.
/DILARA SULTANA/Examiner, Art Unit 2858
/EMAN A ALKAFAWI/Supervisory Patent Examiner, Art Unit 2858 3/9/2026