Detailed Action
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Application Status
Claims 1-20 are presented for examination based on the amendment filed 05/18/2026, claims 1, 3, 10, are 16 are amended.
Response to Arguments
The 35 USC 101 rejections have been withdrawn in view of amendments and persuasive arguments presented by the applicant. The amended claims add a practical application.
The 35 USC 103 rejection is maintained and modified to address amended claim language. Arguments with respect to the amended claims are moot as being directed to newly added claimed features. See below for the modified rejections.
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 1-10 and 15-20 are rejected under 35 U.S.C. 103 as being unpatentable over Usadi et al., US 2013/0096900 A1 (Usadi) in view of Zandbergen, Predicting the optimal CFL number for pseudo time-stepping with machine learning in the COMSOL CFD module (Zandbergen) in further view of Sheth et al., Intelligent Time-Stepping for Practical Numerical Simulation. (Sheth).
Claim 1.
Usadi teaches A method for accelerating numerical solution of a differential equation representing fluid flow in porous media associated with hydrocarbon well environments, the method comprising: (Usadi [0079) “The equations that describe the evolution of state variables such as pressure and composition for each sub region 502 may be represented by a matrix structure characterized by a set of physical, geometrical, or numerical parameters based on the geological characteristics of the matrix structure, such as rock porosity, phase permeability, and the like. A first region may include an injector well 504, and an nth region may include a producer well”
obtaining input data associated with a previous timestep of a numerical solver operating on the differential equation; (Usadi [0097]) “In the above formula, Kv,effective(tn ) equals the coarse scale approximation of the coarse grid cell at a time step, n. The term K, n 700 v equals the discretized phase permeability at each fine grid cell, and the term Sv, equals the phase saturation at each fine grid cell. For two dimensional or three-dimensional models, the effective phase permeability can be written as a tensor as shown in Eqn. 8.” {Examiners note: Conditions a learned function on one prior time step (t^(n-1) See equation 8.}
Usadi does not explicitly teach, but Zandbergen teaches wherein the input data comprises a state of the numerical solver including at least one of residual error, pressure change, or a previous timestep size; From the above list of alternatives the Examiner is selecting "pressure change". (Zandbergen Abstract) “The local data consist of the velocities, pressure and residuals, as well as the cell Reynolds number and the element edge lengths.”
such that the numerical solver exhibits improved convergence and a reduced number of iterations to obtain the simulation output relative to use of a fixed or non-adaptive timestep size. (Zandbergen Abstract) “The convergence is assumed to be accelerated if the network predictions result in convergence in fewer nonlinear iterations compared to solvers that use each one of the two CFL numbers in COMSOL.”
Usadi and Zandbergen are analogous to the claimed invention because they are from the same field of endeavor of learning based simulation.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Usadi and Zandbergen before him or her, to modify the numerical time step solver of Usadi with the prediction methods of Zandbergen to help with the reusability and flexibility of allowing pseudo functions to reach their potential as suggest in Usadi 0009.
Modified Usadi with Zandbergen does not explicitly teach, but Sheth teaches predicting, by a machine learning model, a current timestep size for the numerical solver from the previous timestep to a current timestep immediately following the previous timestep; (Sheth Abstract) “Typical time-step selectors use a limited set of features to heuristically predict the size of the next timestep. . . We have found that history match and uncertainty /optimization studies benefit most from the static approach while the dynamic approach produces optimum step-sizes for prediction studies. We use a confidence monitor to manage the ML time-step selector at runtime. If the confidence level falls below a threshold, we switch to traditional heuristic method for that time-step. This avoids any degradation in the performance when the model features are outside the training space.” (See also pg. 4 “Newton iterations and time-step sizes”)
wherein the current timestep size is determined based on the state of the numerical solver; (Sheth Abstract) “uses machine-learning (ML) techniques to analyze the mathematical and physical state of the system and predict time-step sizes which are large while still being efficient to solve.”
and executing the numerical solver using the current timestep size on the differential equation to generate a simulation output for the current timestep wherein the current timestep size controls advancement of the numerical solver (Sheth abstract) “We present the application of these workflows in a commercial reservoir simulator using distinct types of simulation model including black oil, compositional and thermal steam-assisted gravity drainage (SAGD).” {EXAMINERS NOTE: Under BRI, a timestep selector integrated into a reservoir simulator for runtime timestep simulation necessarily provides the selected time step for execution of the simulator. Sheth’s ML predicted quantity is expressly the next simulator timestep size.}
and is adaptively varied across successive timesteps based on changes in the state of the numerical solver, (Sheth abstract) “. . . data-driven time-step selection algorithms. We propose two workflows --static and dynamic --- that use a diverse set of physical ( e.g., well data) and mathematical ( e.g, CFL) features to build a predictive ML model. This can be pre-trained or dynamically trained to generate an inference model. The trained model can also be reinforced as new data becomes available and efficiently used for transfer learning.”
Usadi, Zandbergen, and Sheth are analogous to the claimed invention because they are from the same field of endeavor of learning based simulation.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Usadi, Zandbergen, and Sheth before him or her, to modify the numerical time step solver of Usadi with the prediction methods of Zandbergen and the timestep prediction of Sheth to “avoids wasted non-linear and linear equation set-up work when the time-step is too small l and avoids highly non-linear systems that take many iterations to solve.” (Sheth Abstract)
Claim 2.
Modified Usadi teaches The method of claim 1, wherein the machine learning model is an artificial neural network (ANN). (Usadi 0012) “A reservoir simulator was used to generate training sets for the Neural Networks. And for these cases, the authors were able to reproduce the narrowly modeled behavior response via the ANN.”
Claim 3.
Modified Usadi with Zandbergen teaches The method of claim 2, wherein the ANN makes the prediction of the current timestep size based on at least one selected from a group consisting of a previous timestep size, pressure changes, and residual errors wherein the residual errors correspond to a convergence status of the numerical solver. (Zandbergen Abstract) “The local data consist of the velocities, pressure and residuals, as well as the cell Reynolds number and the element edge lengths”
Claim 4.
Modified Usadi teaches The method of claim 2, further comprising training the ANN. (Usadi 0011) “Artificial neural networks (ANNs) were trained to predict peak injection volumes and volumes of produced oil and gas at three and seven years after the commencement of injection”
Claim 5.
Modified Usadi teaches The method of claim 4, wherein the training is specific to one hydrocarbon field, using training data associated with the one hydrocarbon field only. (Usadi 0011) “Artificial neural networks (ANNs) were trained to predict peak injection volumes and volumes of produced oil and gas at three and seven years after the commencement of injection” (0087) “In some embodiments, the boundary condition values may be specified based on known conditions of an actual reservoir.”
Claim 6.
Modified Usadi teaches The method of claim 4, wherein the training is performed using training data for a feature set, and wherein the training further comprises reducing the feature set to features relevant to the prediction of the current timestep size. (Usadi 0082) “Examples of parameters that may serve as lookup keys for sub-regions and their surrogate model solutions include, but are not limited to row sum or column sum vectors, diagonal vectors, L1, L2, LN norms of these vectors, and so on.” {Examiners note: Constructs a “feature set” (vectors/norms/physical parameters) used as reduced descriptors.}
Claim 7.
Modified Usadi teaches The method of claim 4, wherein the training further comprises serializing the machine learning model. (Usadi 0089) “At block 620, the solution surrogate may be stored to a database of solution surrogates. Each solution surrogate in the database may be paired with the corresponding physical, geometrical, or numerical parameters used to generate the training set 416. In this way, the solution surrogate may be reused for future reservoir simulations based on a degree of similarity between the physical, geometrical, or numerical parameters used to the generate the solution surrogate and the physical, geometrical, or numerical parameters of subsequent sub regions 102 used for future reservoir simulations” {EXAMINERS NOTE: Serializing is being view as storing the trained model in a persistent form for later reuse.}
Claim 8.
Modified Usadi with Zandbergen teaches The method of claim 1, wherein the numerical solver uses Newton’s method. (Zandbergen Pg 1 Paragraph 2) “And there are a lot of nonlinear solver methods, such as Automatic Newton and Newton Constant.”
Claim 9.
Modified Usadi teaches The method of claim 1, further comprising a preprocessing of the input data, the preprocessing comprising at least one selected from a group consisting of data smoothing and data scaling. (Usadi 0119) “In some embodiments, the constitutive relationship used for training may be that of the fine grid solution after it has been averaged or smoothed.” (0009) “This is handled by both scaling up the absolute permeability and assuming that relative permeability scales uniformly in the volume of the coarse grid cell,”
Claim 10.
Usadi teaches A system, comprising: a plurality of computing systems configured to perform operations comprising: obtaining input data associated with a previous timestep of a numerical solver operating on the differential equation; (Usadi [0097]) “In the above formula, Kv,effective(tn ) equals the coarse scale approximation of the coarse grid cell at a time step, n. The term K, n 700 v equals the discretized phase permeability at each fine grid cell, and the term Sv, equals the phase saturation at each fine grid cell. For two dimensional or three-dimensional models, the effective phase permeability can be written as a tensor as shown in Eqn. 8.” {Examiners note: Conditions a learned function on one prior time step (t^(n-1) See equation 8.}
Usadi does not explicitly teach, but Zandbergen teaches wherein the input data comprises a state of the numerical solver including at least one of residual error, pressure change, or a previous timestep size; From the above list of alternatives the Examiner is selecting "pressure change". (Zandbergen Abstract) “The local data consist of the velocities, pressure and residuals, as well as the cell Reynolds number and the element edge lengths.”
such that the numerical solver exhibits improved convergence and a reduced number of iterations to obtain the simulation output relative to use of a fixed or non-adaptive timestep size. (Zandbergen Abstract) “The convergence is assumed to be accelerated if the network predictions result in convergence in fewer nonlinear iterations compared to solvers that use each one of the two CFL numbers in COMSOL.”
Usadi and Zandbergen are analogous to the claimed invention because they are from the same field of endeavor of learning based simulation.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Usadi and Zandbergen before him or her, to modify the numerical time step solver of Usadi with the prediction methods of Zandbergen to help with the reusability and flexibility of allowing pseudo functions to reach their potential as suggest in Usadi 0009.
Modified Usadi with Zandbergen does not explicitly teach, but Sheth teaches predicting, by a machine learning model, a current timestep size for the numerical solver from the previous timestep to a current timestep immediately following the previous timestep; (Sheth Abstract) “Typical time-step selectors use a limited set of features to heuristically predict the size of the next timestep. . . We have found that history match and uncertainty /optimization studies benefit most from the static approach while the dynamic approach produces optimum step-sizes for prediction studies. We use a confidence monitor to manage the ML time-step selector at runtime. If the confidence level falls below a threshold, we switch to traditional heuristic method for that time-step. This avoids any degradation in the performance when the model features are outside the training space.” (See also pg. 4 “Newton iterations and time-step sizes”)
wherein the current timestep size is determined based on the state of the numerical solver; (Sheth Abstract) “uses machine-learning (ML) techniques to analyze the mathematical and physical state of the system and predict time-step sizes which are large while still being efficient to solve.”
and executing the numerical solver using the current timestep size on the differential equation to generate a simulation output for the current timestep wherein the current timestep size controls advancement of the numerical solver (Sheth abstract) “We present the application of these workflows in a commercial reservoir simulator using distinct types of simulation model including black oil, compositional and thermal steam-assisted gravity drainage (SAGD).” {EXAMINERS NOTE: Under BRI, a timestep selector integrated into a reservoir simulator for runtime timestep simulation necessarily provides the selected time step for execution of the simulator. Sheth’s ML predicted quantity is expressly the next simulator timestep size.}
and is adaptively varied across successive timesteps based on changes in the state of the numerical solver, (Sheth abstract) “. . . data-driven time-step selection algorithms. We propose two workflows --static and dynamic --- that use a diverse set of physical ( e.g., well data) and mathematical ( e.g, CFL) features to build a predictive ML model. This can be pre-trained or dynamically trained to generate an inference model. The trained model can also be reinforced as new data becomes available and efficiently used for transfer learning.”
Usadi, Zandbergen, and Sheth are analogous to the claimed invention because they are from the same field of endeavor of learning-based simulation.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Usadi, Zandbergen, and Sheth before him or her, to modify the numerical time step solver of Usadi with the prediction methods of Zandbergen and the timestep prediction of Sheth to “avoids wasted non-linear and linear equation set-up work when the time-step is too small l and avoids highly non-linear systems that take many iterations to solve.” (Sheth Abstract)
Claim 15.
Claim 15 is rejected as being substantially similar to claim 2 albeit for a system. It is rejected under the same rationale.
Claim 16.
Claim 16 is rejected as being substantially similar to claim 3 albeit for a system. It is rejected under the same rationale.
Claim 17.
Claim 17 is rejected as being substantially similar to claim 4 albeit for a system. It is rejected under the same rationale.
Claim 18.
Claim 18 is rejected as being substantially similar to claim 5 albeit for a system. It is rejected under the same rationale.
Claim 19.
Claim 19 is rejected as being substantially similar to claim 6 albeit for a system. It is rejected under the same rationale.
Claim 20.
Claim 20 is rejected as being substantially similar to claim 7 albeit for a system. It is rejected under the same rationale.
Claims 11-14 are rejected under 35 U.S.C. 103 as being unpatentable over Usadi et al., US 2013/0096900 A1 (Usadi) in view of Zandbergen, Predicting the optimal CFL number for pseudo time-stepping with machine learning in the COMSOL CFD module (Zandbergen) in further view of Sheth et al., Intelligent Time-Stepping for Practical Numerical Simulation. (Sheth), and in further view of Rosebrock Building a simple Keras deep learning rest api. (Rosebrock).
Claim 11.
Modified Usadi does not explicitly teach, but Rosebrock teaches The system of claim 10, wherein a first of the plurality of computing systems is a Flask server, and wherein a second of the plurality of computing systems is a Flask client. (Rosebrock pg. 1) “How to use the Flask web framework to create an endpoint for our API. How to make predictions using our model, JSON-ify them, and return the results to the client” {Examiners note: Flask hosted endpoint (server) and a separate caller (client). Under BRI flask client can read on a client process configured to invoke flask endpoints.}
Usadi, Zandbergen, and Sheth are analogous to the claimed invention because they are from the same field of endeavor of learning-based simulation. Rosebrock is analogous to the claimed invention because they are from the same field of endeavor of flask api.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Usadi, Zandbergen, Sheth and Rosebrock before him or her, to modify the numerical time step solver of Usadi with the prediction methods of Zandbergen, and the timestep prediction of Sheth and the flask server and client of Rosebrock to help with the reusability and flexibility of allowing pseudo functions to reach their potential as suggest in Usadi 0009.
Claim 12.
Modified Usadi with Rosebrock teaches The system of claim 11, wherein the Flask server forwards a request for the (Rosebrock pg. 1) “How to use the Flask web framework to create an endpoint for our API. How to make predictions using our model, JSON-ify them, and return the results to the client” {Examiners note: While Rosebrock does not teach the flask server request being the current timestep size from the numerical solver, it would have been obvious to combine with Zandbergen to forward the timesteps of Zandbergen from a flask server to flask client.}
Claim 13.
Modified Usadi with Zandbergen and Rosebrock teaches The system of claim 11, wherein the(Zandbergen pg. 9 Paragraph 1) “A neural network is used to predict local CFL numbers for pseudo time-stepping” {Examiners note: While Zandbergen does not explicitly use a flask client, it would have been obvious to combine with Rosebrock to have the flask client of Rosebrock perform the prediction of Zandbergen}
Claim 14.
Claim 14 is rejected as being substantially similar to claim 8 albeit for a system and the numerical solver utilizing Newton’s method. It is rejected under the same rationale.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JOHN DAVID HAGLER whose telephone number is (703)756-1339. The examiner can normally be reached Monday - Friday 10am- 6pm.
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, Rehana Perveen can be reached at 5712723676. 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.
/JOHN DAVID HAGLER/Examiner, Art Unit 2189
/REHANA PERVEEN/Supervisory Patent Examiner, Art Unit 2189
Prosecution Insights