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 .
Amendments
This action is in response to amendments filed April 6th, 2026, in which Claims 1, 2, 9, 15, and 19-20 are amended. No claims are cancelled nor added. The amendments have been entered, and Claims 1-20 are currently pending.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1-6, 8-12, and 14-20 are rejected under 35 U.S.C. 103 as being unpatentable over Liu et al., “Multi-resolution Graph Neural Networks for PDE approximation,” in view of Hajgato et al., “Accelerating Convergence of Fluid Dynamics Simulations with Convolutional Neural Networks.”
Regarding Claim 1, Liu teaches a method comprising, in simulation software (Liu, pg. 2, “we introduce two such architectures for GNNs in the context of PDE simulations”), (a) receiving, in analysis engine of the simulation software, two or more numerical models for one or more differential equations to be evaluated by the simulation software, the two or more numerical models including a first numerical model and second numerical model, wherein the two or more numerical models are converging to an exact solution of the one or more differential equations (Liu, pg. 6, Fig. 1, the combination of the “coarse-to-fine mesh” grids along with the input f are converging to an exact solution u of the PDE and are numerical models, see pg. 7, 4th paragraph, “The Input Functions … resulting in 25 control parameters”); (b) generating, by the analysis engine, from the first numerical model, a first numerical solution in a first grid patch in which the first grid patch corresponds to a local domain of dependence, and wherein the first grid patch has a first grid resolution; (c) generating, by the analysis engine, from the second numerical model, a second numerical solution in a second grid patch in which the second grid patch corresponds to the local domain of dependence, wherein the second grid patch has a second resolution that is different from the first grid patch (Liu, pg. 6, Fig. 1, the “coarse-to-fine mesh” grids along with pg. 7, 4th paragraph, “The Input Functions … resulting in 25 control parameters” wherein the input f on each of the two grids are two solutions calculated as “a linear combination of eight Gaussian functions … They are chosen uniformly in domain-dependent intervals” i.e. solutions of values for each grid) (d) generating a high-fidelity numerical solution value at a space-time location determined by the local domain of dependence for the one or more differential equations using a trained neural network receiving as its input the first numerical solution generated in step (b) and the second numerical solution generated in step (c) (Liu, pg. 6, Fig. 1) and repeating steps (b) and (c) to provide as input to determine respective high-fidelity numerical solution value for one or more varying local domains of dependence (Liu, pg. 9-10, Fig. 2, the experiment is performed on “square domain,” “donut shape,” “5.2 Variable domains”) wherein the trained neural network is trained on one or more … [training examples] wherein each of the [training examples] includes (i) an exact or nearly exact solution to the one or more differential equations … and (ii) solution to the two or more numerical models (Liu, pg. 7, “Ground Truth and Los Function … the solution provided by FEniCS on the finest mesh will be considered as the ground truth” & pg. 8, last paragraph, “For each type of domain, a training set of 42 000 examples is generated” where the examples need the input and expected output of Fig. 1, including the input solutions f).
Liu does not teach that the differential equation analyses performed by their invention are design iteration analyses. However, Hajgato teaches an iteration of design analyses of systems, wherein the required PDEs have approximations provided by neural networks (Hajgato, pg. 231, Figs. 1 and 2). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the method of Liu to produce solutions for an engineering design iteration process, such as that of Hajgato. The motivation to do so is that Liu quickly produces the solutions necessary for the design iterations (Hajgato, pg. 231, 1st column, “time-consuming numerical simulation … CFD simulations of an optimization session can be accelerated without losing accuracy”).
Regarding Claim 2, the Liu/Hajgato combination of Claim 1 teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). Liu further teaches wherein training the trained neural network comprises computing the difference between (i) an output of the neural network and (ii) the exact or nearly exact solution for use in a loss function (Liu, pg. 7, 2nd-to-last paragraph, “the solution provided by FEniCS on the fines mesh will be considered as the ground truth. The loss function is defined by the Mean Absolute Error (MAE) between the network output and this ground truth”).
Regarding Claim 3, the Liu/Hajgato combination of Claim 1 teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). The combination has already been shown to teach repeating steps (b), (c), and (d) for an additional design iteration analysis of the plurality of design iteration analyses (Hajgato, pg. 231, Figs. 1 and 2 iterates the process).
Regarding Claim 4, the Liu/Hajgato combination of Claim 1 teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). Liu further teaches interpolating outputs of the two or more numerical models to align common grid points of the first and second grid patches (Liu, pg. 6, 2nd paragraph, “The sampling operators from mesh M1 to mesh M2 use the k-nearest interpolation proposed in PointNet++” on the solutions f of the two numerical models).
Regarding Claim 5, the Liu/Hajgato combination of Claim 1 teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). The combination has already been shown to teach wherein the plurality of design iteration analyses correspond to an engineering model, a scientific model, or financial model (Hajgato, pg. 230, 2nd column, 2nd paragraph, “In aerodynamic shape design,”) wherein the two or more numerical models are used by the trained neural network to generate high-fidelity solutions (Liu, pg. 6, Fig. 1) for an additional design iteration analysis of the plurality of design iteration analyses (Hajgato, pg. 231, Figs. 1,2 & Liu, pg. 9-10 multiple PDE solutions for different f initial conditions for each domain).
Regarding Claim 6, the Liu/Hajgato combination of Claim 1 teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). Liu further teaches wherein the trained neural network is part of a neural network engine, the neural network engine comprising logic or instructions to direct steps (a) – (d) (Liu, pg. 8, 4th paragraph, “Implementation is done in PyTorch”).
Regarding Claim 8, the Liu/Hajgato combination of Claim 6 teaches the method of Claim 6 (and thus the rejection of Claim 6 is incorporated). Liu further teaches wherein the neural network engine is executing on a computing device execution the analysis engine (Liu, pg. 8, 4th paragraph, “Implementation is done in PyTorch” which inherently is performed on a computer).
Claims 9, 10, 11, 12, and 14 recite a system comprising one or more processors; and a memory having instructions stored thereon, wherein the instructions, as part of a simulation software, when executed by a processor, cause the processor to perform the steps of the methods of Claims 1, 2, 5, 6, and 8, respectively. As Liu executes their methods in PyTorch, (Liu, pg. 8, 4th paragraph, “Implementation is done in PyTorch”) and thus inherently on a computer, Claims 9, 10, 11, 12, and 14 are rejected for reasons set forth in the rejections of Claims 1, 2, 5, 6, and 8, respectively.
Claims 15, 16, 17, and 18 recite a non-transitory computer-readable medium having instructions stored thereon, wherein the instructions, when executed by a processor, cause the processor to perform the steps of the methods of Claims 1, 3, 5, and 6, respectively. As Liu executes their methods in PyTorch, (Liu, pg. 8, 4th paragraph, “Implementation is done in PyTorch”) and thus inherently on a computer, Claims 15, 16, 17, and 18 are rejected for reasons set forth in the rejections of Claims 1, 3, 5, and 6, respectively.
Regarding Claim 19, the Liu/Hajgato combination of Claim 18 teaches the non-transitory computer-readable medium of Claim 18 (and thus the rejection of Claim 18 is incorporated). Liu further teaches wherein the neural network is implemented in a library file that can be coupled to the analysis engine (Liu, pg. 8, 4th paragraph, “Implementation is done in PyTorch”).
Regarding Claim 20, the Liu/Hajgato combination of Claim 18 teaches the non-transitory computer-readable medium of Claim 18 (and thus the rejection of Claim 18 is incorporated). The combination further teaches wherein the neural network engine is natively implemented in simulation software comprising the analysis engine (Liu, pg. 8, 4th paragraph, “Implementation is done in PyTorch” in combination with Hajgato, pg. 231, Figs. 1 and 2).
Claims 7 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Liu in view of Hajgato, and further in view of Dirac, US PG Pub 2015/0379424.
Regarding Claim 7, the Liu/Hajgato combination of Claim 6 teaches the method of Claim 6 (and thus the rejection of Claim 6 is incorporated). The combination does not teach, but Dirac does teach, wherein the neural network engine is executing on a cloud infrastructure (Dirac, [0035], “A wide variety of machine learning algorithms may be supported natively … including neural network algorithms” & [0080], “a number of different programming languages … may be supported … including … Python” & [0037], “a machine learning service implemented by using a plurality of network-accessible services of a provider network … cloud-based computing”). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to use the cloud-based machine learning as a service system of Dirac to implement the neural network computation of the Liu/Hajgato combination. The motivation to do so is that “many machine learning techniques can be computationally intensive .. it may not always be advisable or viable for business organizations to build out their own machine learning computational facilities” (Dirac, [0002]) and using a MLaaS solves this problem.
Claim 13 recites a system comprising one or more processors; and a memory having instructions stored thereon, wherein the instructions, as part of a simulation software, when executed by a processor, cause the processor to perform the steps of the methods of Claim 7. As Liu executes their methods in PyTorch, (Liu, pg. 8, 4th paragraph, “Implementation is done in PyTorch”) and thus inherently on a computer, Claim 13 is rejected for reasons set forth in the rejection of Claim 7.
Response to Arguments
Applicant’s arguments filed April 6th, 2026 have been fully considered, but are not fully persuasive.
Due to claim amendments, the Claim Objections and 35 U.S.C. 112(b) rejections of the previous office action have been withdrawn.
Applicant’s arguments regarding the 35 U.S.C. 103 rejections of the claims have been fully considered, but are unpersuasive.
Applicant argues with respect to Claim 1 that Liu fails to teach two different solutions, arguing that “Liu discloses using this exact same source term on different meshes … ‘a projection of the source term on the current mesh’ …”. The rejection of the previous office action has identified the two different numerical models as “the combination of the coarse-to-fine mesh grids along with the input f” and as each model has a different mesh grid, the values of f on the two different mesh grids are two different solutions. The current rejection clarifies this point. That is, there is not one solution (merely f), but two distinct solutions from the two distinct models (the projections of f on the two distinct meshes). The values along the two different meshes, each serving as distinct inputs to the neural network model (see pg. 6, Fig. 1) are distinct sets of values, thus distinct solutions which fall within the claim scope.
Applicant’s arguments regarding the other independent and dependent claims rely upon the features argued with respect to Claim 1, and are thus also unpersuasive.
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 BRIAN M SMITH whose telephone number is (469)295-9104. The examiner can normally be reached Monday - Friday, 8:00am - 4pm Pacific.
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, Kakali Chaki can be reached at (571) 272-3719. 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.
/BRIAN M SMITH/Primary Examiner, Art Unit 2122
Prosecution Insights