DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Responsive to the communication dated 2/5/2026.
Claims 1, 3, 12, 13 are amended.
Claims 1 – 21 are presented for examination.
Final Action
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.
Response to Arguments
Claim Rejections - 35 USC § 101
The Applicant asserts that the identified abstract idea of the independent claims (i.e., claim 1) may not be found to be a mental process if the human mind is not equipped to perform the claim limitation. The Applicant asserts that performing a physics-based simulation cannot practically be performed by the human mind, “at least because a human mind is not practically equipped to use ‘model parameters’ to generate simulated predictions using a ‘physics-based simulation model’, which necessarily requires some computer technology to make use of “model parameters” within the “simulation model’ to generate predictions.”
Therefore, at issue is whether a human can perform a physic-based simulation using model parameters within a simulation model.
In response the argument is not persuasive.
First, the Applicant is asserting that the claim may not be directed towards a mental process because the math is too complicated to be performed in a human mind, and this is an error because the rejection did not assert that the claim is directed towards a mental process. Rather the rejection indicated that under STEP 2A PRONG ONE, the claim recites a mathematical abstract idea. Accordingly, arguing that the claim is not a mental process is incorrect because the Applicant should have directed to argument towards the mathematical abstract idea instead.
Second, the specification clearly discloses that models are mathematical and that model parameters are simply variables in a mathematical equation. For example, the instant specification state:
Par 2: “… a proxy model is a mathematically or statistically defined model that replicates or approximates the response of the output of a full-scale simulation model for some selected input parameters…”
Par 5: “… the physics-based simulation model generates output data comprising simulated predictions that are calculated using the model parameters…”
Par 16: “model parameters are used by a physics-based simulation model together with variables for simulation (making predictions for) a physics process using input data from a plurality of operational data sources. The model parameters in the case of a polynomial equation that can be a first order or higher order equation are the coefficients (constants) of the equation or weights in the case of a neural network…”
Accordingly, the physics-based simulation model is a polynomial equation with coefficients.
Take for example, a fundamental physics-based simulation model: quadratic kinematic equation for displacement (i.e., the dropping an object from rest). The polynomial equation is:
d = V0t +( ½)at2
d: displacement (i.e., distance fallen, in meters or feet)
V0: initial velocity (in m/s or ft/s)
a: Acceleration (due to gravity, approx. 9.8 m/s2 or 32 ft/s2)
t: time (in seconds)
A human child is expected to be able to perform such calculations using only a paper and pencil as part of their remedial science education.
For example:
Problem: A rock is dropped from rest (𝑣0=0) from the top of a cliff. How far does it fall in 3 seconds? (Assume 𝑎=9.8 m/s2)
1. Set up the equation:Because the rock is "dropped from rest," the initial velocity (𝑣0) is 0, which simplifies the equation:𝑑=0⋅𝑡+12⋅9.8⋅𝑡2
PNG
media_image2.png
1
1
media_image2.png
Greyscale
𝑑=4.9𝑡2
2. Plug in the time (𝑡=3):𝑑=4.9⋅(3)2
𝑑=4.9⋅9
d=44.1 meters <=ANSWER (i.e., the prediction)
Therefore, the claim was found to recite a mathematical abstract idea and further a human’s mind is fully capable of performing mathematical calculation. See MPEP 2106.04(a)(2).
Also, while the claim further recites “using an artificial intelligence (AI)-based system including at least one AI-based proxy model that is responsive to receiving an update of the input data, processing the input data and the output data using the at least one AI-based proxy model” these elements simply further recite additional abstract mathematical concepts. These elements merely recite that the AI model (mathematical calculation) has two types of input data. The first time is the input data from operational data and the other type of input data is the output of the physics-based calculation. The AI-based proxy model is merely a series of mathematical calculations that produce a mathematical output.
As evidence that an AI based proxy model is simply a mathematical calculation see:
Cheng_2019 (Polynomial Regression as an Alternative to Neural Nets, April 11, 2019):
Abstract: “… here we present a simple analytic argument that NNs are in fact essentially polynomial regression models (PR)…”
Patwari_2021 (A Neural Network is Just a Glorified Math Equation, Medium Jun 27, 2021)
Shi_2022 (How to Define A Neural Network as A Mathematical Function, Medium TDS Archive, Feb 22, 2022)
page 2: “… a k-layer neural network is a mathematical function f, which is a composition of multivariate functions: fr, f2, …, fk, and g…”
Siemens_2020 (The Elegant Math Behind Neural Networks, m-siemens.de, December 7, 2020)
page 1: “… it’s a computational model that combines simple mathematical functions into a complex calculation graph. Their ability to approximate arbitrary functions makes them frequently used tool…”
Swiffy_2016 (Are the “weights” inside a neural network actually “terms” for a polynomial?” (MATHEMATICS, May 13, 2016)
page 4: “… yes, the weights of a neural network are similar to a function like f(x1) = w1X1 + W0, where you adjust W0, W1 to make the output as similar to some given pairs…”
Accordingly, the Examiner finds that the claim merely recites a mathematical abstract idea in which a first polynomial outputs a numeric result to a second mathematical calculation (AI) which in turn merely outputs a mathematical result.
The Applicant further asserts that the claims integrate the abstract idea into a practical application under STEP 2A PRONG 2 because the claims improve the functioning of a computer or improve a technical field. The Applicant then, cites paragraph 4 of the instant specification, and asserts that the specification provides sufficient details such that one of ordinary skill in the art would recognize how the claims improve the technology. In particular the Applicant asserts that current commercial software vendors do not provide physic-based simulations that provide uncertainty quantification that incorporate prior knowledge, joint parameter tuning independent of the governing laws and that they do not support connecting physics-based simulation models from different vendors.
In response the Examiner notes that while paragraph 4 of the instant specification states that proxy models do not provide uncertainty quantification or the ability to incorporate prior knowledge, or joint parameter tuning, or proxy model training, these elements themselves appear to be mathematical in nature. Improvements to a mathematical abstract idea itself is not a practical application. Moreover, the assertion that the instant claims supports physics-based simulation models from different vendors is not supported by the claims themselves. There are no elements in the claim that articulate/result in the use of a first vendors model from a first vendor in a second commercial software. The Applicant improperly reading into the claim that “a proxy prediction for at least one selected prediction from the simulated prediction or a variable derived from the at least one selected prediction as a replacement for or as a supplement to the at least one selected prediction or the variable derived from the at least one selected prediction” means to replace a first vendors model with an AI model. This is an improper reading of the claim. All the claim requires is that a first abstract mathematical calculation is replace with a second abstract mathematical equation. There is nothing in the claim regarding vendor software or anything else of that nature.
The Applicant then states that one of ordinary skill in the art would recognize the claims as being directed towards improvements to the functioning of a computer because the claim recites “using an artificial intelligence (AI)-based system.” The Applicant asserts that the AI addresses the problems described in the specification (i.e, paragraph 4) because the AI model can replace and be used instead of the physic-based model and in this way the AI-model can be used in place of simulation model from different vendors.
In response the argument is not persuasive. As outlined above, the physics-based model is a mathematical calculation (i.e., a polynomial with coefficients). The Applicant is asserting that there is a practical application because the polynomial with coefficients can be replace with an AI-based proxy model. This is not persuasive because, as outlined above, the AI-based proxy model is merely another mathematical model/calculation. Indeed, Cheng_2019, and Swiffy_2016 indicate that an AI-based proxy model is merely a different polynomial with coefficient. Merely replacing a first mathematical calculation with a second mathematical calculation isn’t a practical application because, even if the second mathematical calculation is improved in some way the improvement may not be to the abstract idea itself. The claim does not recite any elements the rely on or use the output of the AI-based proxy model. Without additional elements the rely upon or use the abstract idea there is not improvement to a technology other than the abstract idea itself.
Additionally, the Applicant asserts that under STEP 2B the claims recite elements which are significantly more than the abstract idea because the claim recites to provide input data from a plurality of different sources of operational data and those elements.
In response this argument is not persuasive. MPEP 2106.05(g) indicates that insignificant extra-solution activity is not significantly more than the abstract idea. Indeed, MPEP 2106.05(g) explicitly identifies “mere data gathering” and indicates that such activities are not significantly more than the abstract idea. The MPEP provides several examples of insignificant extra-solution activity.
One example cited by the MPEP is “Performing clinical tests on individuals to obtain input for an equation.” This is analogous to the claimed invention because obtaining test results from individuals gathers data from a plurality of different sources (i.e., different individuals) and then the data is used in mathematical equations. Accordingly, these elements of the claim are not significantly more than the abstract idea.
Additionally, the Applicant also asserts that using an AI-based system that includes a proxy model that is response to receiving an update of the input data, processing the input data to generate a proxy prediction are elements that are beyond what is well-understood, routine, and conventional.
In response the argument is not persuasive. The Examiner notes that the use of neural networks/AI, at this point, is common in the art. Evidence of this is provided by Siemens_2020 (The Elegant Math Behind Neural Networks, m-siemens.de, December 7, 2020) which states, at page 1: “… it’s a computational model that combines simple mathematical functions into a complex calculation graph. Their ability to approximate arbitrary functions makes them frequently used tool…”. Therefore, the computation model known as artificial intelligence (AI) is a “frequently used tool.”
Also, Siemens_2020 (The Elegant Math Behind Neural Networks, m-siemens.de, December 7, 2020) teaches, at page 1, that neural network have been “researched since the dawn of modern computer science in the 1940s and 50s.” However, it was in “2006 that research made the current breakthrough, where they became the central force in machine learning”. Therefore, the modern breakthrough in machine learning was 20 years ago. Additionally, at page 1, Siemens_2020 states that the math required to understand neural networks is the “vector algebra and analysis as taught in a typical engineer’s math class”. Therefore, the Examiner finds that neural networks (AI) have been in the domain of a typical engineer having only typical math classes for at least 20 years.
Due to the above reasons, the rejection under 35 USC 101 is maintained.
Claim Rejections - 35 USC § 102
The Applicant has amended the claim to recite: “…using an artificial intelligence (AI)-based system including at least one AI-based proxy model that is responsive to receiving an update of the input data, processing the input data and the output data using the at least one AI-based proxy model…”.
The Applicant asserts that the art of record does not make these limitations obvious.
The Examiner agrees that the art of record does not show the output of a physics-based model as the input to an AI-based model. Accordingly, the previous rejection is withdrawn, however, a new ground of rejection is presented below.
End Response to Arguments
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1 - 21 rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception without significantly more.
Claim 1.
STEP 1: Yes. The claim recites “A simulation method comprising:”
STEP 2A PRONG ONE: Yes. The claim recites:
“… providing at least one physics-based simulation model with input data from a plurality of different sources of operational data, the at least one physics-based simulation model including model parameters for simulating a physical process; Generating output data comprising simulated predictions using the at least one physics-based simulation model, the simulated predictions determined using the model parameters; and Using an artificial intelligence (AI)-based system including at least one AI-based proxy model that is responsive to receiving an update of the input data, processing the input data and the output data using the at least one AI-based proxy model to generate a proxy prediction for at least one selected prediction from the simulated predictions or a variable derived from the at least one selected prediction as a replacement for or as a supplement to the at least one selected prediction of the variable derived from the at least one selected prediction” which is a series of mathematical operations that produce an output (i.e., prediction). Note that a physics-based simulation model is simply a mathematical equation. Also an artificial intelligence (AI) is also merely a mathematical contrivance (i.e., mathematical equation).
STEP 2A PRONG TWO: No.
While the claim recites “… with input data from a plurality of different sources of operational data…” this is merely descriptive of the data upon which mathematical calculations are performed. These limitations do not even amount to pre-solution activity because these limitations do not actually require any action. These limitations simply describe where the input data is from. Merely stating that the data is from some non-descript “source of operational data” is not a practical application as the operational source does not rely upon or use the mathematical calculations in any way.
STEP 2B: NO.
The claim does not include any elements beyond the abstract idea which are extra solution activity. Therefore the claim does not include anything beyond the abstract idea which might be more than the abstract idea itself.
Claim 2 recites “wherein: the proxy prediction comprises a measurable; and the method further comprises comparing a measurement value of the measurable to the proxy prediction and, based on the comparing, changing a value of one of the model parameters” which merely further characterizes the mathematical operations. Reciting that the prediction is “a measurable” merely characterizes the output data. Additionally, “comparing a measurement value of the measurable to the proxy prediction” is merely a mathematical equality/inequality and therefore simply a mathematical operation. However, even if this was considered to not be mathematics, it is simply a judgement which is a mental process. Further, changing a value of the model based on a comparision is merely making a change to the mathematical equation and therefore not a practical application because it is merely modifying the abstract mathematical equation itself. Further, comparing numbers and updating a mathematical parameter is not significantly more as this is merely the general concept of fitting a mathematical equation/model to observed data.
Claim 3. The limitations of claim 3 are substantially the same as those of claim 1 and are therefore rejected due to the same reasons as outlined above for claim 1. While claim 3 additionally recites “… normalizing at least a portion of the input data using a normalization engine to generate normalized input data that is time-aligned and in a uniform format” this is merely additional mathematical elements.
Claim 13. The limitations of claim 13 are substantially the same as those of claim 3 and 1 and are therefore rejected due to the same reasons as outlined above for claim 1. While claim 13 additionally recites a system comprising a process, the mere recitation of a computer upon which an abstract idea is executed is not indicative of a practical application nor significantly more than the abstract idea.
Claim 4, 14 recites “wherein the update of the input data is available on a predefined schedule” which is merely a description of input data and simply further characterizes the mathematical abstract idea. At best, this may be considered some description of a pre-solution activity of providing data periodically to the mathematical equation. Providing data in a predefined schedule is routine when using computers because computer perform mathematical operations based on a CPU clock signal.
Claim 5, 15 recites “further comprising: using the AI-based system to train the at least one AI-based proxy model based on at least a portion of pairs of the normalized input data and the output data resulting from the normalized input data” however, training an AI-based model is merely the adjustment of coefficients in a mathematical equation. Therefore, this is simply part of the mathematical abstract idea itself.
Claim 6, 16 recites wherein the at least one physics-based simulation model comprises a first physics-based simulation model and a second physics-based simulation model that are not integrated together, however, this is merely descriptive of the mathematical equations and is simply part of the abstract mathematical idea itself.
Claim 7, 17 recites wherein the AI-based system further provides an uncertainty quantification of the proxy prediction which is merely part of the mathematical abstract idea itself.
Claim 8, 18 recites wherein the proxy prediction comprises a measurable; and the method further comprises comparing a measurement value of the measurable to the proxy prediction and, based on the comparing, changing a value of the model parameters which merely further characterizes the mathematical operations. Reciting that the prediction is “a measurable” merely characterizes the output data. Additionally, “comparing a measurement value of the measurable to the proxy prediction” is merely a mathematical equality/inequality and therefore simply a mathematical operation. However, even if this was considered to not be mathematics, it is simply a judgement which is a mental process. Further, changing a value of the model based on a comparision is merely making a change to the mathematical equation and therefore not a practical application because it is merely modifying the abstract mathematical equation itself. Further, comparing numbers and updating a mathematical parameter is not significantly more as this is merely the general concept of fitting a mathematical equation/model to observed data.
Claim 9, 19 recite wherein the proxy prediction is provided for all of the model parameters which is merely additional mathematical elements.
Claim 10, 20 recites wherein the proxy prediction is used as the replacement for the at least one selected prediction to run the physical process which is insignificant application. See MPEP 2106.05(g) (cutting hair after first determining the hair style, printing or downloading generated menus). In the instant claims, merely, generally recite to use a prediction “to run the physical process” the claim does not provide any particular way in which the mathematical model/prediction informs the general physical process. This is tantamount to merely reciting “and use it”. Such limitations are not indicative of a practical application nor significantly more.
Claim 11, 21 recites wherein the proxy prediction is used as the supplement to the at least one selected prediction by combining the proxy prediction with the at least one selected prediction to run the physical process which further recite elements of the mathematical abstract idea and then merely recites insignificant application. See MPEP 2106.05(g). (cutting hair after first determining the hair style, printing or downloading generated menus). In the instant claims, merely, generally recite to use a prediction “to run the physical process” the claim does not provide any particular way in which the mathematical model/prediction informs the general physical process. This is tantamount to merely reciting “and use it”. Such limitations are not indicative of a practical application nor significantly more.
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 - 21 are rejected under 35 U.S.C. 103 as being unpatentable over Albert_2020 (US 2020/0082041 A1) in view of Chen_2020 (Physics-informed Generative Adversarial Networks for Sequence Generation with Limited Data, 1st NeurlIPS workshop on Interpretable Inductive Biases and Physically Structured Learning, 2020).
Claim 1. Albert_2020 discloses “a simulation method comprising: providing at least one physics-based simulation model with input data from a plurality of different sources of operational data, the at least one physics-based simulation model including model parameters for simulating a physical process (para [0005] "observational data, wherein the observational data includes at least one source of physical
data, comprising one or more sensors sensing the physical data", para [0018]" system for generating simulations of physical variables of a physical system" ); generating output data comprising simulated predictions using the at least one physics based simulation model, the simulated predictions determined using the model parameters (para [0006] "a system for generating simulations of physical variables of a physical system.", para [0005] "obtaining numeric simulation data", para [0031] "For an embodiment, the simulation data includes data generated by running numerical simulation models, whereby certain a-priori specified processes are explicitly modeled (for example, via dynamic or statistical equations), "); and using an artificial intelligence (Al) based system including at least one Al-based proxy model that is responsive to receiving an update of the input data, processing the input data to generate a proxy prediction for at least one selected prediction from the simulated predictions or a variable derived from the at least one selected prediction as a replacement for or as a supplement to the at least one selected prediction or the variable derived from the at least one selected prediction (para [0046] "predicting physical model output comprising applying a current model to the normalized preprocessed observation data, the. numeric simulation data, and the domain interpretable data,", para [0017] "artificial intelligence (Al)-based framework assimilates simulation data ... ability to natively incorporate a wide variety of real-time observational data.", para [0017] "physics-informed, scalable machine learning models as computeefficient, cost-effective, modular, data-driven "emulators" (surrogates)" ).
Albert_2020; however, does not teach “… and the output data using the at least one AI-based proxy model…”
Chen_2020; however, makes obvious and using an artificial intelligence (Al) based system including at least one Al-based proxy model that is responsive to receiving an update of the input data, processing the input data “and the output data using the at least one AI-based proxy model” to generate a proxy prediction for at least one selected prediction from the simulated predictions or a variable derived from the at least one selected prediction as a replacement for or as a supplement to the at least one selected prediction or the variable derived from the at least one selected prediction (page 2: “… we propose to incorporate the physics model into the modeling process… the neural networks parameterized transition function takes the current state as input and predicts the next state…” Figure 2.
PNG
media_image3.png
483
771
media_image3.png
Greyscale
EXAMINER NOTE: The training samples are given to physics model and to the generator and the discriminator-1 makes a distinction between what it identifies as the generated prediction vs. the physics prediction. This is used to “iteratively update the generator, discriminator-1, and discriminator-2” until the neural network is trained.
Figure 1 illustrates that the trained PA-GAN model performs better than the physics based model.
PNG
media_image4.png
410
1042
media_image4.png
Greyscale
Also, Table 1 illustrates that the PI-GAN performed better than the physic-based model for pendulum, spring, and Ebola experiments.
PNG
media_image5.png
452
1039
media_image5.png
Greyscale
Accordingly, once the model is trained it would be obvious to those of ordinary skill in the art to use the trained neural network rather than the physics-based model because the trained neural network performed better than the physics-based model at predicting the actual physics behavior. It would be obvious to those of ordinary skill in the art to replace a poorly performing model with a better performing neural network-based model.
Albert_2020 and Chen_2020 are analogous art because they are from the same field of endeavor called estimating physical parameters of a physical system (i.e., physics simulations). Before the effective filing date, it would have been obvious to a person of ordinary skill in the art to combine Albert_2020 and Chen_2020. The rationale for doing so would have been that Albert_2020 teaches to estimate the parameters/behavior of a physical system using a machine learning model. Chen_2020 teaches that to use physics priors from a physic-based model for training a Generative Adversarial Network that results in a machine learning model that more closely predicts actual behavior of the physical system. Therefore, it would have been obvious to combine Albert_2020 and Chen_2020 for the benefit of having a neural network model that is more accurate than a physic-based model to obtain the invention as specified in the claims.
Claim 3. The limitations of claim 3 are substantially the same as those of claim 1 and are therefore rejected due to the same reasons as outlined above for claim 1. Additionally, Albert_2020 discloses the further limitations of “…Normalizing at least a portion of the input data using a normalization engine to generate normalized input data that is time-aligned and a uniform format (para [0020] " normalizing the preprocessed observation data, the numeric simulation data", para [0039] "the normalization of the data includes a process by which one or multiple input data sources (either observational or simulation, or both) are brought to a common, standardized format that can be easily ingested and acted upon by a
further computational workflow downstream.") and “processing the normalized input data to generate updated normalized input data, and processing the updated normalized input data to generate a proxy prediction for at least one selected prediction from the simulated predictions or a variable derived from the at least one selected prediction as a replacement for or a supplement to the at least one selected prediction of the variable derived from the at least one selected prediction” (para [0046] "predicting physical model output comprising applying a current model to the normalized preprocessed observation data, the numeric simulation data, and the domain interpretable data,", para [0017] "artificial intelligence (Al)-based framework assimilates simulation data ... ability to natively incorporate a wide variety of real-time observational data.", para [0017] "physics-informed, scalable machine learning models as computeefficient, cost-effective, modular, data-driven "emulators" (surrogates)")
Claim 12. The limitations of claim 12 are substantially the same as those of claim 1 and are rejected due to the same reasons as outlined above for claim 1. Additionally, Albert_2020 discloses “A system comprising: at least one processor configured to” (para [0005] "observational data, wherein the observational data includes at least one source of physical data, comprising one or more sensors sensing the physical data", para [0018] "system for generating simulations of physical variables of a physical system"; Figure 3 310 servers; par 6: “another embodiment includes a system for generating simulations of physical variables of a physical system, the system includes a plurality of sensors, sensing physical data, one or more computing devices connected through a network… , and memory including instructions…”).
Claim 13. The limitations of claim 13 are substantially the same as those of claim 3 and are therefore rejected due to the same reasons as outlined above for claim 3. Additionally, Albert_2020 discloses “A system comprising: at least one processor configured to: use a normalization engine…” (para [0006] "a system for generating simulations of physical variables of a physical system.", para [0005] "obtaining numeric simulation data", para [0031] "For an embodiment, the simulation data includes data generated by running numerical simulation models, whereby certain a-priori specified processes are explicitly modeled (for example, via dynamic or statistical equations),", para [0020]" normalizing the preprocessed observation data, the numeric simulation data", para [0039] "the normalization of the data includes a process by which one or multiple input data sources (either observational or simulation, or both).
Claim 2. Albert_2020 discloses “the proxy prediction comprises a measurable (para [0045]-[0047] "estimating one or more physical parameters of the physical system includes atmospheric flow, surface flow, hydrology, climate, or weather variables including at least one of temperature, wind, precipitation, flow speed, or soil moisture."); and the method further comprises comparing a measurement value of the measurable to the proxy prediction and, based on the comparing, changing a value of one of the model parameters (para [0046]-[0047] "testing and adapting the current model by updating model parameters, and 4) iterating on this process until a measure of convergence is achieved.").
Claim 4, 14. Albert_2020 discloses “wherein the update of the input data is available on a predefined schedule (para [0026] "data structured on a coarser grid (i.e., at even or uneven intervals in space and/or time)", para [0017] "ability to natively incorporate a wide variety of real-time observational data." ).
Claim 5, 15. Albert_2020 discloses using the Al-based system to train the at least one Al-based proxy model based on at least a portion of pairs of the normalized input data and the output data resulting from the normalized input data (para [0046]-[0047] "training the emulator model includes 1) retrieving an untrained model, 2) predicting physical model output comprising applying a current model to the normalized preprocessed observation data, the numeric simulation data, and the domain interpretable data, 3) testing and adapting the current model by updating model parameters, and 4) iterating on this process until a measure of convergence is achieved." ).
Claim 6, 16. Albert_2020 discloses wherein the at least one physicsbased simulation model comprises a first physics-based simulation model and a second physics-based simulation model that are not integrated together (para [0049] "further include building a plurality of emulator models for different use cases including at least one of hydroclimate models, climate models, numerical weather prediction models, meso-scale weather models, surface flow models, subsurface flow models, environmental fluid dynamics models. At least some embodiments further include retrieving an initial model for each of the different use cases and building a corresponding trained model of one or more of the use cases."; para [0045]-[0047]).
Claim 7, 17. Albert_2020 discloses wherein the Al-based system further provides an uncertainty quantification for the proxy prediction (para [0075] "implement custom methods and pipelines for quantifying and propagating uncertainty in gridded simulation and remote-sensing imagery data over time" ).
Claim 8, 18. Albert_2020 discloses wherein: the proxy prediction comprises a measurable (para [0045]-[0047] "estimating one or more physical parameters of the physical system includes atmospheric flow, surface flow, hydrology, climate, or weather variables including at least one of temperature, wind, precipitation, flow speed, or soil moisture."); and the method further comprises comparing a measurement value of the measurable to the proxy prediction and, based on the comparing, changing a value of one of the model parameters (para [0046]-[004 7] "testing and adapting the current model by updating model parameters, and 4) iterating on this process until a measure of convergence is achieved.").
Claim 9, 19. Albert_2020 discloses wherein the proxy prediction is provided for all of the model parameters (para [0028] "Further, the one or more computing devices 110 operate to estimate one or more physical parameters of the physical system based on the spatial-temporal emulator model." ).
Claim 10, 20. Albert_2020 discloses wherein the proxy prediction is used as the replacement for the at least one selected prediction to run the physical process (para [0033] "A ninth step 290 includes estimating one or more physical parameters of the physical system based on the spatial-temporal emulator model. A tenth step 295 includes utilizing the estimate one or more physical parameters for at least one of a plurality of applications." ).
Claim 11, 21. Albert_2020 discloses wherein the proxy prediction is used as the supplement to the at least one selected prediction by combining the proxy prediction with the at least one selected prediction to run. the physical process (para [0017] "This artificial intelligence (Al)-based framework assimilates simulation data, observat[onal (remote/ground sensing) data, and explicit physical knowledge (conservation laws, constraints) for modeling realistic spatial-temporal processes.", para [0058] "Further, inputs to the spatial-temporal model(s) 312 may include reanalysis data (combining numerical model output with observational data, obtained from publicly-available databases or from non-public databases).", para [0033] "A ninth step 290 includes estimating one or more physical parameters of the physical system based on the spatial-temporal emulator model. A tenth step 295 includes utilizing the estimate one or more physical parameters for at least one of a plurality of applications." ).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BRIAN S COOK whose telephone number is (571)272-4276. The examiner can normally be reached 8:00 AM - 5:00 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, Emerson Puente can be reached at 571-272-3652. 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 S COOK/Primary Examiner, Art Unit 2187