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 .
This action is non-final and is in response to the RCE filed on Sept 16th, 2025
Claims 1-15 are pending and have been considered
Claims 1, 7, 14 have been amended.
The 35 U.S.C. 101 rejection has been withdrawn in view of the amendments.
Claims 1, 7, and 14, and their respective dependent claims are rejected under 35 U.S.C. 112(a) as the amended independent claim recites “to increase well mass flow performance of the well system based on the predicted well dynamics behavior data”, which has no support in the 11/10/2021 specification.
Claims 1, 7, and 14 are rejected under 35 USC 103 as obvious in view of U.S. Patent Application Publication Nos. 2019/0345797 ("Babic"), 2021/0248500 ("Varadarajan"), 2016/0061003 ("Gottumukkala"), 2022/0294217 ("Spalt"), and 20190072645 A ("Ajanoh").
The dependent claims are also rejected under 35 USC 103.
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 .
Response to Amendments and Arguments
In the amendment filed on Sept 16th, 2025, applicant amended independent claims 1, 7, 14. The amendments have been fully considered.
The amendments overcome the 35 U.S.C. 101 rejection. Applicant asserts that amended independent claims are not directed to mental processes. The Examiner agrees the amended claims no longer recite mental processes. The Examiner however points out that the claims continue to recite an abstract idea, specifically mathematical concepts: interpolation and filtering.
Applicant asserts that amended independent claims are not directed to extra-solution activity. Examiner’s interpretation is that the additional elements in the last limitation integrate the abstract idea into a practical application. This makes the claim(s) eligible under 35 USC 101 – the rejection is withdrawn.
However, one should note that the following additional elements “to increase well mass flow performance of the well system based on the predicted well dynamics behavior data” are not found in the specification filed on 11/10/2021 and attracts a 112(a) rejection. (Also from a 35 U.S.C. 103 perspective the underlined recitation does not carry patentable weight, it is an intended result, without any structure or procedural specificity on how to achieve it being described in the specification. ) The closest expression found in the specification is [0023] “to perform one or more well operations to control well activities (e.g., to control well subsystem A (114), well subsystem B (124) and affect well mass flow data”.
Regarding Claims 1, 7, and 14 rejected as obvious in view of U.S. Patent Application Publication Nos. 2019/0345797 ("Babic"), 2021/0248500 ("Varadarajan"), 2016/0061003 ("Gottumukkala"), 2022/0294217 ("Spalt"), and 2013/0028052 ("Routh"), specifically regarding Applicant’s arguments regarding what Routh supplies, the Examiner withdraws the rejection.
In view of the Amendments, a new 35 USC 103 rejection is made, Claims 1, 7, and 14 being rejected as obvious in view of U.S. Patent Application Publication Nos. 2019/0345797 ("Babic"), 2021/0248500 ("Varadarajan"), 2016/0061003 ("Gottumukkala"), 2022/0294217 ("Spalt"), and 20190072645 A ("Ajanoh").
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-15 rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention. Independent claims 1,7,14 recite “to increase well mass flow performance of the well system based on the predicted well dynamics behavior data”. There is no mentioning on how to increase well mass flow performance in the specification filed on 11/10/2021. The closest expression found in the specification is [0023] “to perform one or more well operations to control well activities (e.g., to control well subsystem A (114), well subsystem B (124) and affect well mass flow data”.
Thus claims 1,7, 14 are rejected under 35 US 112(a). Dependent claims are rejected under the same rationale.
Claims 1 -15 rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the enablement requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to enable one skilled in the art to which it pertains, or with which it is most nearly connected, to make and/or use the invention. Independent claims 1,7,14 recite “wherein the processor filters both of the well dynamics behavior data from the well system and the modeled well dynamics behavior data from the physics-based model such that a frequency spectrum of the filtered well dynamics behavior data from the well system matches a frequency spectrum of the filtered modeled well dynamics behavior data from the physics-based mode.”
How to filter such that the frequency spectrum …matches a frequency spectrum …. is not sufficiently described as to enable one skilled in the art to implement the invention.
Thus claims 1,7, 14 are rejected under 35 US 112(a). Dependent claims are rejected under the same rationale.
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-15 rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Independent claims 1,7,14 recite “wherein the processor filters both of the well dynamics behavior data from the well system and the modeled well dynamics behavior data from the physics-based model such that a frequency spectrum of the filtered well dynamics behavior data from the well system matches a frequency spectrum of the filtered modeled well dynamics behavior data from the physics-based mode.”
The expression matches is indefinite. Does it mean exactly equal? His is unlikely – if matched in spectral amplitude and phases the two signals would be totally identical anyways. If matching only in amplitude, and keeping phase difference in training would make limited or no sense. Does matching mean approximately equal within a tolerance? With similar cut-off frequency?
Because a POSTIA cannot tell with reasonable certainly what scope “matches” covers, the term is indefinite under 112(b)
Thus claims 1,7, 14 are rejected under 35 US 112(a). Dependent claims are rejected under the same rationale.
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 difference 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 the invention was made.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied for establishing a background for determining obviousness under 35 U.S.C. 103(a) are summarized as follows:
i. Determining the scope and contents of the prior art.
ii. Ascertaining the differences between the prior art and the claims at issue.
iii. Resolving the level of ordinary skill in the pertinent art.
iv. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim 7 (system claim) representative of claims 1, 14 is rejected under 35 U.S.C. 103 as being unpatentable over Babic et al (US 2019/0345797), in view of Amur Varadarajan et al (US 2021/0248500), in further view of Gottumukkala US 20160061003, in further view of Spalt et al (US 2022/0294217), in further view of Ajanoh US 20190072645 A. Claims 1, 14 are rejected under the same rationale as presented in the rejection of claim 7.
Claim 7 – Babic discloses: A well system comprising: {see at least [0009] “method for optimizing the operation of one or more oil or gas production wells”},
a well site; {see at least fig1, [0009] “oil or gas production wells, each well comprising a corresponding production device”, [0052] “oil wells 108”};
a physics-based modeling server that outputs modeled well dynamics behavior data for the well site based on a physics-based model ; and {see at least [0080] “the output of the physical-model module 340 are processed using a customizable algorithm submodule 342 to optimize the oil production rate.“, Fig5a 340, the output modeled well dynamics behavior data shown as the output of the physical-model module [0080] “the output of the physical-model module 340”; [0085] “Data collected by the data acquisition/control submodule 332 may optionally also be sent to the physical-model submodule 340 with a static model established based on well operating unit. The static model provide a theoretical limitation for operating envelope.” for the well site see at least fig1, [0009]};
a digital twin manager coupled to the physics-based modeling server that includes a processor {as the management module 312, see at least [0074] “the system 100 comprises a management module 312 for monitoring status of the oil-production devices 106, wells, and other necessary parameters, and for controlling the operation of the oil-production devices”, [0080] “the output of the physical-model module 340 are processed using a customizable algorithm submodule 342 to optimize the oil production rate.”, [0057] "The processing structure 122 may be one or more single-core or multiple-core computing processors" },
wherein the processor of the digital twin manager: {as the management module 312, see at least [0074] “the system 100 comprises a management module 312 for monitoring status of the oil-production devices 106, wells, and other necessary parameters, and for controlling the operation of the oil-production devices"} comprises:
obtains, based on a predetermined monitoring criterion, {as the well-status monitoring system see at least [0043] “The computer-executable code, when executed, causes at least one processing structure to perform actions comprising: for each well, (i) collecting data of the well in real-time from one or more sources including one or more of a data historian module, a database, said production devices, a well-status monitoring subsystem, and an input/output interface”}, well dynamics behavior data of the well system {as the data collected by the data acquisition submodule, see at least [0084] “ As described above, the configuration submodule 328 instructs the data acquisition/control submodule 332 to collect data from a variety of sources including the production devices 106 via the data collection layer 316, the PI system 318, well-status monitoring subsystem 334 collecting well status data such as well on, well off, or well struggling”; [0009] “method for optimizing the operation of one or more oil or gas production wells”}.
obtains modeled well dynamics behavior data for the well system using the physics-based model {as the output data of the physics-based model using a static model established based on well operating unit, see at least [0085] “Data collected by the data acquisition/control submodule 332 may optionally also be sent to the physical-model submodule 340 with a static model established based on well operating unit. The static model provide a theoretical limitation for operating envelope.” [0009], [0086] “the outputs of the data-model submodule 338 and the physical-model submodule 340 are processed by the customizable algorithm submodule 342 for production optimization”}
obtains real-time well dynamics behavior data of the well system; {as the collecting data of the well in real time, see at least [0042] “The system comprises: a memory; a networking interface configured for communicating with each production device; and at least one processing structure functionally coupled to the memory and the networking interface. The at least one processing structure is configured for: for each well, (i) collecting data of the well in real-time from one or more sources including one or more of a data historian module, a database, said production devices, a well-status monitoring subsystem, and an input/output interface;” [0084] “data from a variety of sources, including the production devices 106 via the data collection layer 316”}
the control mechanism of the well site is adjusted to increase well mass flow performance of the well system based on the predicted well dynamics behavior data. . {see at least [0081] “Based on the optimization results, the output submodule 346 sends instructions via the data acquisition/control submodule 332 to the controllers 146 and optionally the drives 144 to adjust the operation of the oil-production devices 106 for optimized production”.} as the instructions sent to controllers to adjust the operation of the oil-producing devices for optimized production based on the optimization results, the control mechanism of the well system interpreted as the controllers and drives.
Babic does not disclose, however, Amur Varadarajan discloses:
trains a physics constrained machine learning model using one or more machine learning algorithms based on the well dynamics behavior data and the modeled well dynamics behavior data as input data; {see at least fig2, fig[0039] “As shown in Fig.2, the method 200 includes training a machine learning model to adjust parameter values generated by a physics-based model over a first or “training” time period, as at 202. The physics-based model may represent a physical system and may be used to simulate parameters of the physical system in the future (i.e., in advance of its current state of operations). For example, the physical system may be a well system, such as that schematically illustrated in FIG. 3 and generally indicated by reference number 300. As shown in FIG. 3, the system 300 may include surface components 302 and downhole components 304, which may be located in a well 306.” , [0040] “In order to calculate these parameters (and/or others), the physics-based model may receive input parameters, which may generally relate to the physical characteristics of the equipment, fluid, and well of the well system 300.” fig4a, fig4b.} (as the training of a machine learning model that adjusts the output of a physics-based model built to simulate a physical well system and receives inputs related to the general characteristics of the well)
generates using the physics constrained machine learning model and the physics based model, to generate and output predicted well dynamics behavior data using the real-time well dynamics behavior data; and { see at least [0043] “If the machine learning model is verified, the method 200 proceeds to using the machine learning model to forecast parameters, as at 206. In particular, at 206, the method 200 may include forecasting parameter values using the physics-based model and the machine learning model over a third or “implementation” time period, which may be at least partially in the future, i.e., ahead of the current state of operation.”, [0044] “As noted above, the method 200 may then proceed to controlling the physical system based at least in part on the forecasted parameter values, as at 208. For example, a computer executing the machine learning model in combination with the physics-based model may prescribe changes in the operating parameters of the well system 300”. fig5c (538) “Predict one or more events in the physical system”, fig5c (540) “Adjust operation of the physical system”, [0041] “[0041] Many of the inputs may be taken in real-time, and the physics-based model may use these inputs to update its prediction of the output parameters.”, fig4a (402), (404), (406), fig4b (402), (404), (406), (452), [0046] FIG. 4B illustrates a control block diagram of an implementation stage 450, generally corresponding to box 206 of FIG. 2, of the method 200, according to an embodiment. The implementation stage 450 may be similar to the training stage 400, except instead of a measured value 408, the output of the machine learning model (e.g., the adjusted value) provides a forecast value at 452. This forecast value 452 thus serves as the basis for subsequent visualizations of operation of the well system 300”}. as the output of the forecast as at 206, and in a different example predicts fig5c (528), at least partially in the future, that method uses to machine learning model that adjusts parameters from the physics-based model and input data including real-time data from the well
In addition, it would have been for one of ordinary skill in the art before the effective filing date of the claimed invention to modify Babic to include the elements of Amur Varadarajan. Babic uses a physics based model to determine optimal parameters to change the plant operation to increase production from the wells. Physics based models have limited accuracy when dealing with complex systems, such as those of flow in production wells. Over more than a decade machine learning models have shown the ability to learn from the data and produce more accurate simulation results. Traditional simulations are based on mathematical models that describe the behavior of a system using a set of equations. While effective in many cases, these models often struggle to capture the complexity and non-linearity of real-world systems. Machine learning, however, excels in handling complex, data-driven scenarios, making it a natural fit for enhancing simulations. Machine learning models can either be completely model free or adjust parameters of the physics model, with the later being known to be more easy to train and to understand. Hence one would have been motivated to use physics constrained machine learning model to determine optimal parameters for adjusting the production according to the model, as such model would be more accurate and result in increased production.
Furthermore, the Supreme Court has supported that combining well known prior art elements, in a well-known manner, to obtain predictable results is sufficient to determine an invention obvious over such combination (see KSR International Co. v. Teleflex Inc. (KSR), 550 U.S.,82 USPQ2d 1385 (2007) & MPEP § 2143). In the instant case, Babic evidently discloses a model and using the results to optimize well functionality. Amur Varadarajan is merely relied upon to improve the model by training it with machine learning algorithms in the same or similar context. Since both training a machine-learning model and using the results to improve prediction and increase production of the wells, as well as obtaining a series of relevant well data and information are implemented through well-known computer technologies in the same or similar context, combining their features as outlined above using such well-known computer technologies (i.e., conventional software/hardware configurations), would be reasonable, according to one of ordinary skill in the art. Moreover, since the elements disclosed by Babic, as well as Amur Varadarajan would function in the same manner in combination as they do in their separate embodiments, it would be reasonable to conclude that their resulting combination would be predictable. Accordingly, the claimed subject matter is obvious over Babic / Amur Varadarajan.
Babic, Amur Varadarajan do not disclose, however Gottumukkala discloses:
wherein the physics-based model includes a valve module that corresponds to and emulates a control mechanism of the well system {[0015] Embodiments of the present disclosure include the use of analytical well modeling tools; [0036] The optimizer module 80 may be designed to generate gradient curves based on the productivity of each well zone 34 at each corresponding flow control valve 36 over various flow control valve settings that are represented in the network model, as illustrated in FIG. 6.; [Claim 4.] The method as recited in claim 1, further comprising building the completion network model based on a layout of the flow control valves, completion tubing, and reservoir parameters}. Analytical well modeling (i.e. based on equations) is interpreted as physics-based model, includes a valve module that corresponds to and emulates a control mechanism of the well system interpreted as flow control valve.
In addition, it would have been for one of ordinary skill in the art before the effective filing date of the claimed invention to modify Babic /Amur Varadarajan to to include the elements of Gottumukkala . They would have been motivated to do so to have the advantage of including in their model the main elements that control the well system. Furthermore, the Supreme Court has supported that combining well known prior art elements, in a well-known manner, to obtain predictable results is sufficient to determine an invention obvious over such combination (see KSR International Co. v. Teleflex Inc. (KSR), 550 U.S.,82 USPQ2d 1385 (2007) & MPEP § 2143). In the instant case, Babic, Amur Varadarajan evidently discloses a model and using the results to optimize well functionality. Gottumukkala is merely relied upon to improve the model by including a model of a control valve. in the same or similar context. Since both training a machine-learning model and using the results to improve prediction and increase production of the wells, as well as obtaining a series of relevant well data and information are implemented through well-known computer technologies in the same or similar context, combining their features as outlined above using such well-known computer technologies (i.e., conventional software/hardware configurations), would be reasonable, according to one of ordinary skill in the art. Moreover, since the elements disclosed by Babic, Amur Varadarajan as well as Gottumukkala would function in the same manner in combination as they do in their separate embodiments, it would be reasonable to conclude that their resulting combination would be predictable. Accordingly, the claimed subject matter is obvious over Babic / Amur Varadarajan/Gottumukkala.
Babic, Amur Varadarajan, Gottumukkala do not disclose, however Spalt discloses:
wherein the processor interpolates the well dynamics behavior data to fill missing data or outliers in the input data for the physics constrained machine learning model { see at least [0035] The data processing system can pre-process the received data or signals in order to increase the accuracy of the machine learning generator. For example, the data processing system can apply one or more filters to the data based on the sampling rate and other characteristics of the data. “, the data processing system 202 can estimate the missing data via time-based interpolation or other interpolation methods to determine the missing values.”} as filling missing values by interpolation when preparing data for machine learning, which is a commonly known procedure to those in the art,
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Babic, Amur Varadarajan, Sandnes to include the elements of Spalt. One would have been motivated to do so, in order to use cleaned training data, which is essential for learning any good model, physics based or otherwise. Furthermore, the Supreme Court has supported that combining well known prior art elements, in a well-known manner, to obtain predictable results is sufficient to determine an invention obvious over such combination (see KSR International Co. v. Teleflex Inc. (KSR), 550 U.S.,82 USPQ2d 1385 (2007) & MPEP § 2143). In the instant case, Babic, Amur Varadarajan, Sandnes evidently discloses a model for optimizing a well system dynamic performance. Spalt is merely relied upon to illustrate the training of the model with cleaned data in the same or similar context. Since both optimizing a well system dynamic performance, as well as cleaning of the data are implemented through well-known computer technologies in the same or similar context, combining their features as outlined above using such well-known computer technologies (i.e., conventional software/hardware configurations), would be reasonable, according to one of ordinary skill in the art. Moreover, since the elements disclosed by Babic, Amur Varadarajan, Sandnes as well as Spalt would function in the same manner in combination as they do in their separate embodiments, it would be reasonable to conclude that their resulting combination would be predictable. Accordingly, the claimed subject matter is obvious over Babic, Amur Varadarajan, Goittumukkala, Spalt.
Babic, Amur Varadarajan, Gottumukkala, Spalt disclose all the limitation of the claim except an amended limitation by which a preprocessing step is applied, to the physics model output data and well data, which in view of the specification are combined (the specification does not indicate how) to form the input vector used for training. Further the specification recites “the modelled and obtained dynamics behavior data for the well system of interest may be filtered using a moving average filter to match frequency spectra of the input dynamics behavior data (Equation 1)”. Equation 1 is a well-known definition of a moving average calculation. It could be applied to either of the two or both, however the specification does not indicate how to match the frequency spectra is not described. Thus, in broadest reasonable interpretation and in view of the specification, the limitation is interpreted as applying filtering on the input vector (effectively the same filtering on input vector components, which have the two sources indicated above, thus matching their spectra). Preprocessing training set by filtering is a most common technique in machine learning used to remove noise in the training data.
Thus, Babic, Amur Varadarajan, Gottumukkala, Spalt do not disclose however Ajanoh discloses
the processor filters both of the well dynamics behavior data from the well system and the modeled well dynamics behavior data from the physics-based model such that a frequency spectrum of the filtered well dynamics behavior data from the well system matches a frequency spectrum of the filtered modeled well dynamics behavior data from the physics-based model; { 16. The system of claim 16, wherein the convolutional neural network includes an input that receives the training set, a convolutional phase that filters the training set to extract interdependent features, and a subsampling phase that reduces computational complexity to produce the centroids.}
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Babic, Amur Varadarajan, Gottumukkala, Spalt with elements of Ajanoh. One would have been motivated to do so to have the advantage of filtering noisy input data for training with a similar filter to obtain corelated spectrum in the joint input vector (of which one becomes input and other becomes target output of the neural model) . Though the Ajanoh applicatgion is in a different field, the signal processing method is generic to all fields. Since both optimizing a well system dynamic performance, as well as filtering of the data are implemented through well-known computer technologies in the same or similar context, combining their features as outlined above using such well-known computer technologies (i.e., conventional software/hardware configurations), would be reasonable, according to one of ordinary skill in the art. Filtering to remove noise, especially filtering by moving average, the simplest form of low pass filtering, would have been an abovious choice to try. Moreover, since the elements disclosed by Babic, Amur Varadarajan, Sandnes as well as Spalt would function in the same manner in combination as they do in their separate embodiments, it would be reasonable to conclude that their resulting combination would be predictable. Accordingly, the claimed subject matter is obvious over Babic, Amur Varadarajan, Goittumukkala, Spalt, Ajanoh.
Claims 2, 8, 15 are rejected under 35 U.S.C. 103 as being unpatentable over Babic et al (US 2019/0345797), in view of Amur Varadarajan et al (US 2021/0248500) in further view of Gottumukkala US 20160061003 in further view of Spalt et al (US 2022/0294217) in further view of Ajanoh US 20190072645 A1 in further view of Liu et al (Faster than Real-time Simulation: Methods, Tools, and Applications, April 2021, DOI:10.48550/arXiv.2104.04149; in further view of Lakshmanan et al ("Artificial Intelligence Based Optimization of Gathering System." November 2020. Doi).
Regarding Claims 2, 8, 15 – Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh discloses the limitations of Claims 1, 7, 14. Babic further discloses:
wherein the [non-linear] well dynamics behavior includes well mass flow, well pressure, and well temperature performance of a well of interest, and {see at least [abstract] … parameters such as economical threshold, water cut, temperature, pressure, flow rate or vibration, failure model prediction, and the like …}
Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh does not disclose, however, Liu discloses:
wherein the predicted well dynamics behavior data have higher time resolution than the real-time well dynamics behavior data and captures non-linear well dynamics behavior of the well system, {as faster than real-time simulation, which in broadest reasonable interpretation includes all simulations, both for linear and non-linear systems, see at least [Abstract] “Real-time simulation enables the understanding of system operating conditions by evaluating simulation models of physical components running synchronized at the real-time wall clock. Leveraging the real-time measurements of comprehensive system models, faster than real-time (FTRT) simulation allows the evaluation of system architectures at speeds faster than real-time. FTRT simulation can assist in predicting the system's behavior efficiently, thus assisting the operation of system processes”}
In addition, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh to include the elements of Liu. One would have been motivated to do so, to have a better visibility at the behavior of the non-linear systems at moments in time which are closer to each other, i.e. at higher rate, that the rate with which data is collected from the well, which is here the real-time data, and that improved visibility give more accurate predictions of the dynamics behavior. Furthermore, the Supreme Court has supported that combining well known prior art elements, in a well-known manner, to obtain predictable results is sufficient to determine an invention obvious over such combination (see KSR International Co. v. Teleflex Inc. (KSR), 550 U.S.,82 USPQ2d 1385 (2007) & MPEP § 2143). In the instant case, Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh evidently discloses optimizing a well system dynamic performance. Liu is merely relied upon to illustrate the functionality of predicted well dynamics behavior data with higher time resolution in the same or similar context. Since both optimizing a well system dynamic performance, as well as predicted well dynamics behavior data with higher time resolution are implemented through well-known computer technologies in the same or similar context, combining their features as outlined above using such well-known computer technologies (i.e., conventional software/hardware configurations), would be reasonable, according to one of ordinary skill in the art. Moreover, since the elements disclosed by Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh, as well as Liu would function in the same manner in combination as they do in their separate embodiments, it would be reasonable to conclude that their resulting combination would be predictable. Accordingly, the claimed subject matter is obvious over Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh / Liu.
Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh, Liu does not disclose, however, Lakshmanan discloses:
wherein the well system comprises interconnected subsystems that include a compressor subsystem and a sales header subsystem {as the network of wellpads connected across compressor stations and sales points, see at least [page 2 – Details of the network] “The network consists of 74 operational wellpads connected across 12 compressor stations (compressor subsystem) with around 150 miles of dry & wet gas pipelines. There are two sales points”}. Lakshmanan fails to explicitly disclose “sales header subsystem”; however, it is reasonable to assume that one of ordinary skills in the art will realize that sales header subsystem is used (a sales header is used to combine production from multiple wells feeding a single sales points)}.
In addition, it would have been obvious to one of ordinary sill in the art before the effective filing date of the claimed invention to modify Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh, Liu to include the elements of Lakshmanan. One would have been motivated to do so, in order to have been able to commercialize the product extracted from the well to customers. Furthermore, the Supreme Court has supported that combining well known prior art elements, in a well-known manner, to obtain predictable results is sufficient to determine an invention obvious over such combination (see KSR International Co. v. Teleflex Inc. (KSR), 550 U.S.,82 USPQ2d 1385 (2007) & MPEP § 2143). In the instant case, Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh, Liu evidently discloses optimizing a well system dynamic performance. Lakshmanan is merely relied upon to illustrate the functionality of a compressor subsystem and a sales header subsystem in the same or similar context. Since both optimizing a well system dynamic performance, as well as a compressor subsystem and a sales header subsystem are implemented through well-known computer technologies in the same or similar context, combining their features as outlined above using such well-known computer technologies (i.e., conventional software/hardware configurations), would be reasonable, according to one of ordinary skill in the art. Moreover, since the elements disclosed by Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh, Liu, as well as Lakshmanan would function in the same manner in combination as they do in their separate embodiments, it would be reasonable to conclude that their resulting combination would be predictable. Accordingly, the claimed subject matter is obvious over Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh, Liu / Lakshmanan.
Claims 3, 9 are rejected under 35 U.S.C. 103 as being unpatentable over Babic et al (US 2019/0345797), in view of Amur Varadarajan et al (US 2021/0248500) in further view of Gottumukkala US 20160061003 in further view of Spalt et al (US 2022/0294217) in further view of Ajanoh US 20190072645 A1in further view of Sun et al (PhysiNet: A Combination of Physics-based Model and Neural Network Model for Digital Twins Preprint · June 2021 DOI: 10.48550/arXiv.2106.14790)
Regarding Claims 3, 9 – Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh discloses the limitations of Claims 1, 7. Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh does not disclose, however, Sun discloses:
wherein the physics-based model emulates components of non-linear well dynamics behavior that are below a predetermined frequency, {as the frequency response of a physics-based model covers entire spectrum, with lower frequencies being of higher amplitude, see at least [Fig. 4] The architecture of a digital twin with a neural network model”, “In this work, we proposed a hybrid model(PhysiNet) which combines the physics-based model and the neural network model for digital twins.” “for a control system in the frequency domain” “Figure 11 shows the prediction results from the model with only neural network at steps #0, #9, #19, #29.”, [4.Conclusions] “It seems the adding of a physics-based model helped PhysiNet find better local minima and capture the non-linear behavior of the system”}
wherein the physics constrained machine learning model is trained to predict components of the non-linear well dynamics behavior {as the frequency response covers all frequency spectrum, see at least [3.2] “ digital twin for a control system in the frequency domain] “Figure 13 shows the frequency response function for the real system and the physics-based model.” “Figure 15 shows the prediction results from the model with only neural network at steps #0, #9, #19, #29. Apart from #0, the neural network model captured the non-linear behavior of the system. “}.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh to include the elements of Sun. One would have been motivated to include the elements of Sun, using well known analysis and matching in the frequency domain, in order to have more accurately match the model of the digital twin with its physical, real system counterpart, and thus improve the prediction and optimization of the model and overall well optimization of production relying on the model. Furthermore, the Supreme Court has supported that combining well known prior art elements, in a well-known manner, to obtain predictable results is sufficient to determine an invention obvious over such combination (see KSR International Co. v. Teleflex Inc. (KSR), 550 U.S.,82 USPQ2d 1385 (2007) & MPEP § 2143). In the instant case, Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh evidently discloses predicting well system dynamics performance through the models. Sun is merely relied upon to further improve the model for in the same or similar context. As best understood by Examiner, both are referring to models used to predict system dynamics performance, and are implemented through well-known computer technologies in the same or similar context, combining their features as outlined above using such well-known computer technologies (i.e., conventional software/hardware configurations), would be reasonable, according to one of ordinary skill in the art. Moreover, since the elements disclosed by Babic, Amur Varadarajan, as well as Sun would function in the same manner in combination as they do in their separate embodiments, it would be reasonable to conclude that their resulting combination would be predictable. Accordingly, the claimed subject matter is obvious over Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh / Sun.
Claims 4, 10 are rejected under 35 U.S.C. 103 as being unpatentable over Babic et al (US 2019/0345797), in view of Amur Varadarajan et al (US 2021/0248500) in further view of Gottumukkala US 20160061003 in further view of Spalt et al (US 2022/0294217) in further view of Ajanoh US 20190072645 A1in further view of Sandnes (US 2021/0181374)
Regarding Claims 4, 10 – Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh discloses the limitations of Claims 1, 7. Babic also discloses:
wherein the well dynamics behavior data include data for both normal operational conditions and shut-down conditions, {see at least [0084] “ As described above, the configuration submodule 328 instructs the data acquisition/control submodule 332 to collect data from a variety of sources including the production devices 106 via the data collection layer 316, the PI system 318, well-status monitoring subsystem 334 collecting well status data such as well on, well off, or well struggling”}
Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh does not disclose, however Sandnes discloses:
wherein the physics constrained machine learning model is trained based on at least six months of the well dynamics behavior data, {see at least [0013] “The models may make use of one or more neural nets that are trained using the respective set of data. [0025] In a typical oil and gas network it may be appropriate for the first time period to be at least six months.”}
In addition, it would have been obvious to one of ordinary skill in the art, at the time of filing, to modify Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh to include the elements of Sandnes. One would have been motivated to do so, in order to ensure a sufficient amount of data for training the machine learning model (and being known in the art that using at least six months of data is recommended). Furthermore, the Supreme Court has supported that combining well known prior art elements, in a well-known manner, to obtain predictable results is sufficient to determine an invention obvious over such combination (see KSR International Co. v. Teleflex Inc. (KSR), 550 U.S.,82 USPQ2d 1385 (2007) & MPEP § 2143). In the instant case, Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh evidently discloses training a machine learning model with well data, for prediction and optimizing a well system dynamic performance. Sandnes is merely relied upon to illustrate a characteristic of the data used for machine learning training, with similar functionality and in the same or similar context. Since both training the machine learning model with some amount of data and as well as using at least six months of training data are implemented through well-known computer technologies in the same or similar context, combining their features as outlined above using such well-known computer technologies (i.e., conventional software/hardware configurations), would be reasonable, according to one of ordinary skill in the art. Moreover, since the elements disclosed by Babic, Amur Varadarajan, as well as Sandnes would function in the same manner in combination as they do in their separate embodiments, it would be reasonable to conclude that their resulting combination would be predictable. Accordingly, the claimed subject matter is obvious over Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh / Sandnes.
Claims 5, 11 are rejected under 35 U.S.C. 103 as being unpatentable over Babic et al (US 2019/0345797), in view of Amur Varadarajan et al (US 2021/0248500) in further view of Gottumukkala US 20160061003 in further view of Spalt et al (US 2022/0294217) in further view of Ajanoh US 20190072645 A1in further view of Dan (Accuracy of Six Interpolation Methods Applied on Pupil Diameter Data, 978-1-7281-7166-1/20/$31.00 ©2020 IEEE, https://tins.ro/publications/repository/Dan_et_al_AQTR_2020.pdf
Regarding Claims 5, 11 – Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh discloses the limitations of Claims 1, 7. Spalt also discloses:
wherein the filter is a moving average filter applied in a sliding window along the input data to compute an average value of the input data in each sliding window, wherein a length of the filter has a predetermined value based on the input data within a predetermined frequency spectrum. {see at least .[0034] ” The data processing system can identify erroneous samples and remove or modify them to clean the data to facilitate the production of accurate prediction results with the machine learning generator. “, [0035] “The data processing system can pre-process the received data or signals in order to increase the accuracy of the machine learning generator. For example, the data processing system can apply one or more filters”, [0078] “Examples of smoothness filters or techniques can include a linear smoother (e.g., where smoothed values are provided as a linear transformation of the observed values), additive smoothing, Butterworth filter, Chebyshev filter, digital filter, elliptic filter, exponential smoothing, Kalman filter, or moving average.” [0075] “The data repository 220 can include filters 228. A filter 228 can refer to or include a signal processing filter, such as a low pass filter, bandpass filter, or high pass filter. A filter 228 can include a filter to smooth utility grid data or signals, such as a smoothness filter. A smoothness filter can create an approximating function that attempts to capture patterns or trends in the data”,” can select a filter configured with a user-specified weighting that applies different weights to each sample based on a desired emphasis of the input data. The length in time of the filter can be selected based on the timescale of the underlying characteristics that are being targeted, e.g., a short filter for transient signals and a long filter for more steady-state characteristics”}
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh to include the further elements of Spalt. One would have been motivated to do so, in order to further clean the data for further processing, eliminating data that is expected to be noise, using traditional filtering techniques. Furthermore, the Supreme Court has supported that combining well known prior art elements, in a well-known manner, to obtain predictable results is sufficient to determine an invention obvious over such combination (see KSR International Co. v. Teleflex Inc. (KSR), 550 U.S.,82 USPQ2d 1385 (2007) & MPEP § 2143). In the instant case, Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh evidently discloses training a model with cleaned data Further elements of Spalt are merely relied upon to illustrate the choices of the filters. Since both cleaning data and implementing specific filtering methods are implemented through well-known computer technologies in the same or similar context, combining their features as outlined above using such well-known computer technologies (i.e., conventional software/hardware configurations), would be reasonable, according to one of ordinary skill in the art. Moreover, since the elements disclosed by Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh, as well as further elements of Spalt would function in the same manner in combination as they do in their separate embodiments, it would be reasonable to conclude that their resulting combination would be predictable. Accordingly, the claimed subject matter is obvious over Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh.
Babic, Amur Varadarajan, Gottumukkala, Spalt, Ajanoh, does not disclose, however, Dan discloses:
wherein the interpolation algorithm is selected from a group consisting of a nearest neighbor interpolation, a linear interpolation, a piecewise cubic spline interpolation, a shape-preserving piecewise cubic spline interpolation, and a modified Akima cubic Hermite interpolation, {see at least [Abstract] “we aim to analyze the accuracy of six interpolation methods”, “ the most accurate reconstruction of pupil loss is obtained by Akima, Makima, and the Piecewise Cubic Hermite Interpolating Polynomial “, [B11 Interpolation Methods] “Therefore, next, we will shortly describe several known methods of interpolation, and provide a critical analysis in terms of complexity and smoothness. We will focus on the following methods: Previous neighbor [16], Linear [2], Cubic Spline [16], Piecewise Cubic Hermite Interpolating Polynomial (Pchip) [16], Akima [17], and Makima [16]”, [B.2. The complexity of the algorithms] “The simplest interpolati