DETAILED ACTION
This Office action is in response to a communication filed on June 25th, 2026 for Application No. 18/365,719, in which claims 1-20 are presented for examination. The amendments filed on June 25th, 2026 have been entered, where claims 1-20 are amended.
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 .
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to abstract ideas without significantly more.
Regarding Claim 1:
Step 1: Claim 1 is a machine claim. Therefore, claims 1-7 are directed to a statutory category of eligible subject matter.
Step 2A Prong 1: If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the "Mental Processes" grouping of abstract ideas. Here, steps of the claimed subject matter are mental processes. Specifically, the claim recites
“for which a future state-of-health (SOH) is to be predicted . . . and predict the future SOH using the battery data” (mental process – amounts to exercising judgement to form an opinion on a future state, with reference to known or observed information, which may be aided by pen and paper);
“identify a selected model from an ensemble of models according to prediction criteria including prediction time and a time length” (mental process – amounts to exercising judgment to form an opinion on which model, from a plurality of known or observed models, should be selected, with reference to known or observed data, which may be aided by pen and paper); and
“temporal parings derived from SOH history curves” (mental process – amounts to form an opinion on pairings, with reference to known or observed data, which may be aided by pen and paper).
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites the additional elements:
“A system, comprising: a memory communicably coupled to a processor and storing instructions that, when executed by the processor, cause the processor to . . . as an input to the selected model” (mere instructions to apply the exception using generic computer components does not provide an inventive concept);
“acquire battery data” (acquiring battery data amounts to insignificant extra-solution activity because the transmission of data is incidental to the claimed subject matter); and
“of a battery . . . at a prediction time after a last point of the battery data . . . of the battery data . . . the ensemble of models including models trained using different . . . of batteries, the temporal pairings corresponding to different values of the prediction criteria and including battery data at a first time point on a respective SOH history curve as model input and an SOH value at a second time point after the first time point on the respective SOH history curve as a supervising signal” (amounts to merely generally linking the use of the judicial exception to a particular technological environment or field of use, which does not impose any meaningful limits on practicing the abstract idea).
Step 2B: The claim does not include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception.
The claim recites the additional elements:
“A system, comprising: a memory communicably coupled to a processor and storing instructions that, when executed by the processor, cause the processor to . . . as an input to the selected model” (mere instructions to apply the exception using generic computer components does not provide an inventive concept);
“acquire battery data” (transmitting data is well‐understood, routine, and conventional, see generally Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362; see also buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014); therefore the limitation, which is recited with a high level of generality, remains insignificant extra-solution activity even upon reconsideration); and
“of a battery . . . at a prediction time after a last point of the battery data . . . of the battery data . . . the ensemble of models including models trained using different . . . of batteries, the temporal pairings corresponding to different values of the prediction criteria and including battery data at a first time point on a respective SOH history curve as model input and an SOH value at a second time point after the first time point on the respective SOH history curve as a supervising signal” (merely generally linking the use of the judicial exception to a particular technological environment or field of use does not provide an inventive concept).
For the reasons above, Claim 1 is rejected as being directed to an abstract idea without significantly more. This rejection applies equally to dependent claims 2-7. The additional limitations of the dependent claims are addressed below.
Regarding Claim 2:
Step 2A Prong 1: See the rejection of Claim 1 above, which Claim 2 depends on. Here, the claim recites additional elements that are mental processes. Specifically, the claim recites
“predict current state-of-health (SOH) values” (mental process – amounts to exercising judgement to form an opinion on a future state, with reference to known or observed information, which may be aided by pen and paper).
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites the additional elements:
“wherein the instructions further cause the processor to train an estimation model to . . . wherein instructions cause the processor to train the estimation model . . . to generate a loss value of the estimation model and update the estimation model” (amounts to mere instructions to apply the judicial exception on generic and unspecialized computer components, which do not impose any meaningful limits on practicing the abstract idea) and
“of batteries at respective training time points . . . by using current values of the battery data at the respective training time points as training data for the estimation model and using lab-measured SOH values of the batteries as a supervising signal for the estimation model” (amounts to merely generally linking the use of the judicial exception to a particular technological environment or field of use, which do not impose any meaningful limits on practicing the abstract idea).
Step 2B: The claim does not include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception.
The claim recites the additional elements:
“wherein the instructions further cause the processor to train an estimation model to . . . wherein instructions cause the processor to train the estimation model . . . to generate a loss value of the estimation model and update the estimation model” (mere instructions to apply the exception using generic computer components does not provide an inventive concept) and
“of batteries at respective training time points . . . by using current values of the battery data at the respective training time points as training data for the estimation model and using lab-measured SOH values of the batteries as a supervising signal for the estimation model” (merely generally linking the use of the judicial exception to a particular technological environment or field of use does not provide an inventive concept).
Accordingly, Claim 2 is rejected as being directed to an abstract idea without significantly more.
Regarding Claim 3:
Step 2A Prong 1: See the rejection of Claim 2 above, which Claim 3 depends on. As discussed above, if a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the "Mental Processes" grouping of abstract ideas. Whereas if a claim limitation, under its broadest reasonable interpretation, recites mathematical relationships, mathematical formulas or equations, or mathematical calculation, then it falls within the “Mathematical Concepts” grouping of abstract ideas. Here, the claim recites additional elements that are mental processes and mathematical concepts. Specifically, the claim recites
“preprocess the current battery data . . . by at least one of: filtering out relevant battery features and applying empirical equations . . . to the current values” (mental process – amounts to exercising judgement to form opinions on values derived from known or observed information, such as relevant subsets of information or calculated transformations of information, with reference to known or observed information, which may be aided by pen and paper; mathematical concepts – applying empirical equations to input data to generate an output is a recitation of a mathematical equation or calculation).
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites the additional elements:
“wherein the instructions further cause the processor to . . . as the training data for the estimation model” (amounts to mere instructions to apply the judicial exception on generic and unspecialized computer components, which do not impose any meaningful limits on practicing the abstract idea) and
“at the respective training time points before use . . . related to physics of the batteries . . . of the battery data” (amounts to merely generally linking the use of the judicial exception to a particular technological environment or field of use, which do not impose any meaningful limits on practicing the abstract idea).
Step 2B: The claim does not include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception.
The claim recites the additional elements:
“wherein the instructions further cause the processor to . . . as the training data for the estimation model” (mere instructions to apply the exception using generic computer components does not provide an inventive concept) and
“at the respective training time points before use . . . related to physics of the batteries . . . of the battery data” (merely generally linking the use of the judicial exception to a particular technological environment or field of use does not provide an inventive concept).
Accordingly, Claim 3 is rejected as being directed to an abstract idea without significantly more.
Regarding Claim 4:
Step 2A Prong 1: See the rejection of Claim 1 above, which Claim 4 depends on. Here, the claim recites additional elements that are mental processes. Specifically, the claim recites
“predict current SOH values . . . to generate historical SOH data” (mental process – amounts to exercising judgement to form an opinion on a future state, with reference to known or observed information, which may be aided by pen and paper) and
“construct the SOH history curves using the historical SOH data” (mental process – amounts to exercising judgement to form an opinion on a visualization of a data curve, with reference to known or observed information, which may be aided by pen and paper).
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites the additional elements:
“wherein the instructions further cause the processor to: apply an estimation model configured to” (amounts to mere instructions to apply the judicial exception on generic and unspecialized computer components, which do not impose any meaningful limits on practicing the abstract idea) and
“of batteries to historical battery data of the batteries, the historical battery data collected over an interval ending at a last time point before the prediction time . . . corresponding to time points in the interval” (amounts to merely generally linking the use of the judicial exception to a particular technological environment or field of use, which do not impose any meaningful limits on practicing the abstract idea).
Step 2B: The claim does not include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception.
The claim recites the additional elements:
“wherein the instructions further cause the processor to: apply an estimation model configured to” (mere instructions to apply the exception using generic computer components does not provide an inventive concept) and
“of batteries to historical battery data of the batteries, the historical battery data collected over an interval ending at a last time point before the prediction time . . . corresponding to time points in the interval” (merely generally linking the use of the judicial exception to a particular technological environment or field of use does not provide an inventive concept).
Accordingly, Claim 4 is rejected as being directed to an abstract idea without significantly more.
Regarding Claim 5:
Step 2A Prong 1: See the rejection of Claim 4 above, which Claim 5 depends on. Here, the claim recites additional elements that are mental processes. Specifically, the claim recites
“characterize the historical SOH data as first time sequency data and second time sequency data relative to a respective SOH history curve” (mental process – amounts to exercising judgement to form an opinion on whether to classify data into one of two groups, with reference to known or observed information depicted in curve form, which may be aided by pen and paper).
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites the additional elements:
“wherein the instructions further cause the processor to” (amounts to mere instructions to apply the judicial exception on generic and unspecialized computer components, which do not impose any meaningful limits on practicing the abstract idea) and
“wherein the first time sequency data includes data corresponding to a first portion of the respective SOH history curve, and wherein the second time sequency data includes data corresponding to a second portion of the respective SOH history curve that is after the first portion” (amounts to merely generally linking the use of the judicial exception to a particular technological environment or field of use, which does not impose any meaningful limits on practicing the abstract idea).
Step 2B: The claim does not include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception.
The claim recites the additional elements:
“wherein the instructions further cause the processor to” (mere instructions to apply the exception using generic computer components does not provide an inventive concept) and
“wherein the first time sequency data includes data corresponding to a first portion of the respective SOH history curve, and wherein the second time sequency data includes data corresponding to a second portion of the respective SOH history curve that is after the first portion” (merely generally linking the use of the judicial exception to a particular technological environment or field of use does not provide an inventive concept).
Accordingly, Claim 5 is rejected as being directed to an abstract idea without significantly more.
Regarding Claim 6:
Step 2A Prong 1: See the rejection of Claim 5 above, which Claim 6 depends on.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites the additional elements:
“wherein the instructions further cause the processor to . . . to train each model in the ensemble of models” (amounts to mere instructions to apply the judicial exception on generic and unspecialized computer components, which do not impose any meaningful limits on practicing the abstract idea) and
“use battery data corresponding to the first time sequency data and the second time sequency data . . . wherein the first time sequency data and the second time sequency data for each model are located at different time points of the respective SOH history curve” (amounts to merely generally linking the use of the judicial exception to a particular technological environment or field of use, which do not impose any meaningful limits on practicing the abstract idea).
Step 2B: The claim does not include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception.
The claim recites the additional elements:
“wherein the instructions further cause the processor to . . . to train each model in the ensemble of models” (mere instructions to apply the exception using generic computer components does not provide an inventive concept) and
“use battery data corresponding to the first time sequency data and the second time sequency data . . . wherein the first time sequency data and the second time sequency data for each model are located at different time points of the respective SOH history curve” (merely generally linking the use of the judicial exception to a particular technological environment or field of use does not provide an inventive concept).
Accordingly, Claim 6 is rejected as being directed to an abstract idea without significantly more.
Regarding Claim 7:
Step 2A Prong 1: See the rejection of Claim 1 above, which Claim 7 depends on. Here, the claim recites additional elements that are mental processes. Specifically, the claim recites
“generate SOH predictions” (mental process – amounts to exercising judgement to form an opinion on a future state, with reference to known or observed information, which may be aided by pen and paper);
“calculate a SOH prediction error . . . by comparing the SOH prediction to estimated SOH values” (mental process – amounts to exercising judgment to calculate a value, with reference to known or observed information, which may be aided by pen and paper); and
“and create an assignment” (mental process - amounts to exercising judgement to form an opinion on an association between known or observed information, which may be aided by pen and paper).
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites the additional elements:
“wherein the instructions further cause the processor to” (amounts to mere instructions to apply the judicial exception on generic and unspecialized computer components, which do not impose any meaningful limits on practicing the abstract idea) and
“for each model in the ensemble of models using historical battery data as an input to each model . . . for each model . . .of each model . . . wherein the prediction criteria include the prediction time and the time length of the battery data, wherein the selected model is chosen based on the assignment” (amounts to merely generally linking the use of the judicial exception to a particular technological environment or field of use, which do not impose any meaningful limits on practicing the abstract idea).
Step 2B: The claim does not include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception.
The claim recites the additional elements:
“wherein the instructions further cause the processor to” (mere instructions to apply the exception using generic computer components does not provide an inventive concept) and
“for each model in the ensemble of models using historical battery data as an input to each model . . . for each model . . .of each model . . . wherein the prediction criteria include the prediction time and the time length of the battery data, wherein the selected model is chosen based on the assignment” (merely generally linking the use of the judicial exception to a particular technological environment or field of use does not provide an inventive concept).
Accordingly, Claim 7 is rejected as being directed to an abstract idea without significantly more.
Regarding Claim 8:
Step 1: Claim 8 is a machine claim. Therefore, claims 8-14 are directed to a statutory category of eligible subject matter.
Step 2A Prong 1: If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the "Mental Processes" grouping of abstract ideas. Here, the claim recites limitations that are substantially the same as the limitations of Claim 1. As a result, and as elaborated above, these limitations are abstract ideas because they are mental processes.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites the additional elements:
“A non-transitory computer-readable medium including instructions that, when executed by a processor, cause the processor to . . . as an input to the selected model” (mere instructions to apply the exception using generic computer components does not provide an inventive concept);
“acquire historical battery data” (acquiring battery data amounts to insignificant extra-solution activity because the transmission of data is incidental to the claimed subject matter); and
“of a battery . . . at a prediction time after a last point of the battery data . . . of the battery data . . . the ensemble of models including models trained using different . . . of the battery data, the temporal pairings corresponding to different values of the prediction criteria and including battery data at a first time point on a respective SOH history curve as model input and an SOH value at a second time point after the first time point on the respective SOH history curve as a supervising signal” (amounts to merely generally linking the use of the judicial exception to a particular technological environment or field of use, which do not impose any meaningful limits on practicing the abstract idea).
Step 2B: The claim does not include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception.
The claim recites the additional elements:
“A non-transitory computer-readable medium including instructions that, when executed by a processor, cause the processor to . . . as an input to the selected model” (mere instructions to apply the exception using generic computer components does not provide an inventive concept);
“acquire historical battery data” (transmitting data is well‐understood, routine, and conventional, see generally Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362; see also buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014); therefore the limitation, which is recited with a high level of generality, remains insignificant extra-solution activity even upon reconsideration); and
“of a battery . . . at a prediction time after a last point of the battery data . . . of the battery data . . . the ensemble of models including models trained using different . . . of the battery data, the temporal pairings corresponding to different values of the prediction criteria and including battery data at a first time point on a respective SOH history curve as model input and an SOH value at a second time point after the first time point on the respective SOH history curve as a supervising signal” (merely generally linking the use of the judicial exception to a particular technological environment or field of use does not provide an inventive concept).
For the reasons above, Claim 8 is rejected as being directed to an abstract idea without significantly more. This rejection applies equally to dependent claims 9-14. The additional limitations of the dependent claims are addressed below.
Regarding Claim 9, the claim recites limitations that are all substantially the same as limitations of Claim 2, in the form of a non-transitory computer-readable medium. The claim is also directed to performing mental processes without integration into a practical component or significantly more.
Accordingly, Claim 9 is rejected under the same rationale.
Regarding Claim 10, the claim recites limitations that are all substantially the same as limitations of Claim 3, in the form of a non-transitory computer-readable medium. The claim is also directed to performing mental processes and mathematical concepts without integration into a practical component or significantly more.
Accordingly, Claim 10 is rejected under the same rationale.
Regarding Claim 11, the claim recites limitations that are all substantially the same as limitations of Claim 4, in the form of a non-transitory computer-readable medium. The claim is also directed to performing mental processes without integration into a practical component or significantly more.
Accordingly, Claim 11 is rejected under the same rationale.
Regarding Claim 12, the claim recites limitations that are all substantially the same as limitations of Claim 5, in the form of a non-transitory computer-readable medium. The claim is also directed to performing mental processes without integration into a practical component or significantly more.
Accordingly, Claim 12 is rejected under the same rationale.
Regarding Claim 13, the claim recites limitations that are all substantially the same as limitations of Claim 6, in the form of a non-transitory computer-readable medium. The claim is also directed to performing mental processes without integration into a practical component or significantly more.
Accordingly, Claim 13 is rejected under the same rationale.
Regarding Claim 14, the claim recites limitations that are all substantially the same as limitations of Claim 7, in the form of a non-transitory computer-readable medium. The claim is also directed to performing mental processes without integration into a practical component or significantly more.
Accordingly, Claim 14 is rejected under the same rationale.
Regarding Claim 15:
Step 1: Claim 15 is a process claim. Therefore, claims 15-20 are directed to a statutory category of eligible subject matter.
Step 2A Prong 1: If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the "Mental Processes" grouping of abstract ideas. Here, the claim recites limitations that are substantially the same as the limitations of Claim 1. As a result, and as elaborated above, these limitations are abstract ideas because they are mental processes.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites the additional elements:
“as an input to the selected model” (mere instructions to apply the exception using generic computer components does not provide an inventive concept);
“acquiring battery data” (acquiring battery data amounts to insignificant extra-solution activity because the transmission of data is incidental to the claimed subject matter); and
“of a battery . . . at a prediction time after a last point of the battery data . . . of the battery data . . . the ensemble of models including models trained using different . . . of the batteries, the temporal pairings corresponding to different values of the prediction criteria and including battery data at a first time point on a respective SOH history curve as model input and an SOH value at a second time point after the first time point on the respective SOH history curve as a supervising signal” (amounts to merely generally linking the use of the judicial exception to a particular technological environment or field of use, which do not impose any meaningful limits on practicing the abstract idea).
Step 2B: The claim does not include additional elements considered individually and in combination that are sufficient to amount to significantly more than the judicial exception.
The claim recites the additional elements:
“as an input to the selected model” (mere instructions to apply the exception using generic computer components does not provide an inventive concept);
“acquire historical battery data” (transmitting data is well‐understood, routine, and conventional, see generally Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362; see also buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014); therefore the limitation, which is recited with a high level of generality, remains insignificant extra-solution activity even upon reconsideration); and
“of a battery . . . at a prediction time after a last point of the battery data . . . of the battery data . . . the ensemble of models including models trained using different . . . of the batteries, the temporal pairings corresponding to different values of the prediction criteria and including battery data at a first time point on a respective SOH history curve as model input and an SOH value at a second time point after the first time point on the respective SOH history curve as a supervising signal” (merely generally linking the use of the judicial exception to a particular technological environment or field of use does not provide an inventive concept).
For the reasons above, Claim 15 is rejected as being directed to an abstract idea without significantly more. This rejection applies equally to dependent claims 16-20. The additional limitations of the dependent claims are addressed below.
Regarding Claim 16, the claim recites limitations that are all substantially the same as limitations of Claim 2, in the form of a method. The claim is also directed to performing mental processes without integration into a practical component or significantly more.
Accordingly, Claim 16 is rejected under the same rationale.
Regarding Claim 17, the claim recites limitations that are all substantially the same as limitations of Claim 3, in the form of a method. The claim is also directed to performing mental processes and mathematical concepts without integration into a practical component or significantly more.
Accordingly, Claim 17 is rejected under the same rationale.
Regarding Claim 18, the claim recites limitations that are all substantially the same as limitations of Claim 4, in the form of a method. The claim is also directed to performing mental processes without integration into a practical component or significantly more.
Accordingly, Claim 18 is rejected under the same rationale.
Regarding Claim 19, the claim recites limitations that are all substantially the same as limitations of Claim 6, which includes the limitations of Claim 5, in the form of a method. The claim is also directed to performing mental processes without integration into a practical component or significantly more.
Accordingly, Claim 19 is rejected under the same rationale.
Regarding Claim 20, the claim recites limitations that are all substantially the same as limitations of Claim 7, in the form of a method. The claim is also directed to performing mental processes without integration into a practical component or significantly more.
Accordingly, Claim 20 is rejected under the same rationale.
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.
Claims 1, 4-8, 11-15, and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Abbaraju et al. (hereinafter Abbaraju) (Patent Pub. No. US 2024/0391352 A1) in view of Chen et al. (hereinafter Chen) (Patent Pub. No. US 2024/0095093 A1) and Tian et al. (hereinafter Tian) (“Data-driven battery degradation prediction: Forecasting voltage-capacity curves using one-cycle data”).
Regarding Claim 1, Abbaraju teaches a system, comprising: a memory communicably coupled to a processor and storing instructions that, when executed by the processor, cause the processor to (Para. [0021], “The examples herein provide an SOH prediction method and system that takes into consideration data from multiple information providers in real time and offers accurate predictions of battery SOH”, where the “system” performs the “SOH prediction” operations; see also Fig. 2 and Para. [0044]- [0048], “FIG. 2 illustrates an example hardware configuration of the system 100 . . . may include one or more processors 216, one or more computer-readable media 218 . . . The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218, which may program the processor(s) 216 to perform the functions described herein . . . The computer-readable media 218 may include volatile and nonvolatile memory”, where “the system 100” comprises a “computer-readable media 218” that is “memory” stores “instructions” and is communicably coupled to a processor, “The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218”, which executes the operations, “the processor(s) 216 to perform the functions described herein”):
acquire battery data of a battery (Abstract, “The system receives discharge data . . . the discharge data indicative of a rate of discharge of the battery at a plurality of locations along the route”, where the “system receives discharge data”, which is battery data of a battery because it is “indicative of a rate of discharge of the battery”)
for which a future state-of-health (SOH) is to be predicted at a prediction time after a last time point of the battery data (Abstract, “the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery”, where “the system . . . predict[s] the state of health of the battery”, which can be a future prediction at a prediction time after a last time point of the battery data, see Para. [0025], “predicting the SOH for the battery of a vehicle, such as before, during, or following the completion of a trip”, where a “predicti[on]” of the “SOH for the battery of a vehicle . . . following the completion of a trip” is a future SOH prediction at a prediction time after a last time point of the battery data);
identify a selected model from an ensemble of models according to . . . [an ensemble learning technique] . . . (Abstract, “the system at least one of trains or updates a first machine learning model”, where “a first machine learning model” must be identified to be “train[ed] or update[ed]”, which is selected, according to an “ensemble learning technique”, from an “ensemble” of models, “multiple locally trained heterogenous SOH MLMs”, see Fig. 5 and Para. [0077], “FIG. 5 illustrates an example 500 of an ensemble learning technique for combining the learnings of multiple locally trained heterogenous SOH MLMs according to some implementations . . . to make an SOH prediction for the battery of the vehicle 102 for a trip”)
. . . the battery data (Abstract, “The system receives discharge data . . . the discharge data indicative of a rate of discharge of the battery at a plurality of locations along the route . . . the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery”, where, as discussed above, “the discharge data” is the battery data),
the ensemble of models including models trained using . . . [data] derived from . . . [profiles] of batteries (Para. [0077], “FIG. 5 illustrates an example 500 of an ensemble learning technique for combining the learnings of multiple locally trained heterogenous SOH MLMs according to some implementations . . . to make an SOH prediction for the battery of the vehicle 102 for a trip”, where the “ensemble” includes models trained using local data, “combining the learnings of multiple locally trained”, which uses data derived from profiles of batteries, see Para. [0139], “Charging stations may obtain and maintain data regarding the charging of different types of cars having batteries with different capacities and characteristics . . . the charging stations 112 may share an SOH MLM that may be based on the historic charging profiles, instead of the data itself”; see also Para. [0033], “some examples herein employ a stacked ensemble learning technique that combines the predictions output by a plurality of different locally trained SOH MLMs 126 received from a plurality of different charging stations 112, respectively”),
the . . . [data] corresponding to different values of . . . [the ensemble learning technique] (Para. [0072], “the locally trained SOH MLMs at the different charging stations 112(1) and 112(2) may be heterogeneous, such as due to differences in the types of chargers present at the respective charging stations 112(1) and 112(2). Based on this, implementations herein may employ ensemble learning techniques to combine the estimations determined based on the multiple locally trained SOH MLMs 126(1) and 126(2)”, where the data, used for the “locally trained SOH MLMs”, correspond to different values, “different charging stations”, “differences in the types of chargers”, and “the estimations”, which are of the ensemble learning techniques, see Fig. 5 and Para. [0077], “FIG. 5 illustrates an example 500 of an ensemble learning technique for combining the learnings of multiple locally trained heterogenous SOH MLMs according to some implementations . . . to make an SOH prediction for the battery of the vehicle 102 for a trip”; see also Para. [0138] – [0139], “electric vehicles can be charged using different types of chargers, such as level 1 chargers that charge at 120 volts, level 2 chargers that charge at 208-240 volts, and level 3 chargers that charge at 400-900 volts. The charging times provided by the different types of chargers can vary significantly . . . Charging stations may obtain and maintain data regarding the charging of different types of cars having batteries with different capacities and characteristics”)
and including battery data . . . as model input and an SOH value . . . as a supervising signal (Para. [0037] – [0038], “The ideal charging profile 124 may be used as input to each of the SOH MLMs 126(1)-126(n) to generate a plurality of SOH predictions as the outputs 136 . . . As one example, a stacked generalization in ensemble learning MLM learns to best combine the predictions (outputs 136) from two or more heterogeneous MLMs 126”; Fig. 5; and Para. [0077], “FIG. 5 illustrates an example 500 of an ensemble learning technique for combining the learnings of multiple locally trained heterogenous SOH MLMs according to some implementations . . . to make an SOH prediction for the battery of the vehicle 102 for a trip”, where battery data “The ideal charging profile 124” is included as model input, “may be used as input to each of the SOH MLMs 126(1)-126(n)”, to generate an SOH value, “to generate a plurality of SOH predictions as the outputs 136”, such that the “discharge data” can reasonably be described as an SOH value, which is used as a supervisory signal for model updating, “Based on the received discharge data, the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery”, see Para. [0002], “The system receives discharge data corresponding to traversal of the vehicle along the route, the discharge data indicative of a rate of discharge of the battery at a plurality of locations along the route. Based on the received discharge data, the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery”);
and predict the future SOH using the battery data as an input to the selected model (Abstract, “Based on the received discharge data, the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery. The vehicle may determine an estimated battery state of health based at least on the first machine learning model”, where the selected model, “a first machine learning model”, uses the battery data, “discharge data”, as input for “train[ing]”, “update[ing]”, or “estimate[ion]”, each of which is a use, directly or indirectly, for prediction of the “battery state of health”, which, as discussed above, can be a future SOH prediction, see Para. [0025], “predicting the SOH for the battery of a vehicle, such as before, during, or following the completion of a trip”).
Abbaraju does not explicitly disclose . . . prediction criteria including the prediction time and a time length of (where, while Abbaraju discloses the selection of a model from an ensemble of models according to an ensemble learning technique, the ensemble learning technique is not specifically described as comprising prediction criteria including the prediction time and a time length of battery data; redundant recitations of prediction criteria omitted)
. . . different temporal pairings . . . SOH history curves . . . temporal pairings . . . at a first time point on a respective SOH history curve . . . at a second time point after the first time point on the respective SOH history curve . . . (where the data used in the ensemble learning technique is not specifically described as different temporal pairings at time points on SOH history curves)
However, Chen teaches . . . [identifying a selected model from an ensemble of models] (Fig. 2 and Para. [0042] – [0045], “The model selection module 206 performs ensemble and model selection . . . the model selection module 206 performs an ensemble operation where two or more model's prediction results are combined to generate a better prediction results than individual models . . . The model selection module 206 includes a model selection operation that is based on the previously described functionalities of the model selection module 206”, where “[t]he model selection module 206” selects “a model” from an “ensemble” of models “based on the previously described functionalities of the model selection module 206”)
[according to] prediction criteria including the prediction time and a time length of [data] . . . (Fig. 2 and Para. [0042] – [0045], “The model selection module 206 performs ensemble and model selection. For example, the model selection module 206 calculates performance metrics. The model selection module 206 calculates accuracy metrics for forecasting . . . the model selection module 206 performs a backtesting operation. Backtesting is used to determine how forecasting models perform in the past, and the historical performance is the best estimate of how these models will perform in the future . . . the model selection module 206 fixes the training data length and forecast horizon, generates a forecast and measures the forecasting accuracy in the past. From this, the model selection module 206 obtains one set of accuracy metrics . . . The model selection module 206 includes a model selection operation that is based on the previously described functionalities of the model selection module 206”, where “the previously described functionalities of the model selection module 206”, which comprises a prediction criteria, “calculate[ion of] accuracy metrics”, in order to determine how the selected model(s) “will perform in the future”, which is according to the prediction time, “forecast horizon”, and a time length of data, “the training data length”).
Before the effective filing date of the invention, it would have been obvious to one of ordinary skill in the art to combine the operation for prediction of a future SOH of a battery based on battery data of the battery, wherein a model is identified and selected from an ensemble of models according to an ensemble learning technique, and wherein ensemble training data includes data corresponding to different values of the ensemble learning technique of Abbaraju with the operation for identifying a selected model from an ensemble of models, according to prediction criteria including the prediction time and a time length of data of Chen in order to select models from an ensemble of methods using historical and combination-based accuracy metrics, which is the best estimate for how the models will perform together in the future (Chen, Para. [0042] - [0044], “the model selection module 206 calculates performance metrics. The model selection module 206 calculates accuracy metrics for forecasting . . . the model selection module 206 performs an ensemble operation where two or more model's prediction results are combined to generate a better prediction results than individual models. Ensemble of two or more forecast together generally produce better forecast . . . the historical performance is the best estimate of how these models will perform in the future”), and where performance of the operation according prediction criteria including the prediction time and a time length of battery data will mimic how forecasts are generated in deployment and will allow for a greater set of metrics to assess accuracy, which generates a more reliable accuracy estimate (Chen, Para. [0044], “The model selection module 206 performs a backtesting procedure that mimics how forecasts are generated. Performance is measured in history. Specifically, the model selection module 206 fixes the training data length and forecast horizon, generates a forecast and measures the forecasting accuracy in the past. From this, the model selection module 206 obtains one set of accuracy metrics. The model selection module 206 moves the training data origin by a few time steps and repeats the procedure to obtain another set of accuracy metrics. By doing this multiple times, the model selection module 206 generates a reliable estimate of the accuracy of the forecasting method”).
Additionally, Tian teaches . . . [the construction of SOH history curves using battery data] (Pg. 1, Abstract, “With the wide deployment of rechargeable batteries, battery degradation prediction has emerged as a challenging issue . . . In this article, we explore the prediction of voltage-capacity curves over battery lifetime based on a sequence to sequence (seq2seq) model. We demonstrate that the data of one present voltage-capacity curve can be used as the input of the seq2seq model to accurately predict the voltage-capacity curves” and Pg. 5, Col. 2, Para. 2, “The seq2seq simultaneously predicts the future voltage capacity curves, which contain fruitful information regarding battery degradation . . . the maximum capacity, which is widely used to evaluate battery state of health, can be extracted from the prediction results”, where a “a sequence to sequence (seq2seq) model” is configured to predict current SOH values of batteries to generate historic SOH data, “voltage-capacity” “over battery lifetime” that are SOH values because they can be used “to evaluate battery state of health”, which is used to construct a SOH history curve, “voltage-capacity curves over battery lifetime”, which, similar to the SOH data, can be considered an SOH history curve because it is data “over battery lifetime” that can be used “to evaluate battery state of health”)[,]
[the battery data comprising] different temporal pairings [derived from] SOH history curves [of batteries], [the] temporal pairings [corresponding to different prediction time and time length of battery data – the prediction criteria] . . . (Pg. 3, Fig. 1 and Pg. 3, Col. 1, Para. 2, “Mathematically, given a battery that has been cycled for s times but the only data of recent a cycles are recorded, it is expected to simultaneously forecast the curves of the battery at s + p, s + 2p, …, s + np cycles, where p is a prediction step and n is the number of the predicted curves. a = 1 is an ideal case, and it means that the data of the present cycle are only required to predict the voltage-capacity curves at the future cycles”, where battery data, “given a battery that has been cycled”, comprising different temporal pairings, “s times . . . s + p, s + 2p, …, s + np cycles”, derived from SOH history curves of batteries, “the curves of the battery”, and where prediction time, “s + p, s + 2p, …, s + np cycles, where p is a prediction step”, and time length of battery data, “only data of recent a cycles are recorded”, as depicted in Fig. 1,correspond to different temporal pairings, see Pg. 4, Col. 1, Para. 2, “A moving window with a length of a cycles is used to scan the entire cycles of a battery to obtain the samples containing inputs and outputs. Given the maximum cycle number of N, we can obtain (N-np-a + 1) samples by moving the window at a step of 1 cycle”, where a “window with a length of a” is used to obtain the different pairs associated with the “the samples containing inputs and outputs”)
[and including battery data] at a first time point on a respective SOH history curve [as input] . . . [and an SOH value] at a second time point after the first time point on the respective SOH history curve [as a supervisory signal] . . . (Pg. 3, Figure 1; Pg. 4, Col. 1, Para. 2, “A moving window with a length of a cycles is used to scan the entire cycles of a battery to obtain the samples containing inputs and outputs. Given the maximum cycle number of N, we can obtain (N-np-a + 1) samples by moving the window at a step of 1 cycle”; and Pg. 3, Col. 1, Para. 2, “Mathematically, given a battery that has been cycled for s times but the only data of recent a cycles are recorded, it is expected to simultaneously forecast the curves of the battery at s + p, s + 2p, …, s + np cycles, where p is a prediction step and n is the number of the predicted curves. a = 1 is an ideal case, and it means that the data of the present cycle are only required to predict the voltage-capacity curves at the future cycles”, where the temporal pairs, “s times . . . s + p, s + 2p, …, s + np cycles”, includes battery data, “given a battery that has been cycled”, at a first time point on a respective SOH history curve, shown in Fig. 1 as “Input” in yellow, as input, “the samples containing inputs”, and an SOH value output, “the samples containing . . . outputs”, at a second time point after the first time point on the respective SOH history curve, shown in Fig. 1 as “Output” in red, which acts as a supervisory signal, see Pg. 10, Col. 2, Para. 2, “The loss function for training is the mean squared error (MSE) of the predicted voltage-capacity curves with different weights”).
Before the effective filing date of the invention, it would have been obvious to one of ordinary skill in the art to combine the identifying of a selected model from an ensemble of models according to prediction criteria including a prediction time and a time length of the battery data, as operationalized through an ensemble learning technique, the ensemble of models including models trained using data derived from profiles of batteries, the data corresponding to different values of the prediction criteria and including battery data as model input and an SOH value as a supervising signal and predicting the future SOH using the battery data as an input to the selected mode of Abbaraju in view of Chen with the construction of SOH history curves using battery data, the battery data comprising different temporal pairings derived from SOH history curves of batteries, where the temporal pairings correspond to different prediction time and time length of battery data and include battery data at a first time point on a respective SOH history curve as input and an SOH value at a second time point after the first time point on the respective SOH history curve as a supervisory signal of Tian in order to train the ensemble models to generate predictions that capture complex degradation paths (compare Tian, Pg. 1, Abstract, “battery life defined by capacity loss provides limited information regarding battery degradation” with Tian, Pg. 2, Col. 2, Para. 1, “batteries may have identical initial capacities and cycle lives, but their Ah/Wh throughput can be different due to different degradation paths . . . Instead, a comprehensive and online prediction of battery characteristics is required to provide more information for the prediction of the consequences of the present operations”), which will allow for updated battery management based on predicted SOH data appearing later on the curve, (Abbaraju, Para. [0002], “The vehicle may determine an estimated battery state of health based at least on the first machine learning model, and may receive control information while traversing the route based on the estimated battery state of health to at least partially minimize battery degradation during traversal of the route” with Tian, Pg. 1, Abstract, “we explore the prediction of voltage-capacity curves over battery lifetime based on a sequence to sequence (seq2seq) model . . . This offers an opportunity to update battery management strategies in response to the predicted consequences”), using data generated with reduced time and energy consumption (Tian, Pg. 1, Abstract, “the model features data generation, that is, we can use the data of only one cycle to generate a large spectrum of aging data at the future cycles for developing other battery diagnosis or prognosis methods. In this way, the time and energy consuming battery degradation tests can be sharply reduced”) and to operationalize the generation of battery state of health predictions from battery discharge data (Abbaraju, Abstract, “The system receives discharge data corresponding to traversal of the vehicle along the route . . . Based on the received discharge data, the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery”), where predictions can be based on all previous cycles or, in an ideal case, a limited subset of cycles (Tian, Pg. 3, Col. 2, “given a battery that has been cycled for s times but the only data of recent a cycles are recorded . . . a = 1 is an ideal case, and it means that the data of the present cycle are only required to predict the voltage-capacity curves at the future cycles”), using a flexible method with proven accuracy for multiple prediction periods (Tian, Pg. 10, Col. 1, Para. 2, “The developed method is flexible to incorporate entire voltage-capacity curves as input and output, respectively. Based on a battery degradation dataset with 45 batteries, we demonstrate that the developed model is able to accurately predict voltage-capacity curves at 100, 200, …, and 1000 cycles ahead of the present cycle”).
Regarding Claim 4, Abbaraju in view of Chen and Tian teach the system of claim 1, wherein the instructions further cause the processor to (Abbaraju, Fig. 2 and Abbaraju, Para. [0044]- [0048], “FIG. 2 illustrates an example hardware configuration of the system 100 . . . may include one or more processors 216, one or more computer-readable media 218 . . . The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218, which may program the processor(s) 216 to perform the functions described herein . . . The computer-readable media 218 may include volatile and nonvolatile memory):
apply an estimation model configured to predict current SOH values of batteries to historical battery data of the batteries, the historical battery data collected over an interval ending at a last time point before the prediction time, to generate historical SOH data corresponding to time points in the interval (Abbaraju, Para. [0002], “The system receives discharge data . . . indicative of a rate of discharge of the battery at a plurality of locations along the route. Based on the received discharge data, the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery. The vehicle may determine an estimated battery state of health based at least on the first machine learning model, and may receive control information while traversing the route based on the estimated battery state of health”, where an estimation model configured to predict current SOH values of batteries, the “machine learning model configured for” “estimate[ion of] battery state of health”, is applied to historical battery data of batteries, “discharge data”, to generate historical SOH data corresponding to time points in the interval, “estimated battery state of health” “indicative of a rate of discharge of the battery at a plurality of locations along the route” that is historical because the battery data can be historical when received, see Abbaraju, Para. [0025], “predicting the SOH for the battery of a vehicle, such as before, during, or following the completion of a trip”, where the data is historical if the battery data is received “before . . . a trip”, and, regardless of its historical status of the battery at the time, becomes historical with the progression of time, such that collected over an interval ending at a last point before the prediction time; see also Abbaraju, Para. [0036], “As one example, considering each trip as an edge scenario, the SOH MLM 122 is trained in federated fashion with the updated data for newer trips. A federated learning approach is used to train a decentralized SOH MLM 122 across multiple edge scenarios, and newly trained SOH MLMs may be added to previously trained SOH MLMs to maintain the influence of historical data on the SOH MLM 122”, where the previously current data is specifically referred to as “historical data” and estimation models “trained in a federated fashion” are configured to predict current values of a plurality of batteries, “considering each trip as an edge scenario, the SOH MLM 122 is trained . . . with the updated data for newer trips . . . across multiple edge scenarios”, where the “SOH MLM” is “trained” with battery data from “each trip” across ”multiple” “trips”); and
construct the SOH history curves using the historical SOH data (Tian, Pg. 1, Abstract, “With the wide deployment of rechargeable batteries, battery degradation prediction has emerged as a challenging issue . . . In this article, we explore the prediction of voltage-capacity curves over battery lifetime based on a sequence to sequence (seq2seq) model. We demonstrate that the data of one present voltage-capacity curve can be used as the input of the seq2seq model to accurately predict the voltage-capacity curves” and Tian, Pg. 5, Col. 2, Para. 2, “The seq2seq simultaneously predicts the future voltage capacity curves, which contain fruitful information regarding battery degradation . . . the maximum capacity, which is widely used to evaluate battery state of health, can be extracted from the prediction results”, where a “a sequence to sequence (seq2seq) model” is configured to predict current SOH values of batteries to generate historic SOH data, “voltage-capacity” “over battery lifetime” that are SOH values because they can be used “to evaluate battery state of health”, which is used to construct a SOH history curve, “voltage-capacity curves over battery lifetime”, which, similar to the SOH data, can be considered an SOH history curve because it is data “over battery lifetime” that can be used “to evaluate battery state of health”; which, in the context of Abbaraju, are a plurality of curves for each model, see Abbaraju, Para. [0033], “locally trained SOH MLMs 126 are shared by the charging stations 112 to the service computing device 108 . . . To realize this approach, some examples herein employ a stacked ensemble learning technique that combines the predictions output by a plurality of different locally trained SOH MLMs 126 received from a plurality of different charging stations 112, respectively, to train and update the SOH MLM 122”).
Regarding Claim 5, Abbaraju in view of Chen and Tian teach the system of claim 4, wherein the instructions further cause the processor to (Abbaraju, Fig. 2 and Abbaraju, Para. [0044]- [0048], “FIG. 2 illustrates an example hardware configuration of the system 100 . . . may include one or more processors 216, one or more computer-readable media 218 . . . The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218, which may program the processor(s) 216 to perform the functions described herein . . . The computer-readable media 218 may include volatile and nonvolatile memory)
characterize the historical SOH data as first time sequency data and second time sequency data relative to the respective SOH history curve (Tian, Pg. 3, Figure 1, where the historical SOH data, “Input” from “Historical cycles”, up to and including the “Present cycle”, are characterized by cycle, such that, the data corresponding with the “s” are characterized as the “Present cycle”, which is within the broadest reasonable interpretation of second time sequence data because it is the latest input data in the time sequence relative to the SOH history curve, “the voltage capacity curves”; and the data corresponding with any recorded cycle before the “Present cycle”, such as “s-a+1”, are characterized as the previous cycle(s), which is within the broadest reasonable interpretation of first time sequence data because it is earlier time sequence data in the respective SOH history curve, “the voltage capacity curves”; see also Tian, Pg. 3, Col. 1, Para. 2, “given a battery that has been cycled for s times but the only data of recent a cycles are recorded, it is expected to simultaneously forecast the curves of the battery at s + p, s + 2p, …, s + np cycles, where p is a prediction step and n is the number of the predicted curves. a = 1 is an ideal case, and it means that the data of the present cycle are only required to predict the voltage-capacity curves at the future cycles”),
wherein the first time sequency data includes data corresponding to a first portion of the respective SOH history curve, and wherein the second time sequency data includes data corresponding to a second portion of the respective SOH history curve that is after the first portion (Tian, Pg. 3, Figure 1, where the first time sequency data, the data corresponding with any recorded cycle before the “Present cycle”, such as “s-a+1”, are corresponding to a first portion of the respective SOH history curve, “the voltage capacity curves”, and the second time sequency data, the data corresponding with the “s” “Present cycle”, are corresponding to a second portion of the respective SOH history curve that is after the first portion of the curve, “the voltage capacity curves”).
The reasons for obviousness were discussed in regard to the rejection of Claim 1 above and remain applicable here.
Regarding Claim 6, Abbaraju in view of Chen and Tian teach the system of claim 5, wherein the instructions further cause the processor to (Abbaraju, Fig. 2 and Abbaraju, Para. [0044]- [0048], “FIG. 2 illustrates an example hardware configuration of the system 100 . . . may include one or more processors 216, one or more computer-readable media 218 . . . The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218, which may program the processor(s) 216 to perform the functions described herein . . . The computer-readable media 218 may include volatile and nonvolatile memory)
use battery data corresponding to the first time sequency data and the second time sequency data to train each model in the ensemble of models (Abbaraju, Abstract, “Based on the received discharge data, the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery”, where the received battery data is used for model training, which, in view of Tian, includes the first time sequence data and the second time sequence SOH data, see Tian, Pg. 3, Figure 1, where the data corresponding with any recorded cycle before the “Present cycle”, such as “s-a+1”, are the first time sequence data and the data corresponding with the “s” “Present cycle” are the second time sequence data, and where each of the “ensemble” of models are “locally trained” as a “SOH MLMs 126”, such that the local battery data is directly used to train each model, see Abbaraju, Para. [0033], “locally trained SOH MLMs 126 are shared by the charging stations 112 to the service computing device 108 . . . To realize this approach, some examples herein employ a stacked ensemble learning technique that combines the predictions output by a plurality of different locally trained SOH MLMs 126 received from a plurality of different charging stations 112, respectively, to train and update the SOH MLM 122”, whereas other battery data is indirectly used in a “fus[ion] federated learning and ensemble learning”, see Abbaraju, Para. [0069] – [], “the employed machine learning techniques may fuse federated learning and ensemble learning to efficiently combine information from different information providers . . . implementations herein may employ ensemble learning techniques to combine the estimations determined based on the multiple locally trained SOH MLMs 126(1) and 126(2). With the limited data shared from the RSUs 109(1)-109(4), and the combined output from the SOH MLMs 126(1) and 126(2) of the charging stations 112(1) and 112(2), respectively, the previous SOH MLM 122 from OEM or fleet operator 110 may be retrained or otherwise updated using a federated learning technique. The updated SOH MLM 122 may be shared back to the vehicle 102 for enabling the vehicle 102 to perform an accurate estimation of the SOH of the battery 115”),
wherein the first time sequency data and the second time sequency data for each model are located at different time points of the respective SOH history curve (Tian, Pg. 3, Figure 1, where, as discussed above, the data corresponding with any recorded cycle before the “Present cycle”, such as “s-a+1”, are the first time sequence data and the data corresponding with the “s” “Present cycle” are the second time sequence data, which, as indicated by the gold dots, are located at different time points of the respective SOH history curve, “the voltage capacity curves”; which, in the context of Abbaraju, are for each model, see Abbaraju, Para. [0033], “locally trained SOH MLMs 126 are shared by the charging stations 112 to the service computing device 108 . . . To realize this approach, some examples herein employ a stacked ensemble learning technique that combines the predictions output by a plurality of different locally trained SOH MLMs 126 received from a plurality of different charging stations 112, respectively, to train and update the SOH MLM 122”).
The reasons for obviousness were discussed in regard to the rejection of Claim 1 above and remain applicable here.
Regarding Claim 7, Abbaraju in view of Chen and Tian teach the system of claim 1, wherein the instructions further cause the processor to (Abbaraju, Fig. 2 and Abbaraju, Para. [0044]- [0048], “FIG. 2 illustrates an example hardware configuration of the system 100 . . . may include one or more processors 216, one or more computer-readable media 218 . . . The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218, which may program the processor(s) 216 to perform the functions described herein . . . The computer-readable media 218 may include volatile and nonvolatile memory):
generate SOH predictions for each model in the ensemble of models using historical battery data as an input to each model (Abbaraju, Abstract, “Based on the received discharge data, the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery. The vehicle may determine an estimated battery state of health based at least on the first machine learning model”, where the “state of health of the battery” prediction is generated using “battery” data as input to the model, where, in view of Chen, it is generated for each model in the ensemble using historic data, see Chen, Fig. 2 and Chen, Para. [0042] – [0044], “The model selection module 206 performs ensemble and model selection . . . the model selection module 206 performs an ensemble operation where two or more model's prediction results are combined to generate a better prediction results than individual models . . . Backtesting is used to determine how forecasting models perform in the past, and the historical performance is the best estimate of how these models will perform in the future”, the “ensemble operation” generates “prediction results” from “historical” data for the various “combin[ations]” of “two or more” models in the “ensemble”);
calculate a SOH prediction error for each model by comparing the SOH prediction to estimated SOH values (Abbaraju, Abstract, “The system receives discharge data . . . the discharge data indicative of a rate of discharge of the battery at a plurality of locations along the route . . . the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery”, where, as discussed above, “the discharge data” is the battery data and the “predict[ion of] the state of health” is the SOH; see also Chen, Para. [0042] – [0044], “The model selection module 206 performs ensemble and model selection . . . The model selection module 206 calculates accuracy metrics for forecasting, such as error . . . the model selection module 206 performs an ensemble operation where two or more model's prediction results are combined to generate a better prediction results than individual models”, where a person of ordinary skill in the art would understand a “forecasting” “error” to require a comparison of the ”forecasting”, which is a SOH prediction in this instance, with a verified truth value, which is a SOH value in this instance, and where, given the necessary practicalities of measuring values, any value is an estimate based on the capabilities of the measurement technique and the necessary rounding decisions required for selecting finite values; see generally Abbaraju, Fig. 7 and Abbaraju, Para. [0087], “the vehicle computing device 104 may perform at least one action based on the estimated battery SOH”, where “estimated . . . SOH” values are used for decision making); and
create an assignment of each model to future SOH prediction criteria, wherein the prediction criteria include the prediction time and the time length of the battery data, wherein the selected model is chosen based on the assignment (Abbaraju, Fig. 6, where each model, “SOH MLM” “122(1)” through “122(N)” are assigned to, and selected for based on this assignment, to a specific SOH prediction criteria, “Edge Scenario”; see also Abbaraju, Para. [0078], “FIG. 6 illustrates an example 600 of federated learning that is used in combination with the ensemble learning of FIG. 5 . . . For example, the SOH MLM 122 may be repeatedly retrained for every trip taken by the vehicle 102 based on current real-time data from the different information providers, such as the charging stations and the RSUs, for the corresponding trip. In the example of FIG. 6, each trip may be a new edge scenario 602. The SOH MLM 122 is trained in a federated fashion with the updated data for a new trip, and the newly trained SOH MLM is added to the previously trained SOH MLMs to create an aggregated SOH MLM 122(A). Thus, a federated learning approach is used to train a decentralized MLM (e.g., deep neural network) across multiple edge scenarios 602(1)-602(N)”, where each “SOH MLM” is assigned and iteratively selected for based on a criteria specific to the “new edge scenario”, such as the “data for a new trip”, which can be for SOH predictions, see Abbaraju, Para. [0025], “predicting the SOH for the battery of a vehicle, such as before, during, or following the completion of a trip”, and which, in view of Chen, the prediction criteria include the prediction time and the time length of the battery data, see Chen, Fig. 2 and Chen, Para. [0042] – [0045], “The model selection module 206 performs ensemble and model selection. For example, the model selection module 206 calculates performance metrics. The model selection module 206 calculates accuracy metrics for forecasting . . . the model selection module 206 performs a backtesting operation. Backtesting is used to determine how forecasting models perform in the past, and the historical performance is the best estimate of how these models will perform in the future . . . the model selection module 206 fixes the training data length and forecast horizon, generates a forecast and measures the forecasting accuracy in the past. From this, the model selection module 206 obtains one set of accuracy metrics . . . The model selection module 206 includes a model selection operation that is based on the previously described functionalities of the model selection module 206”, where “the previously described functionalities of the model selection module 206”, which comprises “calculate[ion of] accuracy metrics” in order to determine how the selected model(s) “will perform in the future”, which is according to a prediction time, “forecast horizon”, and a time length of data, “the training data length”).
The reasons for obvious were discussed in regard to the rejection of Claim 1 above and remain applicable here.
Regarding Claim 8, Abbaraju in view of Chen and Tian teach a non-transitory computer-readable medium including instructions that, when executed by a processor, cause the processor to (Abbaraju, Para. [0021], “The examples herein provide an SOH prediction method and system that takes into consideration data from multiple information providers in real time and offers accurate predictions of battery SOH”, where the “system” performs the “SOH prediction” operations; see also Abbaraju, Fig. 2 and Abbaraju, Para. [0044]- [0048], “FIG. 2 illustrates an example hardware configuration of the system 100 . . . may include one or more processors 216, one or more computer-readable media 218 . . . The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218, which may program the processor(s) 216 to perform the functions described herein . . . the computer-readable media 218 may be a tangible non-transitory medium”, where “the system 100” comprises a “the computer-readable media 218” that is a “non-transitory medium”, which stores “instructions” executed by a processor, “The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218”, which perform the operations, “the processor(s) 216 to perform the functions described herein”):
acquire historical battery data of a battery . . . (Abbaraju, Abstract, “The system receives discharge data . . . the discharge data indicative of a rate of discharge of the battery at a plurality of locations along the route”, where the “system receives discharge data”, which is battery data of a battery because it is “indicative of a rate of discharge of the battery”, which, includes historical data captured by roadside units “RSUs” along the route, see Abbaraju, Fig. 1 and Abbaraju, Para. [0040], “the vehicle data, such as from RSUs 1-m may be captured along a route traversed by the vehicle 102 until the vehicle 102 reaches a destination location, and may include discharge data while the vehicle 102 travels along the route, such as state of charge (SOC) data and SOH estimation data based on this, as indicated at 152. In addition, the vehicle data may include charging data from charging stations 1-n, used along the route, and SOH estimation data based on this, as indicated at 154”; and where, in view of Chen, the battery data includes historical battery data to evaluate accuracy, Chen, Para. [0042] - [0044], “the model selection module 206 calculates performance metrics. The model selection module 206 calculates accuracy metrics for forecasting . . . the historical performance is the best estimate of how these models will perform in the future”; see also Abbaraju, Para. [0079, “This enables the historical charging and discharge data of the battery to continue to be relevant in the updated versions of the SOH MLM 122”)
and predict the future SOH using the historical battery data as an input to the selected model (Abbaraju, Abstract, “Based on the received discharge data, the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery. The vehicle may determine an estimated battery state of health based at least on the first machine learning model”, where the selected model, “a first machine learning model”, uses the battery data, “discharge data”, as input for “train[ing]”, “update[ing]”, or “estimate[ion]”, each of which is a use, directly or indirectly, for prediction of the “battery state of health”, which, as discussed above, can be a future SOH prediction, see Abbaraju, Para. [0025], “predicting the SOH for the battery of a vehicle, such as before, during, or following the completion of a trip”; and where, as discussed above, the data includes historical data, see Abbaraju, Fig. 1 and Abbaraju, Para. [0040], “the vehicle data, such as from RSUs 1-m may be captured along a route traversed by the vehicle 102 until the vehicle 102 reaches a destination location, and may include discharge data while the vehicle 102 travels along the route, such as state of charge (SOC) data and SOH estimation data based on this, as indicated at 152. In addition, the vehicle data may include charging data from charging stations 1-n, used along the route, and SOH estimation data based on this, as indicated at 154”; and where, in view of Chen, the battery data includes historical battery data to evaluate accuracy, Chen, Para. [0042] - [0044], “the model selection module 206 calculates performance metrics. The model selection module 206 calculates accuracy metrics for forecasting . . . the historical performance is the best estimate of how these models will perform in the future”; see also Abbaraju, Para. [0079, “This enables the historical charging and discharge data of the battery to continue to be relevant in the updated versions of the SOH MLM 122”).
The reasons for obviousness were discussed in regard to the rejection of Claim 1 above and remain applicable here.
The remaining limitations are substantially the same as limitations of Claim 1, therefore it is rejected under the same rationale.
Regarding Claim 11, the additional elements of the dependent claim are substantially the same as limitations of Claim 4, therefore it is rejected under the same rationale.
Regarding Claim 12, the additional elements of the dependent claim are substantially the same as limitations of Claim 5, therefore it is rejected under the same rationale.
Regarding Claim 13, the additional elements of the dependent claim are substantially the same as limitations of Claim 6, therefore it is rejected under the same rationale.
Regarding Claim 14, the additional elements of the dependent claim are substantially the same as limitations of Claim 7, therefore it is rejected under the same rationale.
Regarding Claim 15, Abbaraju teaches a method, comprising: . . . (Para. [0021], “The examples herein provide an SOH prediction method and system that takes into consideration data from multiple information providers in real time and offers accurate predictions of battery SOH”).
The remaining limitations are substantially the same as limitations of Claim 8, therefore it is rejected under the same rationale.
Regarding Claim 18, the additional elements of the dependent claim are substantially the same as limitations of Claim 4, therefore it is rejected under the same rationale.
Regarding Claim 19, the additional elements of the dependent claim are substantially the same as limitations of Claim 6, which includes the limitations of Claim 5, therefore it is rejected under the same rationale.
Regarding Claim 20, the additional elements of the dependent claim are substantially the same as limitations of Claim 7, therefore it is rejected under the same rationale.
Claims 2-3, 9-10, and 16-17 are rejected under 35 U.S.C. 103 as being unpatentable over Abbaraju in view of Chen, Tian, and Van Damme et al. (hereinafter Van Damme) (Patent Pub. No. US 2023/0344679 A1).
Regarding Claim 2, Abbaraju in view of Chen and Tian teach the system of claim 1, wherein the instructions further cause the processor to (Abbaraju, Fig. 2 and Abbaraju, Para. [0044]- [0048], “FIG. 2 illustrates an example hardware configuration of the system 100 . . . may include one or more processors 216, one or more computer-readable media 218 . . . The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218, which may program the processor(s) 216 to perform the functions described herein . . . The computer-readable media 218 may include volatile and nonvolatile memory)
train an estimation model to predict current state-of-health (SOH) values of batteries at respective training time points (Abbaraju, Abstract, “the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery. The vehicle may determine an estimated battery state of health based at least on the first machine learning model, and may receive control information while traversing the route based on the estimated battery state of health to at least partially minimize battery degradation during traversal of the route”, where “the first machine learning model” is an estimation model because it “estimated [the] battery state of health”, as the SOH values, which is current values at respective training time points when “degradation during traversal of the route” is the subject of estimation, see also Abbaraju, Para. [0025], “predicting the SOH for the battery of a vehicle, such as before, during, or following the completion of a trip”),
wherein instructions cause the processor to (Abbaraju, Fig. 2 and Abbaraju, Para. [0044]- [0048], “FIG. 2 illustrates an example hardware configuration of the system 100 . . . may include one or more processors 216, one or more computer-readable media 218 . . . The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218, which may program the processor(s) 216 to perform the functions described herein . . . The computer-readable media 218 may include volatile and nonvolatile memory)
train the estimation model by using current values of the battery data at the respective training time points as training data for the estimation model (Abbaraju, Para. [0002], “The system receives discharge data corresponding to traversal of the vehicle along the route, the discharge data indicative of a rate of discharge of the battery at a plurality of locations along the route. Based on the received discharge data, the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery. The vehicle may determine an estimated battery state of health based at least on the first machine learning model,”, where the “first machine learning model” is an estimation model because it “estimated [the] battery state of health”, which is “train[ed]” using current values of battery data at the respective training time points, “receives discharge data corresponding to traversal of the vehicle along the route”)
and . . . SOH values of the batteries . . . for the estimation model . . . of the estimation model and update the estimation model (Abbaraju, Abstract, “the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery. The vehicle may determine an estimated battery state of health based at least on the first machine learning model, and may receive control information while traversing the route based on the estimated battery state of health to at least partially minimize battery degradation during traversal of the route”, where “predicting the state of health of the battery” and an estimation model that is updated, “the first machine learning model” to “estimate[] battery state of health”, are disclosed).
Abbaraju in view of Chen and Tian do not explicitly disclose . . . using lab-measured . . . as a supervising signal . . . to generate a loss value (where the processes of model updating are not specifically described).
However, Van Damme teaches . . . using lab-measured [values] . . . as a supervising signal . . . to generate a loss value [to update a model] (Para. [0061], “The set of weights of the ML model may be updated based on the estimated loss”, where the training datasets, including the ground truth values needed to generate the loss values, can be lab-measured, “systematic lab measurements”, see Para. [0109], “The training data sets could be obtained, for example, by real-time in the field measurements, systematic lab measurements, and/or simulation”; see generally Para. [0110], “Simulations enable the correct labelling of simulator data and may be used to generate millions of training data sets”, where, though discussed in regard to simulator data, the labels for the “lab measur[ed]” “training data” allow it to function as a supervisory signal to generate the loss value).
Before the effective filing date of the invention, it would have been obvious to one of ordinary skill in the art to combine the training of an the estimation model by using current values of the battery data at the respective training time points as training data to update the estimation model to predict SOH values of the batteries of Abbaraju in view of Chen and Tian with the use of lab-measured values as a supervisory signal to generate a loss value to update a model of Van Damme in order to use systematic lab measurements, which will produce consistent and high quality data (Van Damme, Para. [0109], “The training data sets could be obtained, for example, by real-time in the field measurements, systematic lab measurements, and/or simulation”, where the use of “systematic lab measurements” will produce consistent high-quality data, as opposed to “simulations”, which may be based on faulty assumptions, or “field measurements”, which may be altered by unexpected “real-time” conditions outside of the system’s control), to improve the accuracy of the estimation models (Van Damme, Para. [0073] –[0074], “A loss computation module 55c receives the target channel response data (e.g. Target Hlog) of the training batch from the training data generation module 55b, compares each target channel response data with the corresponding predicted/estimated target channel frequency response data and calculates (using an appropriate loss function for the type of ML model topology) a loss for use in updating via update weights module 55f the weights/parameters of the ML model of apparatus . . . [until] a particular model accuracy has been achieved”).
Regarding Claim 3, Abbaraju in view of Chen, Tian, and Van Damme teach the system of claim 2, wherein the instructions further cause the processor to (Abbaraju, Fig. 2 and Abbaraju, Para. [0044]- [0048], “FIG. 2 illustrates an example hardware configuration of the system 100 . . . may include one or more processors 216, one or more computer-readable media 218 . . . The processor(s) 216 may be configured to fetch and execute computer-readable instructions stored in the computer-readable media 218, which may program the processor(s) 216 to perform the functions described herein . . . The computer-readable media 218 may include volatile and nonvolatile memory)
preprocess the current battery data at the respective training time points before use as the training data for the estimation model by at least one of: filtering out relevant battery features and applying empirical equations related to physics of the batteries to the current values of the battery data (Abbaraju, Abstract, “Based on the received discharge data, the system at least one of trains or updates a first machine learning model configured for predicting the state of health of the battery. The vehicle may determine an estimated battery state of health based at least on the first machine learning model, and may receive control information while traversing the route based on the estimated battery state of health to at least partially minimize battery degradation during traversal of the route”, where the current battery data at the respective training time points, “discharge data . . . receive[d] . . . while traversing the route”, is used as “train[ing]” data for a “model”, which is an “estimate[ion]” model, and where, in view of Chen, the data is preprocessed before use, see Chen, Fig. 2 and Chen, Para. [0035] – Para. [0037], “The pre-processing module 202 performs pre-processing operations on the time series data . . . the pre-processing module 202 performs an outlier detection and imputation operation . . . Outliers can have a large influence on the forecasting methods, and imputing the outliers improves the forecast accuracy”, where the “pre-processing operations” include an “outlier detection and imputation operation”, which filters out data relevant data with a “large influence”; see also Abbaraju, Abstract, “the discharge data indicative of a rate of discharge of the battery”, where, as applied to Abbaraju, the outlier values may be data points corresponding to features of the “battery”, such as “rate of discharge”).
Before the effective filing date of the invention, it would have been obvious to one of ordinary skill in the art to combine the use of current battery data at the respective training time points, which includes values of data features, for training of an estimation model of Abbaraju in view of Chen, Tian, and Van Damme with the preprocessing of data before use, wherein the processing includes filtering out relevant data in further view of Chen in order to detect and impute outliers, which will improve prediction accuracy (Chen, Para. [0037], “the pre-processing module 202 performs an outlier detection and imputation operation. This operation detects outliers in the time series as well as impute the outliers. Outliers can have a large influence on the forecasting methods, and imputing the outliers improves the forecast accuracy”).
Regarding Claim 9, the additional elements of the dependent claim are substantially the same as limitations of Claim 2, therefore it is rejected under the same rationale.
Regarding Claim 10, the additional elements of the dependent claim are substantially the same as limitations of Claim 3, therefore it is rejected under the same rationale.
Regarding Claim 16, the additional elements of the dependent claim are substantially the same as limitations of Claim 2, therefore it is rejected under the same rationale.
Regarding Claim 17, the additional elements of the dependent claim are substantially the same as limitations of Claim 3, therefore it is rejected under the same rationale.
Response to Arguments
Applicant's arguments filed on June 25th, 2026 have been fully considered. Each argument is addressed in detail below.
I. Applicant argues the objections to claims 6, 11-13, and 15-20 should be withdrawn (Applicant’s Remarks, 06/25/2026, Pg. 11, Section “I”).
The amendments to the claims have overcome each and every objection to the claims, as previously communicated in the 03/26/2026 Office action. As a result, these objections have been withdrawn.
II. Applicant argues the rejections of claims 1-20, under 35 USC § 101, should be withdrawn (Applicant’s Remarks, 06/25/2026, Pg. 11-15, Section “II”).
1) First, Applicant argues amended claim 1 “does not recite a mental process” because “amended claim 1 is not directed merely to forming an opinion regarding a future state of a battery. Rather, amended claim 1 recites a specific machine-learning architecture and model- selection process in which an ensemble of models includes models trained using different temporal pairings derived from SOH history curves of batteries. The claimed temporal pairings correspond to different values of prediction criteria and include battery data at a first time point on a respective SOH history curve as model input and an SOH value at a second time point after the first time point on the respective SOH history curve as a supervising signal” (Applicant’s Remarks, Pg. 12, Para. 2). Additionally, Applicant argues “[i]dentifying a selected model from such an ensemble according to prediction criteria including a prediction time and a time length of the battery data and then predicting a future SOH using the selected model, cannot practically be performed in the human mind or with pen and paper” (id.).
However, identifying a selected model from such an ensemble according to prediction criteria including a prediction time and a time length and then predicting a future SOH are well-within the definition of evaluations, judgments, and opinions that can be performed by a human using pen and paper (see MPEP 2106.04(a)(2)(III), “The courts consider a mental process (thinking) that "can be performed in the human mind, or by a human using a pen and paper" to be an abstract idea . . . the "mental processes" abstract idea grouping is defined as concepts performed in the human mind, and examples of mental processes include observations, evaluations, judgments, and opinions”). Additionally, while Applicant is correct that the claim recites machine-learning architecture and ensemble models, a claim that requires a computer may still recite a mental process (see MPEP 2106.04(a)), “A Claim That Requires a Computer May Still Recite a Mental Process . . . examiners should review the specification to determine if the claimed invention is described as a concept that is performed in the human mind and applicant is merely claiming that concept performed 1) on a generic computer, or 2) in a computer environment, or 3) is merely using a computer as a tool to perform the concept. In these situations, the claim is considered to recite a mental process”). Here, as discussed in detail above, the additional elements relating to machine-learning architecture, ensemble models, battery data, and SOH curves amount to merely claiming that concept performed on a generic computer, in a computer environment or for a particular field of use, and merely using a computer as a tool to perform the concept (id.)
As a result, the arguments are not persuasive.
2) Second, Applicant argues “even if the claim were considered to recite an abstract idea, amended claim 1 integrates any alleged abstract idea into a practical application” because it “is directed to a specific improvement in battery state-of-health prediction technology. In particular, the claim uses battery data having a last time point before a prediction time, identifies a selected model from an ensemble of differently trained models based on prediction criteria including the prediction time and the time length of the battery data, and uses the selected model to predict a future SOH of the battery” (Applicant’s Remarks, Pg. 12, Para. 3). Additionally, Applicant argues “the claimed ensemble is not an arbitrary collection of models; instead, the models are trained using different temporal pairings derived from SOH history curves, where each temporal pairing uses battery data at a first time point as model input and an SOH value at a later second time point as a supervising signal”, which “provides a technical solution to a technical problem in battery diagnostics” because “[t]he claimed system improves future SOH prediction by selecting a model that is appropriate for the prediction time and available battery-data time length, thereby accounting for variations in the amount and timing of available battery data” (Applicant’s Remarks, Pg. 12-13, Para. 3-1). Furthermore, Applicant argues this alleged improvement is described in the specification and “the claim therefore does not merely apply an abstract idea using generic computer components. Instead, amended claim 1 recites a particular machine-learning model-selection and prediction process tied to battery data, SOH history curves, temporal training pairings, prediction criteria, and a future SOH prediction at a prediction time after the last time point of the battery data”, which “impose meaningful limits on the claimed subject matter and integrate the alleged abstract idea into a practical application” (Applicant’s Remarks, Pg. 13, Para. 1-2).
However, the specification sets forth the alleged improvements to in battery state-of-health prediction technology, through ensemble model selection using SOH curves and battery data, in a conclusory manner (MPEP 2106.04(d)(1), “Conversely, if the specification explicitly sets forth an improvement but in a conclusory manner (i.e., a bare assertion of an improvement without the detail necessary to be apparent to a person of ordinary skill in the art), the examiner should not determine the claim improves technology”; see also Applicant’s Spec. Para. [0008], [0025], [0051], and [0060]). Additionally, the claim recites limitations, such as selection of ensemble models and use of battery data and SOH curves with various temporal characteristics have broad applicability across many fields of endeavor endeavor (see MPEP 2106.05(f), “A claim having broad applicability across many fields of endeavor may not provide meaningful limitations that integrate a judicial exception into a practical application or amount to significantly more”), such as to fail to reflect the alleged improvements (see MPEP 2106.04(d)(1)), “the claim must be evaluated to ensure that the claim itself reflects the disclosed improvement”). Furthermore, as discussed in detail above, the recitations of ensemble models as well as battery data and SOH curves with temporal characteristics amounts to merely indicating a field of use or technological environment in which to apply a judicial exception, such as to fail to impose meaningful limits on the claimed subject matter or to integrate the alleged abstract idea into a practical application (see MPEP 2106.05(h), “limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application”).
As a result, the arguments are not persuasive.
3) Third, Applicant argues amended claim 1 “the ordered combination of the limitations of amended claim 1 amounts to significantly more than any alleged abstract idea” because “[t]he claim recites a particular arrangement of an ensemble of models trained using different temporal pairings derived from SOH history curves, selection of a model according to prediction criteria including the prediction time and time length of the battery data, and prediction of a future SOH using the selected model. This ordered combination is not merely conventional data acquisition or generic computer implementation but instead provides a specific technological approach for improving future battery SOH prediction” (Applicant’s Remarks, Pg. 13, Para. 3).
However, as discussed in detail above, Applicant-cited limitations, such as selection of ensemble models and use of battery data and SOH curves with various temporal characteristics have broad applicability across many fields of endeavor, such that it fails to amount to significantly more than the judicial exception (see MPEP 2106.05(f), “A claim having broad applicability across many fields of endeavor may not provide meaningful limitations that integrate a judicial exception into a practical application or amount to significantly more”). Furthermore, as discussed in detail above, the recitations of ensemble models as well as battery data and SOH curves with temporal characteristics amounts to merely indicating a field of use or technological environment in which to apply a judicial exception, such as to fail to impose meaningful limits on the claimed subject matter or to integrate the alleged abstract idea into a practical application (see MPEP 2106.05(h), “limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application”).
As a result, the argument is not persuasive.
4) Fourth, Applicant argues the remaining claims are subject matter eligible because independent claims 8 and 15 were amended in a similar matter as claim 1, which Applicant argues recites eligible subject matter, and, as a result, dependent claims 2-7, 9-14, and 16-20 are also subject matter eligible because each depends on one of the independent claims (Applicant’s Remarks, Pg. 14-15, Para. 2-1). However, as discussed in detail above, arguments in favor of the subject matter eligibility of claim 1 are not persuasive.
As a result, the arguments are not persuasive.
III. Applicant argues the rejections of claims 1-20, under 35 USC § 103, should be withdrawn (Applicant’s Remarks, 06/25/2026, Pg. 15-18, Sections “III”, “IV”, “V”).
In response to Applicant’s amendments, the previously communicated rejections under 35 U.S.C. § 103, have been withdrawn. However, Applicants arguments are not persuasive in light of the new grounds for rejection, under 35 U.S.C. § 103, discussed in detail above. The new grounds of rejection rely on new combinations of the existing prior art of record to teach the new combinations of elements in the amended claims, which were not presented in these arrangements in any of the previously presented claims. As a result, Applicant arguments against the previously communicated rejections under 35 U.S.C. § 103 are rendered moot.
However, for clarity of the record and in the interest of compact prosecution, arguments still relevant to the new grounds of rejection are discussed below.
Specifically, as discussed in detail above, the new grounds of rejection rely on a combination of Abbaraju in view of Chen and Tian to teach the amended independent claims. However, Applicant argues “Neither Abbaraju nor Chen, individually or in any combination, discloses or suggests the recitations of amended claim 1” and “Tian fails to cure the deficiencies of the aforementioned references” (Applicant’s Remarks, Pg. 15-17, Para. 2-2 and Pg. 18, Para. 4). This argument is insufficient because it amounts to a general allegation that the claim defines a patentable invention without specifically pointing out how the language of the claim patentably distinguishes it from the references (see 37 C.F.R. 1.111(b), “In order to be entitled to reconsideration or further examination, . . . The reply by the applicant or patent owner must . . . specifically points out the supposed errors in the examiner’s action . . . A general allegation that the claims define a patentable invention without specifically pointing out how the language of the claims patentably distinguishes them from the references does not comply with the requirements of this section”).
As a result, the argument is not persuasive.
IV. Applicant argues the rejections of claims 1-20, under 35 USC § 112, should be withdrawn (Applicant’s Remarks, 06/25/2026, Pg. 19, Section “Claims 1-20 are rejected under 35 U.S.C. § 112(b) as being indefinite”).
The amendments to the claims have overcome all of the rejections of the claims under 35 USC § 112, as previously communicated in the 03/26/2026 Office action. As a result, these rejections have been withdrawn.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MATTHEW BRYCE GOLAN whose telephone number is (571)272-5159. The examiner can normally be reached Monday through Friday, 8:00 AM to 5:00 PM ET.
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, Alexey Shmatov can be reached at (571) 270-3428. 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.
/MATTHEW BRYCE GOLAN/Examiner, Art Unit 2123
/ALEXEY SHMATOV/Supervisory Patent Examiner, Art Unit 2123
Prosecution Insights