DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
The amendment filed on 12/05/2025 has been entered. Claims 1-4, 6-14, and 16-20 remain pending in the application. Claims 5 and 5 have been canceled. Applicant’s amendments to the claims have overcome each and every 101 rejection set forth in the Non-Final Office Action mailed 09/05/2025.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-4 and 6-10 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 1 recites the limitation "the plurality of machine-learning models" in lines 5-6 of the claim. There is insufficient antecedent basis for this limitation in the claim as "the plurality of machine-learning models" was not introduced prior to its cited use.
Claims 2-4 and 6-10 are rejected based on their dependency to claim 1.
The following is a quotation of 35 U.S.C. 112(d):
(d) REFERENCE IN DEPENDENT FORMS.—Subject to subsection (e), a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
The following is a quotation of pre-AIA 35 U.S.C. 112, fourth paragraph:
Subject to the following paragraph [i.e., the fifth paragraph of pre-AIA 35 U.S.C. 112], a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
Claims 8 and 18 are rejected under 35 U.S.C. 112(d) or pre-AIA 35 U.S.C. 112, 4th paragraph, as being of improper dependent form for failing to further limit the subject matter of the claim upon which it depends, or for failing to include all the limitations of the claim upon which it depends. Dependent claims 8 and 18 stating that the kernels are ran “on different hardware devices” does not further limit the independent claims that state that the kernels are ran “on a plurality of different hardware devices”. These dependent claims fail to add new limitations that further limit the independent claims. Applicant may cancel the claim(s), amend the claim(s) to place the claim(s) in proper dependent form, rewrite the claim(s) in independent form, or present a sufficient showing that the dependent claim(s) complies with the statutory requirements.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-2, 4, 6-10, 11-12, 14, and 16-20 are rejected under 35 U.S.C. 103 as being obvious over Chen et al. (US 20210181739) in view of Suplin et al. (US 20230243662 A1), Dubey et al. (US 20250030766 A1), and Olin et al. (US 20240177528 A1)
Regarding claim 1, Chen teaches a system comprising: a controller ([0049]) configured to:
determine a machine-learning model comprising the kernels among the plurality of machine-learning models based on a task to be performed by a vehicle ([0033]; [0052], models are neural networks, which are comprised of kernels); …
and estimate, using the selected predictor, energy consumption of running the machine-learning model for performing the task on the hardware device of the vehicle ([0055]). …
Chen further teaches that this estimation is done to allocate resources within the vehicle, including power required ([0036] and [0055]). However, Chen does not explicitly teach that the controller is programmed to estimate remaining range of the vehicle based on the estimated energy consumption, and information of the vehicle, and a route of the vehicle.
In the same field of endeavor, reference Suplin teaches a system of using machine-learning models to predict energy consumption of a vehicle, wherein a controller is programmed to:
estimate remaining range of the vehicle based on the estimated energy consumption, and information of the vehicle, and a route of the vehicle ([0037], anticipate range by analyzing the power consumed per segments of a trip; [0030], estimation based on range of the vehicle; [0034-0035], and estimate based off predicted vehicle information and predicted energy consumption);
Suplin additionally teaches that this predicted energy consumption is based on the total energy consumed by the system ([0037]). A skilled artisan would have understood that because these tasks consume power, they must be accounted for in a total energy consumption prediction in order for it to be accurate. Therefore, when this teaching of range estimation is combined with Chen, the estimated range predictably accounts for the energy consumption of running its machine-learning models.
Suplin is analogous to the art of using trained machine-learning models to estimate the power consumption of a vehicle. Therefore, it would have been obvious to modify the teachings of Chen and have the resource and power allocation calculate the range of the vehicle. The motivation for this combination is that it allows the vehicle to know its remaining range, thereby allowing the system to allocate various computing tasks elsewhere, or to plan on stopping at a charging station, when the range is not sufficient for a vehicle to reach its destination.
The process of the prior combination is run a different number of generic processors. However, the prior combination does not teach that this controller is configured to: train a plurality of predictors using a data set including energy consumptions of kernels running on a plurality of different hardware devices, and select one of the plurality of predictors based on a hardware device of the vehicle among the plurality of different hardware devices, each of the plurality of predictors predicting energy consumption of the kernels in corresponding hardware device; and use this selected predictor in the estimation of energy consumption.
In the field of vehicle energy analysis using machine learning, Dubey teaches a system configured to:
train a plurality of predictors using a data set including energy consumptions of kernels running on a plurality of different hardware devices ([0030], [0051], and [0057], where a plurality of neural networks are trained to predict energy consumptions of different vehicles, i.e. different hardware devices);
and select one of the plurality of predictors based on a hardware device of the vehicle among the plurality of different hardware devices, each of the plurality of predictors predicting energy consumption of the kernels in corresponding hardware device ([0057], where additional vehicle-specific layers of a plurality of vehicle-specific layers are included based on the vehicle class being analyzed, such as ).
One of ordinary skill in the art would have been able to modify the prior combination with these teachings, thereby using predictor neural networks corresponding to the type of vehicle being analyzed. As running various machine-learning models, such as the models of Suplin, on a vehicle is well known to consume power, it would have been obvious to one of ordinary skill in the art that these predictor neural networks would also predict the energy consumption of kernels of said machine learning models to ensure a more accurate prediction inclusive of all power consumption in the vehicle.
It would have been obvious to one of ordinary skill in the art at the effective date of filing to modify the prior combination with the vehicle-class specific predictors of Dubey based on a reasonable expectation of success and motivation of better predicting the energy consumption of vehicles based on the traditional consumption of energy for that vehicle class. As one of ordinary skill in the art would recognize that the energy consumption of vehicles varies based on the class of vehicle, having class-specific predictors allows for a better prediction of consumed energy.
The prior combination does not explicitly teach that the controller is further configured to display the estimated remaining range of the vehicle to a driver of the vehicle.
Displaying an estimated range is known in the art as in the same field of endeavor, reference Olin teaches a system that predicts the range of the vehicle, wherein its controller is configured to display the estimated remaining range of the vehicle to a driver of the vehicle ([0055], suggest remaining range).
As Olin is analogous to the art of range estimation for vehicles, it would have been obvious to modify the prior combination to display the remaining range of the vehicle to the driver for the motivation of providing an accurate estimate of range to the driver so they can plan their routes accordingly, and stop at a charging station in the event of the displayed range not being sufficient to get to their destination.
Regarding claim 2, the prior art remains as applied in claim 1. Olin teaches:
wherein the information of the vehicle comprises a vehicle model, a battery remained capacity, a fuel tank remained capacity, velocity of the vehicle, acceleration of the vehicle, or combinations thereof ([0070]; [0043-004]).
Regarding claim 4, the prior art remains as applied in claim 1. Chen teaches that:
the task to be performed by the vehicle includes an automated drive, eye tracking, virtual assistance, mapping systems, driver monitoring, gesture controls, speech recognition, voice recognition, path planning, real-time path monitoring, surrounding object detection, lane changing, or combinations thereof ([0033], task may be “a request for object detection, pedestrian detection, collaborative simultaneous localization and mapping, collaborative perception, path planning, and the like.”).
Regarding claim 6, the prior art remains as applied in claim 1. Olin teaches that its controller is further configured to:
display a refuel or recharge suggestion to the vehicle based on the estimated remaining range of the vehicle. ([0055], suggest a charging point based on charging station availability; [0100], second indication warning that range is insufficient).
Regarding claim 7, the prior art remains as applied in claim 1. Chen teaches that:
the plurality of machine-learning models include a deep neural network, a convolutional neural network, and a recurrent neural network ([0052]).
Regarding claim 8, the prior art remains as applied in claim 1. Dubey teaches that the controller is further configured to:
train the plurality of predictors using a data set including energy consumptions of the kernels running on different hardware devices ([0030], [0051], and [0057], where a plurality of neural networks are trained to predict energy consumptions of different vehicles, i.e. different hardware devices);
Regarding claim 9, the prior art remains as applied in claim 8. Chen teaches that:
the data set including energy consumptions of the kernels running on different hardware devices are obtained by monitoring a power of the vehicle while running the kernels on the different hardware devices ([0070], training data set is obtained and updated by analyzing “historical data of particular computing tasks and the amount of computing resources that were needed to perform those computing tasks.” According to the prior combination, this includes analyzing the computing task on a corresponding processor to predict computing resources, i.e. power of the hardware device(s) of the vehicle).
Regarding claim 10, the prior art remains as applied in claim 1. Olin teaches that the controller configured to:
update the route of the vehicle based on the estimated remaining range of the vehicle ([0046], display charging stations to driver when range of vehicle is too low, and operate the vehicle to reach those charging stations)
Regarding claim 11, Chen teaches a method for estimating remaining range of a vehicle comprising:
determining a machine-learning model comprising the kernels among a plurality of machine-learning models based on a task to be performed by a vehicle ([0033]; [0052], models are neural networks, which are comprised of kernels); …
and estimating, using the selected predictor, energy consumption of running the machine-learning model for performing the task on the hardware device of the vehicle ([0055]). …
Chen further teaches that this estimation is done to allocate resources within the vehicle, including power required ([0036] and [0055]). However, Chen does not explicitly teach that the method includes estimating the remaining range of the vehicle based on the estimated energy consumption, and information of the vehicle, and a route of the vehicle.
In the same field of endeavor, reference Suplin teaches a system of using machine-learning models to predict energy consumption of a vehicle, wherein a controller is programmed to:
estimating the remaining range of the vehicle based on the estimated energy consumption, and information of the vehicle, and a route of the vehicle ([0037], anticipate range by analyzing the power consumed per segments of a trip; [0030], estimation based on range of the vehicle; [0034-0035], and estimate based off predicted vehicle information and predicted energy consumption);
Suplin additionally teaches that this predicted energy consumption is based on the total energy consumed by the system ([0037]). A skilled artisan would have understood that because these tasks consume power, they must be accounted for in a total energy consumption prediction in order for it to be accurate. Therefore, when this teaching of range estimation is combined with Chen, the estimated range predictably accounts for the energy consumption of running its machine-learning models.
Suplin is analogous to the art of using trained machine-learning models to estimate the power consumption of a vehicle. Therefore, it would have been obvious to modify the teachings of Chen and have the resource and power allocation calculate the range of the vehicle. The motivation for this combination is that it allows the vehicle to know its remaining range, thereby allowing the method to allocate various computing tasks elsewhere, or to plan on stopping at a charging station, when the range is not sufficient for a vehicle to reach its destination.
The process of the prior combination is run a different number of generic processors. However, the prior combination does not teach that method further includes: training a plurality of predictors using a data set including energy consumptions of kernels running on a plurality of different hardware devices, and selecting one of the plurality of predictors based on a hardware device of the vehicle among the plurality of different hardware devices, each of the plurality of predictors predicting energy consumption of the kernels in corresponding hardware device; and use this selected predictor in the estimation of energy consumption.
In the field of vehicle energy analysis using machine learning, Dubey teaches a method comprising:
training a plurality of predictors using a data set including energy consumptions of kernels running on a plurality of different hardware devices ([0030], [0051], and [0057], where a plurality of neural networks are trained to predict energy consumptions of different vehicles, i.e. different hardware devices);
and selecting one of the plurality of predictors based on a hardware device of the vehicle among the plurality of different hardware devices, each of the plurality of predictors predicting energy consumption of the kernels in corresponding hardware device ([0057], where additional vehicle-specific layers of a plurality of vehicle-specific layers are included based on the vehicle class being analyzed, such as ).
One of ordinary skill in the art would have been able to modify the prior combination with these teachings, thereby using predictor neural networks corresponding to the type of vehicle being analyzed. As running various machine-learning models, such as the models of Suplin, on a vehicle is well known to consume power, it would have been obvious to one of ordinary skill in the art that these predictor neural networks would also predict the energy consumption of kernels of said machine learning models to ensure a more accurate prediction inclusive of all power consumption in the vehicle.
It would have been obvious to one of ordinary skill in the art at the effective date of filing to modify the prior combination with the vehicle-class specific predictors of Dubey based on a reasonable expectation of success and motivation of better predicting the energy consumption of vehicles based on the traditional consumption of energy for that vehicle class. As one of ordinary skill in the art would recognize that the energy consumption of vehicles varies based on the class of vehicle, having class-specific predictors allows for a better prediction of consumed energy.
The prior combination does not explicitly teach that the method includes displaying the estimated remaining range of the vehicle to a driver of the vehicle.
Displaying an estimated range is known in the art as in the same field of endeavor, reference Olin teaches a method that predicts the range of the vehicle, wherein the method includes displaying the estimated remaining range of the vehicle to a driver of the vehicle ([0055], suggest remaining range).
As Olin is analogous to the art of range estimation for vehicles, it would have been obvious to modify the prior combination to display the remaining range of the vehicle to the driver for the motivation of providing an accurate estimate of range to the driver so they can plan their routes accordingly, and stop at a charging station in the event of the displayed range not being sufficient to get to their destination.
Regarding claim 12, the prior art remains as applied in claim 11. Olin teaches:
wherein the information of the vehicle comprises a vehicle model, a battery remained capacity, a fuel tank remained capacity, velocity of the vehicle, acceleration of the vehicle, or combinations thereof ([0070]; [0043-004]).
Regarding claim 14, the prior art remains as applied in claim 11. Chen teaches that:
the task to be performed by the vehicle includes an automated drive, eye tracking, virtual assistance, mapping systems, driver monitoring, gesture controls, speech recognition, voice recognition, path planning, real-time path monitoring, surrounding object detection, lane changing, or combinations thereof ([0033], task may be “a request for object detection, pedestrian detection, collaborative simultaneous localization and mapping, collaborative perception, path planning, and the like.”).
Regarding claim 16, the prior art remains as applied in claim 11. Olin teaches that the method is further comprising:
displaying a refuel or recharge suggestion to the vehicle based on the estimated remaining range of the vehicle. ([0055], suggest a charging point based on charging station availability; [0100], second indication warning that range is insufficient).
Regarding claim 17, the prior art remains as applied in claim 11. Chen teaches that:
the plurality of machine-learning models include a deep neural network, a convolutional neural network, and a recurrent neural network ([0052]).
Regarding claim 18, the prior art remains as applied in claim 11. Dubey teaches that the method is further comprising:
training the plurality of predictors using a data set including energy consumptions of the kernels running on different hardware devices ([0030], [0051], and [0057], where a plurality of neural networks are trained to predict energy consumptions of different vehicles, i.e. different hardware devices);
Regarding claim 19, the prior art remains as applied in claim 18. Chen teaches that:
the data set including energy consumptions of the kernels running on different hardware devices are obtained by monitoring a power of the vehicle while running the kernels on the different hardware devices ([0070], training data set is obtained and updated by analyzing “historical data of particular computing tasks and the amount of computing resources that were needed to perform those computing tasks.” According to the prior combination, this includes analyzing the computing task on a corresponding processor to predict computing resources, i.e. power of the hardware device(s) of the vehicle).
Regarding claim 20, the prior art remains as applied in claim 11. Olin teaches that the method further comprises:
updating the route of the vehicle based on the estimated remaining range of the vehicle ([0046], display charging stations to driver when range of vehicle is too low, and operate the vehicle to reach those charging stations)
Claims 3 and 13 are rejected under 35 U.S.C. 103 as being obvious over Chen in view of Suplin, Dubey, and Olin as applied to claims 1 and 11 above, and in further view of Brittain et al. (US 20220335287 A1), referred to further as Brittain.
Regarding claim 3, the prior art remains as applied in claim 1. Dubey further teaches that:
the selected predictor estimates energy consumption of running a composite operation on the hardware device of the vehicle and energy consumption of running remaining kernels on the hardware device of the vehicle ([0057], where the total energy of the composite operations that are apart of driving the vehicle are estimated) Note that the energy consumption of kernels in Dubey is for all kernels used in the performance of a program, which is a neural network according to the prior combination.
The prior combination does not teach that the controller is further configured to fuse two or more of the kernels into a composite operation. However, in the same field of endeavor, reference Brittain does teach a process of improving a neural network’s performance, wherein a controller is configured to fuse two or more of the kernels into a composite operation ([0006], “kernel fusing”). Brittain further teaches that these kernels are fused so that similar kernels are fused based on the proximity of their memory bound operations ([0004]).
As is understood in the art, different nodes or kernels that used by a program, such as a neural network, often access the same or similar memory as other nodes or kernels that are used by said program. As such, when the prior combination incorporates the kernel fusion technique of Brittain, a skilled artisan would have understood that the kernels associated with a specific model will be fused for a composite operation, with the program now being ran on the fused kernels. This predictably results in the energy consumption of the fused kernels of a program being considered, with the other kernels associated with said program being considered subsequently.
As Brittain is analogous to the art of analyzing and increasing efficiency and accuracy of kernel metrics, it would have been obvious to modify the prior combination by fusing kernels of a neural network for the motivation, as taught by Brittain, of making overall execution time faster and consuming less power ([0004]).
Regarding claim 13, the prior art remains as applied in claim 11. Dubey further teaches that:
the selected predictor estimates energy consumption of running a composite operation on the hardware device of the vehicle and energy consumption of running remaining kernels on the hardware device of the vehicle ([0057], where the total energy of the composite operations that are apart of driving the vehicle are estimated) Note that the energy consumption of kernels in Dubey is for all kernels used in the performance of a program, which is a neural network according to the prior combination.
The prior combination does not teach that method is further comprising fusing two or more of the kernels into a composite operation. However, in the same field of endeavor, reference Brittain does teach a method of improving a neural network’s performance, wherein the method includes fusing two or more of the kernels into a composite operation ([0006], “kernel fusing”). Brittain further teaches that these kernels are fused so that similar kernels are fused based on the proximity of their memory bound operations ([0004]).
As is understood in the art, different nodes or kernels that used by a program, such as a neural network, often access the same or similar memory as other nodes or kernels that are used by said program. As such, when the prior combination incorporates the kernel fusion technique of Brittain, a skilled artisan would have understood that the kernels associated with a specific model will be fused for a composite operation, with the program now being ran on the fused kernels. This predictably results in the energy consumption of the fused kernels of a program being considered, with the other kernels associated with said program being considered subsequently.
As Brittain is analogous to the art of analyzing and increasing efficiency and accuracy of kernel metrics, it would have been obvious to modify the prior combination by fusing kernels of a neural network for the motivation, as taught by Brittain, of making overall execution time faster and consuming less power ([0004]).
Response to Arguments
Applicant's arguments filed 12/05/2025 have been fully considered.
Applicant argues over the amended independent claim 1, contesting that “The cited references do not teach or suggest a controller configured to ‘select one of the plurality of predictors based on a hardware device of the vehicle among the plurality of different hardware devices, each of the plurality of predictors predicting energy consumption of the kernels in corresponding hardware device””. This argument is persuasive as reference Zou did not teach the predictor being selected “based on a hardware device of the vehicle among the plurality of different hardware devices” as the claim is amended to include. As a result, the previous rejection in view of Zou has been withdrawn.
However, referring further to the language of claim 1, examiner recognizes that “the plurality of different hardware devices” are not explicitly defined to be comprised in a singular vehicle. As a result, under a broadest reasonable interpretation, reference Dubey is now relied upon to teach the aforementioned amended limitations, and limitations previously taught by Zou, as it teaches that a plurality of predictors are trained and selected based on a hardware device of the vehicle, i.e. vehicle type or class, among the plurality of different hardware devices, i.e. among the plurality of vehicle types or classes.
Therefore, a new rejection is made under 35 U.S.C. 103 over Chen in view of Suplin, Dubey, and Olin as necessitated by applicant’s amendments to the independent claims.
Conclusion
The following prior art made of record and not relied upon is considered pertinent to applicant’s disclosure:
Wu et al. (US 20220363140 A1)
Kim et al. (US 20250068938 A1)
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 JACK R. BREWER whose telephone number is (571)272-4455. The examiner can normally be reached 9AM-6PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Angela Ortiz can be reached at 571-272-1206. 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.
JACK R. BREWER
Examiner
Art Unit 3663
/ADAM D TISSOT/Primary Examiner, Art Unit 3663