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 Objections
Claim 9 objected to because of the following informality. In the preamble, Examiner suggests a word or phrase is missing after the word “neural” and in further examination below the claim is being read as “neural networks.” Appropriate correction is required.
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 1–15 rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding claims 1–2:
These claims recite “[a] system for forecasting energy consumption by vehicles in a mixed-vehicle fleet, the system comprising”
“non-transitory computer readable storage media that stores” a collection of data
“a network interface that receives” a collection of data
“a multi-task learning model comprising” a collection of components
“an inductive transfer learning model comprising” a collection of components
Examiner finds that no structure is recited for performing the functions of the “multi-task learning model” or “inductive transfer learning model.” Whether the models are intended to be hardware models, software, or neither (paper models) is unclear; further, because of this confusion, whether the “system” of the claim is intended to be machine, a manufacture, or a process is unclear. Thus, a person having ordinary skill in the art would be unable to determine the metes and bounds of the claim.
Regarding claims 3-15:
These claims recite a system comprising media storing a collection of data and “one or more neural networks” but no structure is recited for performing the functions of the networks. Whether the networks are intended to be a hardware model, software, or neither is unclear; further, because of this confusion, whether the “system” of the claim is intended to be machine, a manufacture, or a process is unclear. Thus, a person having ordinary skill in the art would be unable to determine the metes and bounds of the claim.
Further, claim 4 recites a “multi-task learning model” and claim 5 recites “an inductive transfer learning model,” again without structure for the performance of the model functions, giving rise to the arguments described under claims 1–2, above.
Examiner’s note: Examiner here notes that claims 16–20 are to a method, and are not found to be indefinite under 112(b). However, Examiner suggests considering whether the claims should recite a computer-implemented method and a structure for performance of the method steps.
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 rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Analysis is provided for the claims under the guidelines of MPEP 2106.
Regarding claim 1:
Step 1:
The claim recites “[a] system for forecasting energy consumption by vehicles in a mixed-vehicle fleet, the system comprising” the components that follow. Because of the ambiguity described in the 112(b) rejection above, it is unclear whether the claim is to a machine, a manufacture, or a process, however, the claim is to one of these statutory categories.
Step 2A prong 1:
The limitation (bold only) “a multi-task learning model comprising: a plurality of shared layers that identify features indicative of energy consumption and a marginal probability distribution over each of the identified features” recites a mental process. A person could identify features in the input data indicative of energy consumption with a marginal probability distribution over each of the identified features, using observation and judgement.
The limitation (bold only) “a set of vehicle-specific layers for each of a plurality of classes of vehicles, each set of vehicle-specific layers identifying a conditional probability distribution over each of the identified features to generate a predictive function for predicting energy consumption by vehicles in the class” recites a mental process. A person could identify a conditional probability distribution over each of the identified features and use this probability as predictive function for predicting energy consumption by vehicles, using judgement and evaluation.
The combination of the limitations (bold only) “a multi-task learning model comprising: a plurality of shared layers that identify features indicative of energy consumption and a marginal probability distribution over each of the identified features” and (bold only) “a set of vehicle-specific layers for each of a plurality of classes of vehicles, each set of vehicle-specific layers identifying a conditional probability distribution over each of the identified features to generate a predictive function for predicting energy consumption by vehicles in the class” recites a mental process. The bolded portions indicate that the feature identification step is performed for all vehicle classes as a whole, while the predictive energy consumption function is performed for a specific class of vehicles, using the features produced by the all-vehicle-class feature selection process. A person could perform the mental processes, recited above, in this manner. Therefore, the entire “multi-task learning model” recites a mental process.
The limitation (bold only) “an inductive transfer learning model comprising: the plurality of shared layers transferred from the multi-task learning model; and at least one set of vehicle-specific layers for an additional class of vehicles that identifies a conditional probability distribution over each of the identified features to generate a predictive function for predicting energy consumption by vehicles in the additional class” recites a mental process. The limitation recites using the all-vehicle-class feature identification mental process to produce an energy consumption predictive function for an additional vehicle class. A person could perform this process using judgement and evaluation.
Thus, the claim recites an abstract idea.
Step 2A prong 2:
The further elements “non-transitory computer readable storage media that stores: route data identifying vehicle routes in a geographic area, each vehicle route comprising a plurality of route segments; and elevation data indicative of elevations along each route segment” and “a network interface that receives: traffic data indicative of traffic conditions along at least some of the route segments; weather data indicative of weather conditions in the geographic area; vehicle locations indicative of locations of each of the vehicles; and energy consumption data indicative of energy consumed by each of the vehicles” recite the input of the vehicle data used in the recited mental processes identified above. Thus, the further elements recite mere data gathering, which is insignificant extra-solution activity (MPEP 2106.05(g)).
Thus, the additional elements merely recite insignificant extra-solution activity. Taken alone, the additional elements do not integrate the abstract idea into a practical application. Considering the elements together as an ordered combination adds nothing that is not present from examining the elements individually. The elements, individually or together, do not describe an improvement in the functioning of technology.
Step 2B:
The claim as a whole does not amount to significantly more than the recited judicial exception.
The elements “non-transitory computer readable storage media that stores: route data identifying vehicle routes in a geographic area, each vehicle route comprising a plurality of route segments; and elevation data indicative of elevations along each route segment” and “a network interface that receives: traffic data indicative of traffic conditions along at least some of the route segments; weather data indicative of weather conditions in the geographic area; vehicle locations indicative of locations of each of the vehicles; and energy consumption data indicative of energy consumed by each of the vehicles” recite mere data gathering, which is recognized as well-understood, routine, and conventional activity in the art (see MPEP § 2106.05(d)(II)(i)).
Even when considered in combination, the additional elements represent insignificant extra-solution activity, which do not provide an inventive concept. The claim is not eligible under 35 U.S.C. 101.
Regarding claim 2:
For step 2A prong 1, claim 2 further limits claim 1 and the same elements in claim 2 still recite an abstract idea.
The further element “wherein the features indicative of energy consumption comprise: length of each route; average past travel speed along each route; past time to travel along each route; change in elevation along each route; maximum elevation change along each route; speed ratio along each of at least some of the routes; jam factor along each of at least some of the routes; temperature in the geographic area; precipitation in the geographic area; visibility in the geographic area; wind speed in the geographic area; humidity in the geographic area; and wind gust in the geographic area” further limit the data used by the mental processes identified under claim 1, but they remain mental processes.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claim 3:
Step 1:
The claim recites “[a] system for forecasting energy consumption by vehicles in a mixed-vehicle fleet, the system comprising” the components that follow. Because of the ambiguity described in the 112(b) rejection above, it is unclear whether the claim is to a machine, a manufacture, or a process, however, the claim is to one of these statutory categories.
Step 2A prong 1:
The limitation “and one or more neural networks that generates a predictive function for predicting energy consumption by vehicles in each of a plurality of classes of vehicles, the one or more neural networks comprising a plurality of shared layers for all of the classes of vehicles and a set of vehicle-specific layers for each of the plurality of classes” recites a mental process. A person could generate a function for predicting energy consumptions by a vehicle class using observation and judgement. Further, the person could perform this process using both process steps that are shared among all vehicle classes and that are specific to a vehicle class.
Thus, the claim recites an abstract idea.
Step 2A prong 2:
The element “non-transitory computer readable storage media that stores route data identifying vehicle routes in a geographic area, each vehicle route comprising a plurality of route segments; a network interface that receives: vehicle locations indicative of locations of each of the vehicles; and energy consumption data indicative of energy consumed by each of the vehicles” recites the input of the vehicle data used in the recited mental process identified above. Thus, the element recites mere data gathering, which is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B:
The claim as a whole does not amount to significantly more than the recited judicial exception.
The element “non-transitory computer readable storage media that stores route data identifying vehicle routes in a geographic area, each vehicle route comprising a plurality of route segments; a network interface that receives: vehicle locations indicative of locations of each of the vehicles; and energy consumption data indicative of energy consumed by each of the vehicles” recites mere data gathering, which is recognized as well-understood, routine, and conventional activity in the art (see MPEP § 2106.05(d)(II)(i)).
Even when considered in combination, the additional elements represent insignificant extra-solution activity, which do not provide an inventive concept. The claim is not eligible under 35 U.S.C. 101.
Regarding claim 4:
For step 2A prong 1, claim 4 further limits claim 3 and the same elements in claim 4 still recite an abstract idea.
The further limitation “wherein the one or more neural networks comprise a multi-task learning model comprising the plurality of shared layers for all of the classes of vehicles and the sets of vehicle-specific layers for each of the plurality of classes” further limits the mental process of the claim but it remains a mental process. A person could perform the process using steps shared for all of the classes of vehicles and other steps that are vehicle-specific.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claim 5:
For step 2A prong 1, claim 5 further limits claim 4 and the same elements in claim 5 still recite an abstract idea.
The further limitation “wherein the one or more neural networks further comprise an inductive transfer learning model comprising the plurality of shared layers, transferred from the multi-task learning model, and at least one set of vehicle-specific layers for at least one additional class” further limits the mental process of the claim but it remains a mental process. A person could perform extend the process to an additional vehicle class in the manner described.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claim 6:
For step 2A prong 1, claim 6 further limits claim 3 and the same elements in claim 6 still recite an abstract idea.
The further limitation “the plurality of shared layers identifies features indicative of energy consumption and a marginal probability distribution over each of the identified features for all of the classes of vehicles; and for each class, the set of vehicle-specific layers identifies a conditional probability distribution over each of the identified features” further limits the mental process of the claim but it remains a mental process. A person could identify features indicative of energy consumption and a marginal probability distribution over each of the identified features for all of the classes of vehicles, using observation and judgement. Further, a person could, per vehicle class, identify a conditional probability distribution, using judgement and evaluation.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claim 7:
For step 2A prong 1, claim 7 further limits claim 3 and the same elements in claim 7 still recite an abstract idea.
The further limitations “the plurality of shared layers identifies features indicative of energy consumption for all of the classes of vehicles; and for each class, the set of vehicle-specific layers identifies a marginal probability distribution over each of the identified features and a conditional probability distribution over each of the identified features” further limits the mental process of the claim but it remains a mental process. A person could identify features indicative of energy consumption for all of the classes of vehicles, using observation and judgement. Further, a person could, per vehicle class, identify a marginal probability distribution over each of the identified features and a conditional probability distribution over each of the identified features, using judgement and evaluation.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claim 8:
For step 2A prong 1, claim 8 further limits claim 3 and the same elements in claim 8 still recite an abstract idea.
For step 2A prong 2, the further element “the non-transitory computer readable storage media further stores elevation data indicative of elevations along each route segment; and a network interface further receives: traffic data indicative of traffic conditions along at least some of the route segments; and weather data indicative of weather conditions in the geographic area” recites the input of the vehicle data used in the recited mental process identified above. Thus, the element recites mere data gathering, which is insignificant extra-solution activity (MPEP 2106.05(g)).
For step 2B, the claim as a whole does not amount to significantly more than the recited judicial exception.
The element “the non-transitory computer readable storage media further stores elevation data indicative of elevations along each route segment; and a network interface further receives: traffic data indicative of traffic conditions along at least some of the route segments; and weather data indicative of weather conditions in the geographic area” recites mere data gathering, which is recognized as well-understood, routine, and conventional activity in the art (see MPEP § 2106.05(d)(II)(i)).
Even when considered in combination, the additional elements represent insignificant extra-solution activity, which do not provide an inventive concept. The claim is not eligible under 35 U.S.C. 101.
Regarding claim 9:
For step 2A prong 1, claim 9 further limits claim 8 and the same elements in claim 9 still recite an abstract idea.
The further element “wherein the one or more neural generates each predictive function by identifying features indicative of energy consumption along each route segment” further limits the mental process of the claim but it remains a mental process. A person could identify features predictive of energy consumption along a route segment, using observation and judgement.
For step 2A prong 2, the further element “the features indicative of energy consumption along each route segment including: one or more static road segment features for each route segment, identified using the route data and the elevation data; one or more past vehicle speed features for each route segment, identified using the vehicle locations; one or more weather conditions, identified using the weather data; and one or more traffic conditions, identified using the traffic data” recites the input of the vehicle data used in the recited mental process identified above. Thus, the element recites mere data gathering, which is insignificant extra-solution activity (MPEP 2106.05(g)).
For step 2B, the claim as a whole does not amount to significantly more than the recited judicial exception.
The element “the features indicative of energy consumption along each route segment including: one or more static road segment features for each route segment, identified using the route data and the elevation data; one or more past vehicle speed features for each route segment, identified using the vehicle locations; one or more weather conditions, identified using the weather data; and one or more traffic conditions, identified using the traffic data” recites mere data gathering, which is recognized as well-understood, routine, and conventional activity in the art (see MPEP § 2106.05(d)(II)(i)).
Even when considered in combination, the additional elements represent insignificant extra-solution activity, which do not provide an inventive concept. The claim is not eligible under 35 U.S.C. 101.
Regarding claim 10:
For step 2A prong 1, claim 10 further limits claim 3 and the same elements in claim 10 still recite an abstract idea.
The further element “one of the plurality of classes is internal combustion vehicles and one of the plurality of classes is electric vehicles; and the energy consumption data is indicative of fuel consumed by the internal combustion vehicles and electric energy consumed and generated by the electric vehicles” further limits the data used by the mental process, but it remains a mental process.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claim 11:
For step 2A prong 1, claim 11 further limits claim 10 and the same elements in claim 11 still recite an abstract idea.
The further element “one of the plurality of classes is hybrid vehicles” further limits the data used by the mental process, but it remains a mental process.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claim 12:
For step 2A prong 1, claim 12 further limits claim 3 and the same elements in claim 12 still recite an abstract idea.
The further element “wherein the one or more neural networks comprise a set of vehicle-specific layers for each model of vehicle in the mixed-vehicle fleet” further limits the mental process, but it remains a mental process. A person could perform a set of steps specific to a model of vehicle to perform the process.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claim 13:
For step 2A prong 1, claim 13 further limits claim 3 and the same elements in claim 13 still recite an abstract idea.
The further element “wherein the one or more neural networks comprise a set of vehicle-specific layers for each year of each model in the mixed-vehicle fleet” further limits the mental process, but it remains a mental process. A person could perform a set of steps specific to a model year of vehicle to perform the process.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claim 14:
For step 2A prong 1, claim 14 further limits claim 3 and the same elements in claim 14 still recite an abstract idea.
The further element “a vehicle trajectory mapping module that maps each vehicle location to one of the route segments in the route data” recites a mental process. A person could map vehicle locations to route segments using observation and judgement.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claim 15:
For step 2A prong 1, claim 15 further limits claim 3 and the same elements in claim 15 still recite an abstract idea.
The further element “wherein the vehicles are buses” further limits the data used by the mental process, but it remains a mental process.
For step 2A prong 2, and step 2B, no further elements remain to be considered. The claim as a whole does not amount to significantly more than the recited judicial exception and is ineligible under 35 U.S.C. 101.
Regarding claims 16–20:
These claims are to “[a] method of forecasting energy consumption by vehicles in a mixed-vehicle fleet, the method comprising” the steps that follow. Thus, the claims are to a process, which is a statutory category of invention. The claims are otherwise analogous to claims 3, 5–6, or 8–9, respectively, and are found ineligible by the same arguments.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 3–11, 14, and 16–20 rejected under 35 U.S.C. 103 over Meyer et al., US Pre-Grant Publication No. 2016/0061611 (hereafter Meyer) in view of Lee, US Pre-Grant Publication No. 2022/0121883 (hereafter Lee).
Regarding claim 1:
Meyer teaches:
“A system for forecasting energy consumption by vehicles in a mixed-vehicle fleet, the system comprising: non-transitory computer readable storage media that stores”: Meyer, paragraph 0077, “The computing system 500 [system] can also include a disk drive unit 521 for receiving a computer readable medium 528 [non-transitory computer readable storage media]. In a particular embodiment, the disk drive unit 521 may receive the computer-readable medium 528 in which one or more sets of instructions 527, such as the software corresponding to the energy prediction tool, can be embedded. Further, the instructions 527 may embody one or more of the methods or logic as described herein. In a particular embodiment, the instructions 527 may reside completely, or at least partially, within any one or more of the main memory 512, the static memory 522, computer readable medium 528, and/or within the processor 511 during execution of the instructions 527 by the processor 511”; Meyer, paragraph 0020, “For traditional petroleum based vehicles, the energy consumption prediction may be generated by the energy prediction tool in terms of an amount of petroleum fuel ( e.g., gasoline, diesel fuel) predicted to be consumed in gallons, liters or other amount of measurable fuel usage, and/or in terms of an energy usage amount ( e.g., kWh, Joules, or other similar unit of energy usage) by one or more vehicle batteries that are included in the vehicle system. For vehicles that rely, at least in part, on one or more batteries for powering the propulsion of the vehicle, the energy consumption prediction may be generated by the energy prediction tool in terms of amount of battery energy predicted to be consumed in term of an energy usage amount ( e.g., kWh, Joules, or other similar unit of energy usage) by one or more vehicle batteries that are included in the vehicle system [hence, a system for estimating a mixed-vehicle fleet, where traditional combustion engine vehicles and electric vehicles are two different vehicle classes].
“route data identifying vehicle routes in a geographic area, each vehicle route comprising a plurality of route segments”: Meyer, paragraph 0006, “The apparatus may include a memory configured to store road segment information for a road segment a vehicle is set to travel on [route data identifying vehicle routes in a geographic area, each vehicle route comprising a plurality of route segments], and a processor in communication with the memory.”
“elevation data indicative of elevations along each route segment”: Meyer, paragraph 0018, “The energy consumption tool may incorporate road segment information ( e.g., posted speed limit, elevation [elevation data indicative of elevations along each route segment], number of traffic stops) into an algorithm for generating the energy consumption profile for the vehicle.”
“a network interface that receives: traffic data indicative of traffic conditions along at least some of the route segments; weather data indicative of weather conditions in the geographic area”: Meyer, paragraph 0076, “In a networked deployment, the computing system 500 may operate in the capacity of a server or as a client user computer within a vehicle in a server-client user network environment, or as a peer computer system within a vehicle in a peer-to-peer ( or distributed) network environment [a network interface]”; Meyer, paragraph 0027, “The road segment information may include, but is not limited to, posted speed limit on the identified road segment, an elevation profile for the identified road segment, current and/or predicted traffic information for the identified road segment [traffic data indicative of traffic conditions along at least some of the route segments], road condition information for the identified road segment, weather information for the identified road segment [weather data indicative of weather conditions in the geographic area], stop sign and traffic light information for the identified road segment, tum sequence information for the identified road segment, or some other identifiable road segment attribute for the identified road segment.”
“vehicle locations indicative of locations of each of the vehicles”: Meyer, paragraph 0023, “In some embodiments the starting location 110 may correspond to a current location of the vehicle 101 obtained by the energy prediction tool via driver input, or via location information received from a GPS unit that is part of the vehicle's vehicle system [vehicle locations indicative of locations of each of the vehicles]. In some embodiments the starting location 110 may correspond to a location manually input by the driver that may not correspond to an actual current location of the vehicle.”
“energy consumption data indicative of energy consumed by each of the vehicles”: Meyer, paragraph 0016, “However, even under such embodiments where the historical driving information may not be relied upon, the energy prediction tool may provide an accurate energy consumption prediction for the vehicle based on historical energy consumption information for the vehicle [energy consumption data indicative of energy consumed by each of the vehicles], external information that may affect energy consumption by the vehicle, vehicle systems information that may affect energy consumption by the vehicle, and/or selected road segment information.”
“generate a predictive function for predicting energy consumption by vehicles in the additional class,” (bold only) “a plurality of shared layers that identify features indicative of energy consumption and a marginal probability distribution over each of the identified features” (bold only) “a set of vehicle-specific layers for each of a plurality of classes of vehicles, each set of vehicle-specific layers identifying a conditional probability distribution over each of the identified features to generate a predictive function for predicting energy consumption by vehicles in the class,” and (bold only) “and at least one set of vehicle-specific layers for an additional class of vehicles that identifies a conditional probability distribution over each of the identified features to generate a predictive function for predicting energy consumption by vehicles in the additional class”: Meyer, paragraph 0058, “After generating the energy consumption predictions from one or more of the wheel energy model 204, elevation model 205, warm up model 206, braking/accelerating model 207, auxiliary load model 208, and climate usage 209, a summation function may be implemented by the energy prediction tool at 210 and 211 [a predictive function for predicting energy consumption by vehicles in the additional class][indicative of energy consumption]. For example, at 210 the energy prediction tool may implement the summation of the energy consumption predictions from the wheel energy model 204, elevation model 205, warm up model 206, and braking/accelerating model 207 to generate a propulsive energy consumption prediction.”
Meyer does not explicitly teach:
“a multi-task learning model comprising”
(bold only) “a plurality of shared layers that identify features indicative of energy consumption and a marginal probability distribution over each of the identified features”
(bold only) “a set of vehicle-specific layers for each of a plurality of classes of vehicles, each set of vehicle-specific layers identifying a conditional probability distribution over each of the identified features to generate a predictive function for predicting energy consumption by vehicles in the class”
“and an inductive transfer learning model comprising”
“the plurality of shared layers transferred from the multi-task learning model”
(bold only) “and at least one set of vehicle-specific layers for an additional class of vehicles that identifies a conditional probability distribution over each of the identified features to generate a predictive function for predicting energy consumption by vehicles in the additional class”
Lee teaches:
“a multi-task learning model comprising”: Lee, paragraph 0014, “An apparatus for training an image classification model according to an embodiment disclosed includes a first trainer that trains a model body and a first head through supervised learning based on a labeled data set subjected to type 1 labeling, a second trainer that trains the model body, the first head, and a second head through multi-task learning [a multi-task learning model] based on the labeled data set and an unlabeled data set, and a third trainer that trains a plurality of third heads through supervised learning based on the labeled data set subjected to type 2 labeling while freezing the model body, in which the model body extracts feature vector for input data, and each of the first head, the second head, and the third head generates a classification result based on the feature vector.”
(bold only) “a plurality of shared layers that identify features indicative of energy consumption and a marginal probability distribution over each of the identified features”: Lee, Fig. 4,
PNG
media_image1.png
491
806
media_image1.png
Greyscale
; Lee, paragraphs 0095–0096, “Thereafter, the model body 220 generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420 [a plurality of shared layers that identify features]. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector”; Lee, paragraph 0065, “In addition, the second trainer 120 may train the model body based on the total loss function obtained by performing weighted summation of each of the loss functions calculated by supervised learning, self-supervised learning, and unsupervised learning [interpreting the last layer of a weighted classification network as a probability distribution; and such a network with multiple classification heads as a marginal probability distribution over each of the identified features].”
(bold only) “a set of vehicle-specific layers for each of a plurality of classes of vehicles, each set of vehicle-specific layers identifying a conditional probability distribution over each of the identified features to generate a predictive function for predicting energy consumption by vehicles in the class”: Lee, paragraphs 0095–0096, “Thereafter, the model body 220 generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420 [a set of …-specific layers for each of a plurality of classes]. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector [interpreting the last layer of a weighted classification network as a conditional probability distribution, hence, identifying a conditional probability distribution over each of the identified features].”
“and an inductive transfer learning model comprising”: Lee, paragraph 0043–0045, “Thereafter, the first trainer 110 may update a training parameter of each of the model body and the first head based on the calculated loss function value. In this case, the training parameter may be, for example, a weight or a bias applied to a layer included in the network structure of each of the model body or the first head. Meanwhile, according to an embodiment, the first trainer 110 may calculate the loss function value described above by using a cross entropy function as a loss function. The second trainer 120 trains the model body, the first head, and a second head through multi-task learning based on the labeled data set subjected to type 1 labeling and an unlabeled data set [an inductive transfer learning model].”
“the plurality of shared layers transferred from the multi-task learning model”: Lee, paragraph 0043–0045, “Thereafter, the first trainer 110 may update a training parameter of each of the model body and the first head based on the calculated loss function value. In this case, the training parameter may be, for example, a weight or a bias applied to a layer included in the network structure of each of the model body or the first head. Meanwhile, according to an embodiment, the first trainer 110 may calculate the loss function value described above by using a cross entropy function as a loss function. The second trainer 120 trains the model body [the plurality of shared layers transferred from the multi-task learning model], the first head, and a second head through multi-task learning based on the labeled data set subjected to type 1 labeling and an unlabeled data set.”
(bold only) “and at least one set of vehicle-specific layers for an additional class of vehicles that identifies a conditional probability distribution over each of the identified features to generate a predictive function for predicting energy consumption by vehicles in the additional class”: Lee, paragraphs 0095–0096, “Thereafter, the model body 220 generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420 [at least one set of …-specific layers for an additional class]. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector [that identifies a conditional probability distribution over each of the identified features].”
Lee and Meyer are analogous arts as they are both related to data modelling. It would have been obvious to a person having ordinary skill in the art prior to the effective filing date of the claimed invention to have combined the multi-task modelling of Lee with the teachings of Meyer to arrive at the present invention, in order to improve the model and its training, as stated in Lee, paragraph 0023, “In addition, according to the disclosed embodiments, it is possible to reduce the amount of time and resources required for training by making lighten the overall structure of the image classification model by constructing several model heads in parallel while sharing a model body.”
Regarding claim 3 and analogous claim 16:
Meyer teaches:
“A system for forecasting energy consumption by vehicles in a mixed-vehicle fleet, the system comprising: non-transitory computer readable storage media”: Meyer, paragraph 0077, “The computing system 500 [system] can also include a disk drive unit 521 for receiving a computer readable medium 528 [non-transitory computer readable storage media]. In a particular embodiment, the disk drive unit 521 may receive the computer-readable medium 528 in which one or more sets of instructions 527, such as the software corresponding to the energy prediction tool, can be embedded. Further, the instructions 527 may embody one or more of the methods or logic as described herein. In a particular embodiment, the instructions 527 may reside completely, or at least partially, within any one or more of the main memory 512, the static memory 522, computer readable medium 528, and/or within the processor 511 during execution of the instructions 527 by the processor 511”; Meyer, paragraph 0020, “For traditional petroleum based vehicles, the energy consumption prediction may be generated by the energy prediction tool in terms of an amount of petroleum fuel ( e.g., gasoline, diesel fuel) predicted to be consumed in gallons, liters or other amount of measurable fuel usage, and/or in terms of an energy usage amount ( e.g., kWh, Joules, or other similar unit of energy usage) by one or more vehicle batteries that are included in the vehicle system. For vehicles that rely, at least in part, on one or more batteries for powering the propulsion of the vehicle, the energy consumption prediction may be generated by the energy prediction tool in terms of amount of battery energy predicted to be consumed in term of an energy usage amount ( e.g., kWh, Joules, or other similar unit of energy usage) by one or more vehicle batteries that are included in the vehicle system [hence, a system for estimating a mixed-vehicle fleet, where traditional combustion engine vehicles and electric vehicles are two different vehicle classes].
“that stores route data identifying vehicle routes in a geographic area, each vehicle route comprising a plurality of route segments”: Meyer, paragraph 0006, “The apparatus may include a memory configured to store road segment information for a road segment a vehicle is set to travel on [stores route data identifying vehicle routes in a geographic area, each vehicle route comprising a plurality of route segments], and a processor in communication with the memory.”
“a network interface that receives: vehicle locations indicative of locations of each of the vehicles”: Meyer, paragraph 0076, “In a networked deployment, the computing system 500 may operate in the capacity of a server or as a client user computer within a vehicle in a server-client user network environment, or as a peer computer system within a vehicle in a peer-to-peer ( or distributed) network environment [a network interface]”; Meyer, paragraph 0023, “In some embodiments the starting location 110 may correspond to a current location of the vehicle 101 obtained by the energy prediction tool via driver input, or via location information received from a GPS unit that is part of the vehicle's vehicle system [vehicle locations indicative of locations of each of the vehicles]. In some embodiments the starting location 110 may correspond to a location manually input by the driver that may not correspond to an actual current location of the vehicle.”
“energy consumption data indicative of energy consumed by each of the vehicles”: Meyer, paragraph 0016, “However, even under such embodiments where the historical driving information may not be relied upon, the energy prediction tool may provide an accurate energy consumption prediction for the vehicle based on historical energy consumption information for the vehicle [energy consumption data indicative of energy consumed by each of the vehicles], external information that may affect energy consumption by the vehicle, vehicle systems information that may affect energy consumption by the vehicle, and/or selected road segment information.”
“that generates a predictive function for predicting energy consumption by vehicles in each of a plurality of classes of vehicles” and (bold only) “the one or more neural networks comprising a plurality of shared layers for all of the classes of vehicles and a set of vehicle-specific layers for each of the plurality of classes”: Meyer, paragraph 0058, “After generating the energy consumption predictions from one or more of the wheel energy model 204, elevation model 205, warm up model 206, braking/accelerating model 207, auxiliary load model 208, and climate usage 209, a summation function may be implemented by the energy prediction tool at 210 and 211 [generates a predictive function for predicting energy consumption by vehicles in each of a plurality of classes of vehicles]. For example, at 210 the energy prediction tool may implement the summation of the energy consumption predictions from the wheel energy model 204, elevation model 205, warm up model 206, and braking/accelerating model 207 to generate a propulsive energy consumption prediction.”
Meyer does not explicitly teach:
(bold only) “and one or more neural networks that generates a predictive function for predicting energy consumption by vehicles in each of a plurality of classes of vehicles”
(bold only) “the one or more neural networks comprising a plurality of shared layers for all of the classes of vehicles and a set of vehicle-specific layers for each of the plurality of classes”
Lee teaches:
(bold only) “and one or more neural networks that generates a predictive function for predicting energy consumption by vehicles in each of a plurality of classes of vehicles”: Lee, paragraph 0036, “In the following embodiments, the 'model body' may mean a network structure for extracting a feature vector for input data [one or more neural networks ].”
(bold only) “the one or more neural networks comprising a plurality of shared layers for all of the classes of vehicles and a set of vehicle-specific layers for each of the plurality of classes”: Lee, Fig. 4,
PNG
media_image1.png
491
806
media_image1.png
Greyscale
; Lee, paragraphs 0095–0096, “Thereafter, the model body 220 [a plurality of shared layers for all of the classes] generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420 [a set of …-specific layers for each of the plurality of classes]. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector [interpreting the last layer of a weighted classification network as a conditional probability distribution, hence, identifying a conditional probability distribution over each of the identified features].”
Lee and Meyer are analogous arts as they are both related to data modelling. It would have been obvious to a person having ordinary skill in the art prior to the effective filing date of the claimed invention to have combined the multi-task modelling of Lee with the teachings of Meyer to arrive at the present invention, in order to improve the model and its training, as stated in Lee, paragraph 0023, “In addition, according to the disclosed embodiments, it is possible to reduce the amount of time and resources required for training by making lighten the overall structure of the image classification model by constructing several model heads in parallel while sharing a model body.”
Regarding claim 4:
Meyer as modified by Lee teaches “[t]he system of claim 3.”
Meyer further teaches (bold only) “wherein the one or more neural networks comprise a multi-task learning model comprising the plurality of shared layers for all of the classes of vehicles and the sets of vehicle-specific layers for each of the plurality of classes”: Meyer, paragraph 0016, “However, even under such embodiments where the historical driving information may not be relied upon, the energy prediction tool may provide an accurate energy consumption prediction for the vehicle based on historical energy consumption information for the vehicle [of vehicles], external information that may affect energy consumption by the vehicle, vehicle systems information that may affect energy consumption by the vehicle, and/or selected road segment information.”
Lee further teaches:
“wherein the one or more neural networks comprise a multi-task learning model”: Lee, paragraph 0014, “An apparatus for training an image classification model according to an embodiment disclosed includes a first trainer that trains a model body and a first head through supervised learning based on a labeled data set subjected to type 1 labeling, a second trainer that trains the model body, the first head, and a second head through multi-task learning [a multi-task learning model] based on the labeled data set and an unlabeled data set, and a third trainer that trains a plurality of third heads through supervised learning based on the labeled data set subjected to type 2 labeling while freezing the model body, in which the model body extracts feature vector for input data, and each of the first head, the second head, and the third head generates a classification result based on the feature vector.”
(bold only) “comprising the plurality of shared layers for all of the classes of vehicles and the sets of vehicle-specific layers for each of the plurality of classes”: Lee, paragraphs 0095–0096, “Thereafter, the model body 220 [the plurality of shared layers for all of the classes] generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420 [the sets of …-specific layers for each of the plurality of classes]. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector.”
Lee and Meyer are combinable for the rationale given under claim 3.
Regarding claim 5 and analogous claim 17:
Meyer as modified by Lee teaches “[t]he system of claim 4.”
Meyer further teaches (bold only) “at least one set of vehicle-specific layers for at least one additional class”: Meyer, paragraph 0016, “However, even under such embodiments where the historical driving information may not be relied upon, the energy prediction tool may provide an accurate energy consumption prediction for the vehicle based on historical energy consumption information for the vehicle [of vehicles], external information that may affect energy consumption by the vehicle, vehicle systems information that may affect energy consumption by the vehicle, and/or selected road segment information.”
Lee further teaches:
“wherein the one or more neural networks further comprise an inductive transfer learning model comprising the plurality of shared layers”: Lee, paragraph 0043–0045, “Thereafter, the first trainer 110 may update a training parameter of each of the model body and the first head based on the calculated loss function value. In this case, the training parameter may be, for example, a weight or a bias applied to a layer included in the network structure of each of the model body or the first head. Meanwhile, according to an embodiment, the first trainer 110 may calculate the loss function value described above by using a cross entropy function as a loss function. The second trainer 120 trains the model body, the first head, and a second head through multi-task learning based on the labeled data set subjected to type 1 labeling and an unlabeled data set [an inductive transfer learning model].”
“transferred from the multi-task learning model”: Lee, paragraph 0043–0045, “Thereafter, the first trainer 110 may update a training parameter of each of the model body and the first head based on the calculated loss function value. In this case, the training parameter may be, for example, a weight or a bias applied to a layer included in the network structure of each of the model body or the first head. Meanwhile, according to an embodiment, the first trainer 110 may calculate the loss function value described above by using a cross entropy function as a loss function. The second trainer 120 trains the model body [transferred from the multi-task learning model], the first head, and a second head through multi-task learning based on the labeled data set subjected to type 1 labeling and an unlabeled data set.”
(bold only) “at least one set of vehicle-specific layers for at least one additional class”: Lee, paragraphs 0095–0096, “Thereafter, the model body 220 generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420 [at least one set of … -specific layers for at least one additional class]. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector.”
Lee and Meyer are combinable for the rationale given under claim 3.
Regarding claim 6 and analogous claim 18:
Meyer as modified by Lee teaches “[t]he system of claim 3.”
Meyer further teaches (bold only) “the plurality of shared layers identifies features indicative of energy consumption and a marginal probability distribution over each of the identified features for all of the classes of vehicles” and (bold only) “the set of vehicle-specific layers identifies a conditional probability distribution over each of the identified features”: Meyer, paragraph 0016, “However, even under such embodiments where the historical driving information may not be relied upon, the energy prediction tool may provide an accurate energy consumption prediction for the vehicle based on historical energy consumption information for the vehicle [identifies features indicative of energy consumption][of vehicles], external information that may affect energy consumption by the vehicle, vehicle systems information that may affect energy consumption by the vehicle, and/or selected road segment information.”
Lee further teaches:
(bold only) “the plurality of shared layers identifies features indicative of energy consumption and a marginal probability distribution over each of the identified features for all of the classes of vehicles”: Lee, Fig. 4,
PNG
media_image1.png
491
806
media_image1.png
Greyscale
; Lee, paragraphs 0095–0096, “Thereafter, the model body 220 generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420 [plurality of shared layers]. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector”; Lee, paragraph 0065, “In addition, the second trainer 120 may train the model body based on the total loss function obtained by performing weighted summation of each of the loss functions calculated by supervised learning, self-supervised learning, and unsupervised learning [interpreting the last layer of a weighted classification network as a probability distribution; and such a network with multiple classification heads as a marginal probability distribution over each of the identified features for all of the classes].”
(bold only) “and for each class, the set of vehicle-specific layers identifies a conditional probability distribution over each of the identified features”: Lee, paragraphs 0095–0096, “Thereafter, the model body 220 generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420 [the set of … -specific layers ]. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector [interpreting the last layer of a weighted classification network as a conditional probability distribution, hence, identifies a conditional probability distribution over each of the identified features].”
Lee and Meyer are combinable for the rationale given under claim 3.
Regarding claim 7:
Meyer as modified by Lee teaches “[t]he system of claim 3.”
Meyer further teaches (bold only) “the plurality of shared layers identifies features indicative of energy consumption for all of the classes of vehicles” and (bold only) “for each class, the set of vehicle-specific layers identifies a marginal probability distribution over each of the identified features and a conditional probability distribution over each of the identified features”: Meyer, paragraph 0016, “However, even under such embodiments where the historical driving information may not be relied upon, the energy prediction tool may provide an accurate energy consumption prediction for the vehicle based on historical energy consumption information for the vehicle [identifies features indicative of energy consumption for all of the classes of vehicles][vehicle], external information that may affect energy consumption by the vehicle, vehicle systems information that may affect energy consumption by the vehicle, and/or selected road segment information.”
Lee further teaches:
(bold only) “the plurality of shared layers identifies features indicative of energy consumption for all of the classes of vehicles”: Lee, Fig. 4,
PNG
media_image1.png
491
806
media_image1.png
Greyscale
; Lee, paragraphs 0095–0096, “Thereafter, the model body 220 generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420 [plurality of shared layers]. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector”; Lee, paragraph 0065, “In addition, the second trainer 120 may train the model body based on the total loss function obtained by performing weighted summation of each of the loss functions calculated by supervised learning, self-supervised learning, and unsupervised learning.”
(bold only) “for each class, the set of vehicle-specific layers identifies a marginal probability distribution over each of the identified features and a conditional probability distribution over each of the identified features”: Lee, paragraphs 0095–0096, “Thereafter, the model body 220 generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420 [the set of …-specific layers]. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector [interpreting the last layer of a weighted classification network as a probability distribution; and such a network with multiple classification heads as a marginal probability distribution over each of the identified features][interpreting the last layer of a weighted classification network as a conditional probability distribution, hence, a conditional probability distribution over each of the identified features].”
Lee and Meyer are combinable for the rationale given under claim 3.
Regarding claim 8 and analogous claim 19:
Meyer as modified by Lee teaches “[t]he system of claim 3.”
Meyer further teaches:
“the non-transitory computer readable storage media further stores elevation data indicative of elevations along each route segment”: Meyer, paragraph 0018, “The energy consumption tool may incorporate road segment information (e.g., posted speed limit, elevation [elevation data indicative of elevations along each route segment], number of traffic stops) into an algorithm for generating the energy consumption profile for the vehicle.”
“and a network interface further receives: traffic data indicative of traffic conditions along at least some of the route segments and weather data indicative of weather conditions in the geographic area”: Meyer, paragraph 0076, “In a networked deployment, the computing system 500 may operate in the capacity of a server or as a client user computer within a vehicle in a server-client user network environment, or as a peer computer system within a vehicle in a peer-to-peer ( or distributed) network environment [a network interface]”; Meyer, paragraph 0027, “The road segment information may include, but is not limited to, posted speed limit on the identified road segment, an elevation profile for the identified road segment, current and/or predicted traffic information for the identified road segment [traffic data indicative of traffic conditions along at least some of the route segments], road condition information for the identified road segment, weather information for the identified road segment [weather data indicative of weather conditions in the geographic area], stop sign and traffic light information for the identified road segment, tum sequence information for the identified road segment, or some other identifiable road segment attribute for the identified road segment.”
Regarding claim 9 and analogous claim 20:
Meyer as modified by Lee teaches “[t]he system of claim 8.”
Meyer further teaches:
(bold only) “wherein the one or more neural [networks, see objection, above] generates each predictive function by identifying features indicative of energy consumption along each route segment, the features indicative of energy consumption along each route segment including”: Meyer, paragraph 0027, “The road segment information may include, but is not limited to, posted speed limit on the identified road segment, an elevation profile for the identified road segment, current and/or predicted traffic information for the identified road segment [by identifying features indicative of energy consumption along each route segment], road condition information for the identified road segment, weather information for the identified road segment, stop sign and traffic light information for the identified road segment, tum sequence information for the identified road segment, or some other identifiable road segment attribute for the identified road segment”; Meyer, paragraph 0058, “After generating the energy consumption predictions from one or more of the wheel energy model 204, elevation model 205, warm up model 206, braking/accelerating model 207, auxiliary load model 208, and climate usage 209, a summation function may be implemented by the energy prediction tool at 210 and 211 [generates each predictive function]. For example, at 210 the energy prediction tool may implement the summation of the energy consumption predictions from the wheel energy model 204, elevation model 205, warm up model 206, and braking/accelerating model 207 to generate a propulsive energy consumption prediction.”
“one or more static road segment features for each route segment, identified using the route data and the elevation data”: Meyer, paragraph 0018, “The energy consumption tool may incorporate road segment information ( e.g., posted speed limit, elevation [one or more static road segment features for each route segment, identified using the route data and the elevation data], number of traffic stops) into an algorithm for generating the energy consumption profile for the vehicle.”
“one or more past vehicle speed features for each route segment, identified using the vehicle locations”: Meyer, paragraph 0023, “In some embodiments the starting location 110 may correspond to a current location of the vehicle 101 obtained by the energy prediction tool via driver input, or via location information received from a GPS unit that is part of the vehicle's vehicle system [vehicle locations]. In some embodiments the starting location 110 may correspond to a location manually input by the driver that may not correspond to an actual current location of the vehicle”; Meyer, paragraph 0030, “In terms of the individual models, a speed prediction model 202 may be utilized by the energy prediction tool to generate an estimated speed that is a prediction for a speed at which the vehicle 101 will travel along the identified road segment. The energy prediction tool may determine the estimated speed for the vehicle 101 based on posted speed limit information and traffic information that may have been extracted as part of the road segment information at 201. For example, the energy prediction tool may initially determine that a default speed for the vehicle 101 traveling along the identified road segment should equal the posted speed limit identified in the road segment information [one or more past vehicle speed features for each route segment, identified using the vehicle locations].”
“one or more weather conditions, identified using the weather data; and one or more traffic conditions, identified using the traffic data”: Meyer, paragraph 0027, “The road segment information may include, but is not limited to, posted speed limit on the identified road segment, an elevation profile for the identified road segment, current and/or predicted traffic information for the identified road segment [one or more traffic conditions, identified using the traffic data], road condition information for the identified road segment, weather information for the identified road segment [one or more weather conditions, identified using the weather data], stop sign and traffic light information for the identified road segment, tum sequence information for the identified road segment, or some other identifiable road segment attribute for the identified road segment.”
Lee further teaches (bold only) “wherein the one or more neural [networks, see objection, above] generates each predictive function by identifying features indicative of energy consumption along each route segment, the features indicative of energy consumption along each route segment including”: Lee, paragraph 0036, “In the following embodiments, the 'model body' may mean a network structure for extracting a feature vector for input data [one or more neural networks].”
Lee and Meyer are combinable for the rational given under claim 8.
Regarding claim 10:
Meyer as modified by Lee teaches “[t]he system of claim 3.”
Meyer further teaches “one of the plurality of classes is internal combustion vehicles and one of the plurality of classes is electric vehicles; and the energy consumption data is indicative of fuel consumed by the internal combustion vehicles and electric energy consumed and generated by the electric vehicles”: Meyer, paragraph 0020, “For traditional petroleum based vehicles [internal combustion vehicles ], the energy consumption prediction may be generated by the energy prediction tool in terms of an amount of petroleum fuel (e.g., gasoline, diesel fuel) predicted to be consumed in gallons, liters or other amount of measurable fuel usage [the energy consumption data is indicative of fuel consumed by the internal combustion vehicles], and/or in terms of an energy usage amount ( e.g., kWh, Joules, or other similar unit of energy usage) by one or more vehicle batteries that are included in the vehicle system. For vehicles that rely, at least in part, on one or more batteries for powering the propulsion of the vehicle [electric vehicles], the energy consumption prediction may be generated by the energy prediction tool in terms of amount of battery energy predicted to be consumed in term of an energy usage amount ( e.g., kWh, Joules, or other similar unit of energy usage) by one or more vehicle batteries that are included in the vehicle system [electric energy consumed and generated by the electric vehicles].”
Regarding claim 11:
Meyer as modified by Lee teaches “[t]he system of claim 10.”
Meyer further teaches “one of the plurality of classes is hybrid vehicles”: Meyer, paragraph 0020, “For battery and petroleum fuel hybrid based vehicles [hybrid vehicles], the energy consumption prediction may be generated by the energy prediction tool in terms of an amount of petroleum fuel consumed and an amount of battery energy predicted to be consumed. For alternative fuel based vehicles ( e.g., biodiesel, solar power, liquefied petroleum gas, compressed natural gas, neat ethanol, fuel cells), the energy consumption prediction may be generated by the energy prediction tool in terms of an amount of the alternative fuel predicted to be consumed.”
Regarding claim 14:
Meyer as modified by Lee teaches “[t]he system of claim 3.”
Meyer further teaches “a vehicle trajectory mapping module that maps each vehicle location to one of the route segments in the route data”: Meyer, paragraph 0021, “The specific road segment for which the energy prediction tool generates an energy consumption profile may correspond to one or more available routes to a desired destination. For example, FIG. 1 illustrates an exemplary route planning display that identifies a vehicle 101 located at a starting location 110, and a destination location 120 representing a location where the driver of the vehicle has determined is a desired destination. In between the starting location 110 and the destination location 120, is a first route 1, second route 2, and a third route 3, where each of the routes represent available driving routes for the vehicle 101 to reach the destination location 120 when starting from the starting location 110 [that maps each vehicle location to one of the route segments in the route data].”
Claim 2 rejected under 35 U.S.C. 103 over Meyer as modified by Lee in view of Yogesha et al., US Pre-Grant Publication No. 2021/0407303 (hereafter Yogesha).
Meyer as modified by Lee teaches “[t]he system of claim 1.”
Meyer further teaches:
“wherein the features indicative of energy consumption comprise: length of each route”: Meyer, paragraph 0042, “Some of the factors received by the warm up model 206 for determining the warm up energy consumption prediction may include, but not be limited to, trip distance information [length of each route], initial ambient temperature information, initial tire pressure information, initial coolant temperature information, initial exhaust temperature information, and initial oil temperature information.”
“average past travel speed along each route”: Meyer, paragraph 0030, “In terms of the individual models, a speed prediction model 202 may be utilized by the energy prediction tool to generate an estimated speed that is a prediction for a speed at which the vehicle 101 will travel along the identified road segment. The energy prediction tool may determine the estimated speed for the vehicle 101 based on posted speed limit information and traffic information that may have been extracted as part of the road segment information at 201. For example, the energy prediction tool may initially determine that a default speed for the vehicle 101 traveling along the identified road segment should equal the posted speed limit identified in the road segment information [average past travel speed along each route].”
“past time to travel along each route”: Meyer, paragraph 0033, “A stop prediction model 203 may be utilized by the energy prediction tool to determine a stopping likelihood profile and a travel time estimate for the vehicle 101 [past time to travel along each route].”
“change in elevation along each route”: Meyer, paragraph 0017, “The energy prediction tool may also be configured to distinguish between available routes that have different rates of change in their elevation grades [change in elevation along each route].”
“maximum elevation change along each route”: Meyer, paragraph 0017, “The energy prediction tool may also be configured to distinguish between available routes that have different rates of change in their elevation grades [hence, identifying less elevation change with less energy use, and thus incorporating maximum elevation change along each route].”
“speed ratio along each of at least some of the routes” and “jam factor along each of at least some of the routes”: Meyer, paragraph 0022, “Each road segment may be identified according to a road segment attribute that may include, but not be limited to, a specific road ( e.g., part of a same street or road), speed limit ( e.g., segment of road having a same speed limit, or a speed limit within a predetermined range [speed ratio along each of at least some of the routes]), traffic congestion [jam factor along each of at least some of the routes] (e.g., a segment of road having a same traffic condition, or a traffic condition within a predetermined range of traffic conditions), road conditions (e.g., a segment of road sharing a same, or similar, road condition such as construction), road segment type (e.g., city road, country road, main.”
“temperature in the geographic area”: Meyer, paragraph 0042, “Some of the factors received by the warm up model 206 for determining the warm up energy consumption prediction may include, but not be limited to, trip distance information [length of each route], initial ambient temperature information, initial tire pressure information, initial coolant temperature information, initial exhaust temperature information, and initial oil temperature information.”
Meyer does not explicitly teach:
“precipitation in the geographic area”
“visibility in the geographic area”
“wind speed in the geographic area“
”humidity in the geographic area”
“wind gust in the geographic area”
Yogesha teaches:
“precipitation in the geographic area,” “wind speed in the geographic area,“ and ”humidity in the geographic area”: Yogesha, paragraph 0003, “For example, strong cross winds [wind speed in the geographic area], extreme temperatures, higher humidity [humidity in the geographic area], and/or precipitation [precipitation in the geographic area] may result in conditions in which a UAM vehicle's batteries are drained quickly, resulting in an uncompleted mission.”
“visibility in the geographic area”: Yogesha, paragraph 0068, “The historical flight data 403 may include data that identifies an impact that fog has on energy drain. Moisture may affect the operation of UAM vehicles, regardless of the form of the moisture. Visibility may be measured as the distance at which an object or light can be clearly discerned and may be defined in miles or kilometers. If visibility is less than 0.5 miles, then there is a high probability of fog [visibility in the geographic area].”
“wind gust in the geographic area”: Yogesha, paragraph 0066, “For example, if an average wind speed for a given day is 20 MPH, but there may be wind gusts up to 40 MPH, then the vehicle should plan for and account for the 40 MPH wind [wind gust in the geographic area]. Additionally, flying a vehicle in a windy environment (e.g., headwinds) uses more power than in less windy environments, and the battery may drain faster than normal. Thus, the machine learning model 400 should take into account wind speeds, and may recommend to not fly in high winds or headwinds.”
Yogesha and Meyer are analogous arts as they are both related to modelling vehicle energy consumption through the inclusion of environmental factors. It would have been obvious to a person having ordinary skill in the art prior to the effective filing date of the claimed invention to have combined the additional environmental factors taught in Yogesha with the teachings of Meyer to arrive at the present invention, in order to improve energy consumption modelling, as stated in Yogesha, paragraph 0003, “However, weather conditions may impact a battery residual charge and alter its current supplying properties, which may reduce a mission duration, posing a threat to the safety of the UAM vehicles and their surroundings. For example, strong cross winds, extreme temperatures, higher humidity, and/or precipitation may result in conditions in which a UAM vehicle's batteries are drained quickly, resulting in an uncompleted mission.”
Claims 12–13 rejected under 35 U.S.C. 103 over Meyer as modified by Lee in view of Shirazi et al., US Pre-Grant Publication No. 2022/0388424 (hereafter Shirazi).
Regarding claim 12:
Meyer as modified by Lee teaches “[t]he system of claim 3.”
Meyer further teaches (bold only) “wherein the one or more neural networks comprise a set of vehicle-specific layers for each model of vehicle in the mixed-vehicle fleet”: Meyer, paragraph 0020, “For traditional petroleum based vehicles, the energy consumption prediction may be generated by the energy prediction tool in terms of an amount of petroleum fuel ( e.g., gasoline, diesel fuel) predicted to be consumed in gallons, liters or other amount of measurable fuel usage, and/or in terms of an energy usage amount ( e.g., kWh, Joules, or other similar unit of energy usage) by one or more vehicle batteries that are included in the vehicle system. For vehicles that rely, at least in part, on one or more batteries for powering the propulsion of the vehicle, the energy consumption prediction may be generated by the energy prediction tool in terms of amount of battery energy predicted to be consumed in term of an energy usage amount ( e.g., kWh, Joules, or other similar unit of energy usage) by one or more vehicle batteries that are included in the vehicle system [hence, a system for estimating a vehicle in the mixed-vehicle fleet, where traditional combustion engine vehicles and electric vehicles are two different vehicle classes].
Lee teaches (bold only) “the one or more neural networks comprise a set of vehicle-specific layers for each model of vehicle in the mixed-vehicle fleet”: Lee, paragraphs 0095–0096, “Thereafter, the model body 220 generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420 [the one or more neural networks comprise a set of …-specific layers]. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector.”
Meyer as modified by Lee does not explicitly teach (bold only) “wherein the one or more neural networks comprise a set of vehicle-specific layers for each model of vehicle in the mixed-vehicle fleet.”
Shirazi teaches (bold only) “wherein the one or more neural networks comprise a set of vehicle-specific layers for each model of vehicle in the mixed-vehicle fleet”: Shirazi, paragraph 0032, “According to developments of the method according to the invention, provision is made for not only the vehicle to be started by the energy storage system which is examined (vehicle model [for each model], vehicle make and/or vehicle variant), but also the temperature or a temperature range of a vehicle engine to be started by the electrical energy storage system to be taken into account, with the output result of the evaluation relating to a prediction with respect to the engine start performance of the electrical energy storage system at different temperatures of the vehicle engine.”
Shirazi and Meyer are analogous arts as they are both related to modelling energy performance of vehicles. It would have been obvious to a person having ordinary skill in the art prior to the effective filing date of the claimed invention to have combined the specific vehicle information of Shirazi with the teachings of Meyer to arrive at the present invention, in order to better predict energy use for particular vehicle variants, as stated in Shirazi, paragraph 0035, “In the engine-start prediction algorithm, the engine-start data may be divided according to classifications into different categories, wherein the various patterns differ. In particular, in this case the engine-start data are divided into different categories depending on a vehicle make, a vehicle model and/or a vehicle variant of the vehicle to be started by the electrical energy storage system.”
Regarding claim 13:
Meyer as modified by Lee teaches “[t]he system of claim 3.”
Meyer further teaches (bold only) “the one or more neural networks comprise a set of vehicle-specific layers for each year of each model in the mixed-vehicle fleet”: Meyer, paragraph 0020, “For traditional petroleum based vehicles, the energy consumption prediction may be generated by the energy prediction tool in terms of an amount of petroleum fuel ( e.g., gasoline, diesel fuel) predicted to be consumed in gallons, liters or other amount of measurable fuel usage, and/or in terms of an energy usage amount ( e.g., kWh, Joules, or other similar unit of energy usage) by one or more vehicle batteries that are included in the vehicle system. For vehicles that rely, at least in part, on one or more batteries for powering the propulsion of the vehicle, the energy consumption prediction may be generated by the energy prediction tool in terms of amount of battery energy predicted to be consumed in term of an energy usage amount ( e.g., kWh, Joules, or other similar unit of energy usage) by one or more vehicle batteries that are included in the vehicle system [hence, a system for estimating each model in the mixed-vehicle fleet, where traditional combustion engine vehicles and electric vehicles are two different vehicle classes].
Lee teaches (bold only) “the one or more neural networks comprise a set of vehicle-specific layers for each year of each model in the mixed-vehicle fleet”: Lee, paragraphs 0095–0096, “Thereafter, the model body 220 generates a feature vector obtained by extracting features of the input image data d, and the generated feature vector is input a third head 420. In this case, in FIG. 4, although a feature vector for one image data is illustrated as being simultaneously input to each of the third heads 420-1, 420-2, . . . , 420-N, this is for simplicity of expression, and actually, it should be noted that a feature vector obtained by extracting features of image data classified based on a specific class is input to only one of the third heads 420 [the one or more neural networks comprise a set of …-specific layers]. Thereafter, each of the third heads 420-1, 420-2, ... , 420-N classifies the image data based on the input feature vector.”
Meyer as modified by Lee does not explicitly teach (bold only) “the one or more neural networks comprise a set of vehicle-specific layers for each year of each model in the mixed-vehicle fleet.”
Shirazi teaches (bold only) “wherein the one or more neural networks comprise a set of vehicle-specific layers for each model of vehicle in the mixed-vehicle fleet”: Shirazi, paragraph 0032, “According to developments of the method according to the invention, provision is made for not only the vehicle to be started by the energy storage system which is examined (vehicle model, vehicle make and/or vehicle variant [for each year of each model ), but also the temperature or a temperature range of a vehicle engine to be started by the electrical energy storage system to be taken into account, with the output result of the evaluation relating to a prediction with respect to the engine start performance of the electrical energy storage system at different temperatures of the vehicle engine.”
Shirazi and Meyer are analogous arts as they are both related to modelling energy performance of vehicles. It would have been obvious to a person having ordinary skill in the art prior to the effective filing date of the claimed invention to have combined the specific vehicle information of Shirazi with the teachings of Meyer to arrive at the present invention, in order to better predict energy use for particular vehicle variants, as stated in Shirazi, paragraph 0035, “In the engine-start prediction algorithm, the engine-start data may be divided according to classifications into different categories, wherein the various patterns differ. In particular, in this case the engine-start data are divided into different categories depending on a vehicle make, a vehicle model and/or a vehicle variant of the vehicle to be started by the electrical energy storage system.”
Claim 15 rejected under 35 U.S.C. 103 over Meyer as modified by Lee in view of Shane et al., US Pre-Grant Publication No. 2014/0028254 (hereafter Shane).
Meyer as modified by Lee teaches “[t]he system of claim 3.”
Meyer as modified by Lee does not explicitly teach “wherein the vehicles are buses.”
Shane teaches “wherein the vehicles are buses”: Shane, paragraph 0020, “FIG. 1 illustrates an exemplary embodiment of an exemplary environment 100 in which the present technique is practiced. The environment 100 includes at least one charging station 102, a plurality of electric vehicles 104, 106, 108, and 110, and a processing sub-system 112. In various embodiments, the charging station 102 is located at sites such as residences, offices, commercial or industrial parking lots, and warehouses. For example, for a freight delivery service provider, the charging station 102 may be located at a base parking area for vehicles operated by the freight delivery service provider. The term ‘vehicles’ refers to machines for carrying passengers, cargo, or equipment. The plurality of electric vehicles 104, 106, 108, and 110, may represent automobiles, buses [wherein the vehicles are buses], trucks, equipment carrier trucks, aircrafts, ships, and baggage handling equipment in warehouses, or airports, or dockyards, for example. Each of the plurality of electric vehicles 104, 106, 108, and 110 includes, among other propulsion or drive components of a vehicle, a chargeable energy storage mechanism such as a battery to power the propulsion components.”
Shane and Meyer are analogous arts as they are both related to modeling vehicle energy use. It would have been obvious to a person having ordinary skill in the art prior to the effective filing date of the claimed invention to have combined the application to buses of Shane with the teachings of Meyer to arrive at the present invention, in order to apply the methods to a wider variety of vehicles, as stated in Shane, paragraph 0020, “The term ‘vehicles’ refers to machines for carrying passengers, cargo, or equipment. The plurality of electric vehicles 104, 106, 108, and 110, may represent automobiles, buses, trucks, equipment carrier trucks, aircrafts, ships, and baggage handling equipment in warehouses, or airports, or dockyards, for example.”
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Ayman et al., “Data-Driven Prediction of Route-Level Energy Use for Mixed-Vehicle Transit Fleets,” 2020, arXiv:2004.06043, discloses methods of using trained models to predict energy use in a mixed fleet of vehicles of different classes.
Miftakhov et al., US Pre-Grant Publication No. 2016/0257214, discloses a method for controlling a vehicle charging network that includes energy load predictions, in which the computing system is organized in a hierarchy, such that some components are shared throughout a charging station or the entire system, and others are specific to a vehicle.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to VINCENT SPRAUL whose telephone number is (703) 756-1511. The examiner can normally be reached M-F 9:00 am - 5:00 pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, MICHAEL HUNTLEY can be reached at (303) 297-4307. 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.
/VAS/Examiner, Art Unit 2129
/MICHAEL J HUNTLEY/Supervisory Patent Examiner, Art Unit 2129
Prosecution Insights