Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 20, 21, 22, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, and 39 are rejected under 35 U.S.C. 103 as being unpatentable over Singh (EP 3378679 A1), in view of Ciaravola (US 20210188017 A1).
Regarding Claim 20,
Disclosure by Singh
Singh teaches:
• A computer-implemented method
See at least:
“Aspects of the tire wear estimation system 50 preferably are executed on a processor that is accessible through the vehicle CAN bus, which enables input of data from the sensor 24, as well as input of data from a lookup table or a database that is stored in a suitable storage medium and is in electronic communication with the processor.” Singh, paragraph [0030].
Singh expressly teaches executing the tire wear estimation operations on a processor that electronically receives sensed and stored predictor data. The disclosed tire wear estimation method is therefore computer implemented.
• for calculating a tire wear rate of a vehicle,
See at least:
“The system 50 instead utilizes a tire wear estimation model that receives multiple input parameters to generate a high-accuracy estimation of the rate of tire wear.” Singh, paragraph [0029].
Singh expressly teaches processing multiple input parameters with a tire wear estimation model to produce an estimated rate of tire wear for a tire supporting a vehicle.
• the method comprising:
See at least:
“The present invention also includes a method of estimating the wear rate of a tire 12. The method includes steps in accordance with the description that is presented above and shown in Figures 1 through 14.” Singh, paragraph [0049].
Singh expressly characterizes the disclosed operations as steps of a method for estimating tire wear rate.
• Training a data-driven mathematical tire wear model
See at least:
“The tire wear estimation system 50 generates the estimated wear rate 60 through model fitting, and any appropriate model may be selected. For example, a Multiple Regression Linear Model may be used. By way of background, linear regression is a simple approach to supervised learning.” Singh, paragraph [0040].
“The model fitting is done using stepwise regression, in turn using a forward selection technique, with p-value criteria.” Singh, paragraph [0041].
Singh expressly teaches fitting a mathematical regression model through supervised learning and statistical predictor selection. Model fitting determines the model relationship from data and thus constitutes training a data-driven mathematical tire wear model.
• obtaining technical data of at least one tire of a vehicle;
See at least:
“Each tire 12 preferably is equipped with a sensor or transducer 24 that is mounted to the tire for the purpose of detecting certain real-time tire parameters, such as tire pressure and temperature. The sensor 24 preferably also includes a tire identification for each specific tire 12, and transmits measured parameters and tire ID data to a remote processor.” Singh, paragraph [0028].
Singh expressly teaches obtaining tire pressure, tire temperature, tire identification, and other measured tire parameters from at least one tire and transmitting those data to a processor.
• obtaining technical data of the vehicle;
See at least:
“A first one of the predictors 52 for the tire wear estimation system 50 includes vehicle effects 54. More particularly, one vehicle effect 54 is a wheel position 56 on the vehicle 10.” Singh, paragraph [0031].
“Another vehicle effect 54 is the vehicle drivetrain type 58.” Singh, paragraph [0032].
Singh expressly teaches obtaining technical vehicle information, including the tire wheel position and the vehicle drivetrain type, for use as model predictors.
• obtaining telematics information of the vehicle; and
See at least:
“The route and driver effects 62 in turn include route severity 64 and driver severity 66. The route severity 64 takes into account the amount of turns, starts and stops in a route driven by the vehicle 10.” Singh, paragraph [0034].
“The driver severity 66 takes into account the driving style of the driver of the vehicle 10.” Singh, paragraph [0035].
Singh expressly teaches obtaining information concerning vehicle travel and operation, including route severity, turns, starts, stops, and driving style. These operational data constitute telematics information of the vehicle.
• calculating a tire wear rate
See at least:
“All of the predictors 52 are input into a model 86 to generate the estimated wear rate 60 for a given tire 12.” Singh, paragraph [0040].
Singh expressly teaches evaluating the model using the predictor inputs to produce estimated wear rate 60 as a numerical model output.
• based at least in part on the obtained technical data of the at least one tire of the vehicle,
See at least:
“The dimensional tire effects 68 in turn include the tire rim size 70, the tire width 72, and the tire outer diameter 74.” Singh, paragraph [0036].
“Therefore, the dimensional tire effects 68 comprise one of the predictors 52 to be input into the tire wear estimation system 50.” Singh, paragraph [0036].
Singh expressly uses technical tire characteristics, including rim size, tire width, and tire outer diameter, as predictors supplied to the tire wear estimation model.
• the obtained technical data of the vehicle
See at least:
“Therefore, the wheel position 56 is one of the predictors 52 to be input into the tire wear estimation system 50.” Singh, paragraph [0031].
“Therefore, the drivetrain type 58 has a significant impact on tire wear, and is one of the predictors 52 to be input into the tire wear estimation system 50.” Singh, paragraph [0033].
Singh expressly uses wheel position and drivetrain type as vehicle technical predictors for calculating the tire wear rate.
• and the obtained telematics information of the vehicle
See at least:
“The route and driver effects 62 in turn include route severity 64 and driver severity 66.” Singh, paragraph [0034].
“The route and driver effects 62 may be sensed by the sensor 24, may be included in the tire ID data, and may be stored in the above-described storage medium.” Singh, paragraph [0035].
Singh expressly teaches obtaining route and driver information and supplying those operational variables as predictors for tire wear estimation.
• according to a data-driven mathematical tire wear model.
See at least:
“All of the predictors 52 are input into a model 86 to generate the estimated wear rate 60 for a given tire 12. The tire wear estimation system 50 generates the estimated wear rate 60 through model fitting, and any appropriate model may be selected. For example, a Multiple Regression Linear Model may be used.” Singh, paragraph [0040].
Singh calculates the tire wear rate according to a fitted mathematical regression model. The model is data driven because the relationship between the wear rate and its predictors is determined through model fitting and supervised learning.
Claim Limitations Not Explicitly Taught by Singh
Singh does not explicitly teach the following claim limitations:
• based on a plurality of calculated tire wear rates,
• wherein the data-driven mathematical model is continuously adapted
• to data obtained from a plurality of vehicles,
• wherein the obtained data comprise technical data of at least one tire of each of the vehicles,
• technical data of each of the vehicles,
• and telematics information of each of the vehicles;
Disclosure by Ciaravola
Ciaravola teaches the complementary fleet based training arrangement used to address the remaining limitations.
• based on a plurality of calculated tire wear rates,
See at least:
“Preferably, the ANN is trained based on a given database, for example, a wear fleet database, including tire-usage-related statistical data and corresponding RTM-related statistical data.” Ciaravola, paragraph [0117].
“Training the ANN by carrying out a supervised learning technique including applying to the ANN, for each used tire, the recorded tire-usage-related quantities associated with said used tire as inputs and the respective second correction factor CF2 as output.” Ciaravola, paragraph [0123].
Ciaravola expressly teaches supervised training from multiple respective used-tire records, with each record paired with a calculated quantitative output. Ciaravola does not expressly identify that output as a calculated tire wear rate. Singh, however, expressly teaches a supervised mathematical model whose intended output is tire wear rate. It would have been obvious, when applying Ciaravola’s fleet based supervised training arrangement to Singh’s model, to pair each fleet input record with a respective calculated tire wear rate because supervised fitting uses labeled outputs corresponding to the quantity the model is intended to calculate. This modification uses Ciaravola’s disclosed training structure without changing Singh’s intended model output.
• wherein the data-driven mathematical model is continuously adapted
See at least:
“Preferably, the ANN is trained based on a given database, for example, a wear fleet database, including tire-usage-related statistical data and corresponding RTM-related statistical data.” Ciaravola, paragraph [0117].
“The ANN training may be conveniently performed so as to obtain different ANN specifically trained for different tire models and different vehicle models.” Ciaravola, paragraph [0144].
Ciaravola does not expressly state that the model is continuously adapted. Under the broadest reasonable interpretation, continuous adaptation encompasses repeatedly updating the model as additional fleet data become available. Repeating Ciaravola’s disclosed supervised training process using newly accumulated records would have been obvious because the new records are of the same type used for the original training and would predictably maintain or improve model accuracy as additional tires, vehicles, routes, and operating conditions are represented.
• to data obtained from a plurality of vehicles,
See at least:
“Preferably, the ANN is trained based on a given database, for example, a wear fleet database, including tire-usage-related statistical data and corresponding RTM-related statistical data.” Ciaravola, paragraph [0117].
“The ANN training may be conveniently performed so as to obtain different ANN specifically trained for different tire models and different vehicle models.” Ciaravola, paragraph [0144].
Ciaravola teaches training from a wear fleet database and expressly contemplates training for different vehicle models. A person of ordinary skill would have understood or found it obvious to populate such a fleet database with records obtained from multiple vehicles so that the training data represent the disclosed variations among vehicle models and vehicle operating conditions.
• wherein the obtained data comprise technical data of at least one tire of each of the vehicles,
See at least:
“The tire-usage-related statistical data are indicative of recorded tire-usage-related quantities, for example, vehicle and route parameters typically causing irregular tread wear, such as route severity expressed as Root Mean Square of transversal and longitudinal accelerations, vehicle wheel alignment and tire inflation pressure, associated with used tires.” Ciaravola, paragraph [0118].
“Each pair of RTM amounts is related to a respective used tire and corresponds to respective recorded tire-usage-related quantities associated with said respective used tire.” Ciaravola, paragraph [0119].
Ciaravola expressly maintains respective records associated with respective used tires and identifies tire inflation pressure and remaining tread material quantities as tire related data. In implementing the fleet training database, it would have been obvious to include the selected technical tire variables for at least one tire associated with each vehicle record because those variables form part of the corresponding feature set used for supervised training.
• technical data of each of the vehicles,
See at least:
“The tire-usage-related statistical data are indicative of recorded tire-usage-related quantities, for example, vehicle and route parameters typically causing irregular tread wear, such as route severity expressed as Root Mean Square of transversal and longitudinal accelerations, vehicle wheel alignment and tire inflation pressure.” Ciaravola, paragraph [0118].
“Inputting into the trained ANN the mean-acceleration-related quantities, at least a pressure-related quantity indicative of a tire inflation pressure associated with the tire under tread wear monitoring, and quantities related to toe, camber and load that are associated with the tire and the vehicle under tread wear monitoring.” Ciaravola, paragraph [0137].
Ciaravola expressly teaches using vehicle parameters, vehicle wheel alignment, toe, camber, and vehicle associated load quantities as model inputs. It would have been obvious to retain the selected vehicle technical variables for each fleet record so that corresponding feature values are available for each labeled training observation.
• and telematics information of each of the vehicles;
See at least:
“The tire-usage-related statistical data are indicative of recorded tire-usage-related quantities, for example, vehicle and route parameters typically causing irregular tread wear, such as route severity expressed as Root Mean Square of transversal and longitudinal accelerations.” Ciaravola, paragraph [0118].
Ciaravola expressly teaches recording route severity and longitudinal and transversal acceleration information associated with the used-tire records. These data describe vehicle movement and use during operation and therefore constitute telematics information. Including those selected operational variables for each fleet record would have been an obvious and predictable implementation of Ciaravola’s supervised training database because the model requires corresponding input values for each labeled observation.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to train Singh’s data-driven mathematical tire wear model using Ciaravola’s fleet based supervised training arrangement, with respective tire, vehicle, and telematics records paired with corresponding calculated tire wear rates, and to update the model as additional fleet records became available.
Singh teaches fitting a supervised mathematical model to calculate tire wear rate from tire, vehicle, route, and driver predictors. Ciaravola teaches supervised model training using a wear fleet database containing corresponding usage records and calculated output values for respective used tires. The teachings are technically compatible because Ciaravola’s recorded variables, including tire inflation pressure, vehicle parameters, route severity, acceleration, alignment, toe, camber, and load, correspond to the categories of predictors used by Singh.
Using calculated tire wear rates as the labeled outputs would preserve Singh’s expressly intended dependent variable while applying Ciaravola’s known fleet training arrangement according to its established function. Updating the model as additional fleet observations became available would predictably improve the statistical reliability and representative coverage of the model across different tires, vehicles, routes, and operating conditions. The combination therefore would have constituted the predictable use of known supervised learning and fleet data techniques to improve the accuracy and reliability of Singh’s tire wear rate estimation.
Regarding Claim 21,
The combination of Singh and Ciaravola establishes the method of Claim 20, which is the basis for Claim 21.
Disclosure by Singh
Singh teaches:
• further comprising:
See at least:
“The second embodiment of the wear estimation system 100 incorporates the first embodiment of the wear estimation system 50 as described above, and adds certain real-time predictors 102.” Singh, paragraph [0044].
Singh expressly teaches supplementing the underlying tire wear estimation method with additional real time predictor data.
• obtaining data of at least one in-operational measurement
See at least:
“The second embodiment of the wear estimation system 100 adds predictors 102 that include real-time measurements of sensed conditions of the tire 12.” Singh, paragraph [0044].
Singh expressly teaches obtaining data representing real time measurements of sensed tire conditions. Because the measurements are acquired by the operative tire wear monitoring system from sensors associated with the tire and vehicle, Singh at least implicitly teaches that the measurements are obtained during vehicle and tire operation.
• of at least one property of the at least one tire of the vehicle;
See at least:
“Such real-time measurements include changes in the physical attributes or characteristics of the tire, such as the stiffness of the tread 16.” Singh, paragraph [0045].
Singh expressly identifies tread stiffness as a measured physical property of tire 12, which supports vehicle 10.
• wherein calculating the tire wear rate further comprises calculating the tire wear rate
See at least:
“When the first embodiment of the wear estimation system 50 is integrated with the real-time predictors 102, a predicted wear state 104 is calculated. The predicted wear state 104 includes the above-described wear rate 60 with the addition of corrected real-time predictors.” Singh, paragraph [0046].
Singh teaches an additional calculation in which wear rate 60 is integrated with the real time tire condition predictors to produce a refined predicted wear state.
• based at least in part on the obtained data of the at least one in-operational measurement
See at least:
“On the server 110, the predictors 52 are input into the model 86 for estimation of the wear rate 60, which is integrated with the real-time predictors 102 to yield the predicted wear state 104.” Singh, paragraph [0047].
“The second embodiment of the wear estimation system 100 provides additional refinement and accuracy, as it adds the predictors 102 of real-time measurements of sensed conditions of the tire 12 to the estimation of the wear rate 60.” Singh, paragraph [0048].
Singh expressly teaches adding the obtained real time measurement data to the estimation of wear rate 60. The resulting wear rate determination is therefore based at least in part on those measurement data. To the extent “calculating the tire wear rate” is interpreted as requiring recalculation of the numerical wear rate rather than refinement of a predicted wear state containing that rate, such recalculation would have been obvious because Singh expressly adds the real time predictor to the wear rate estimation for the stated purposes of improving refinement and accuracy.
• of the at least one property of the at least one tire of the vehicle.
See at least:
“Sensors on the tire 12 and/or the vehicle 10 are a first source 114 that measure real-time predictors 102, which are wirelessly transmitted by means known in the art 112 to the server 110.” Singh, paragraph [0047].
Singh expressly teaches measuring the real time predictor with sensors associated with tire 12 of vehicle 10 and transmitting the resulting data for use in the tire wear calculation.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to train and update Singh’s data-driven tire wear rate model using Ciaravola’s fleet based supervised training arrangement, while retaining Singh’s use of real time measurements of sensed tire properties to refine the tire wear rate calculation.
Singh teaches a supervised mathematical model that calculates tire wear rate from tire, vehicle, route, and driver predictors and further teaches adding real time tire condition measurements, including tread stiffness, to improve the refinement and accuracy of the wear estimation. Singh, paragraphs [0040] through [0048].
Ciaravola teaches supervised model training using a wear fleet database containing respective tire usage records and corresponding calculated outputs. The records include tire, vehicle, and operating information, such as tire inflation pressure, vehicle parameters, route severity, acceleration, wheel alignment, toe, camber, and load. Ciaravola, paragraphs [0117] through [0123] and [0135] through [0144].
The teachings are technically compatible because Ciaravola’s fleet data correspond to the same general categories of predictors processed by Singh’s model. Applying Ciaravola’s fleet based training arrangement would improve the representative coverage and reliability of Singh’s underlying model, while Singh’s real time tire property measurements would refine the wear rate calculation during vehicle operation. The combination would therefore have constituted the predictable use of a known fleet based supervised learning technique to improve the accuracy and responsiveness of a known tire wear estimation method.
Regarding Claim 22,
The combination of Singh and Ciaravola establishes the method of Claim 21, which is the basis for Claim 22.
Disclosure by Singh
Singh teaches a processor-based tire wear estimation method in which data received from a tire sensor and data retrieved from a lookup table or database stored in a suitable storage medium are provided to a processor for estimating tire wear rate. Singh further teaches obtaining real-time measurements of sensed tire conditions, including physical tire characteristics such as tread stiffness, and integrating those measurements with the tire wear rate estimation. Singh, paragraphs [0030] and [0044] through [0048].
Singh does not explicitly teach the newly added limitations requiring selection from a plurality of pre-stored tire wear algorithms or requiring that such selection occur after the in-operational tire measurement data are obtained.
Claim Limitations Not Explicitly Taught by Singh
Singh does not explicitly teach:
• selecting one of a plurality of pre-stored algorithms for calculating tire wear rate
• after obtaining data of the at least one in-operational measurement of at least one property of the at least one tire of the vehicle.
Disclosure and Obviousness Based on Ciaravola
Ciaravola renders obvious:
• selecting one of a plurality of pre-stored algorithms for calculating tire wear rate
See at least:
“The tread wear model calibration step 1 may be carried out to obtain, for each type/model of tire, a plurality of respective specifically-calibrated TWMs depending on mounting position of a tire on a motor vehicle, such as a first calibrated TWM for tires mounted on front wheels and a second calibrated TWM for tires mounted on rear wheels, or even, in case of four-wheeled motor vehicles, a first calibrated TWM for a tire mounted on a front right wheel, a second calibrated TWM for a tire mounted on a front left wheel, a third calibrated TWM for a tire mounted on a rear right wheel, and a fourth calibrated TWM for a tire mounted on a rear left wheel.” Ciaravola, paragraph [0042].
“In the tread wear monitoring step 2, the given parameters . . . are known from the tread wear model calibration step 1.” Ciaravola, paragraph [0060].
Ciaravola expressly teaches a plurality of previously calibrated tread wear models corresponding to different tire mounting positions. Each calibrated TWM comprises a mathematical function together with calibration parameters that transform operational inputs into a tire wear result. When executed by a processor, each specifically calibrated TWM constitutes an algorithm for calculating tire wear. Ciaravola further teaches that the calibration parameters are already known when operational monitoring begins, establishing that the models are prepared before the operational calculation.
Ciaravola does not expressly state that all of the calibrated TWMs are stored together and selected during operation. Singh, however, teaches a processor in electronic communication with a suitable storage medium and teaches that tire wheel-position information may be sensed, included in tire identification data, or stored in that storage medium. Singh, paragraphs [0030] and [0031]. It would therefore have been obvious to store executable representations of Ciaravola’s previously calibrated TWMs in Singh’s processor-accessible storage and retrieve the TWM corresponding to the monitored tire position. This would permit the processor to use Ciaravola’s models for their expressly intended position-specific purpose without recalibrating a model during each operational calculation.
Although Ciaravola describes calculation of tread wear rather than using the precise term “tire wear rate,” Singh expressly calculates tire wear rate as its model output. Applying Ciaravola’s plurality of position-specific calibrated models to Singh’s tire wear rate calculation would have predictably resulted in a plurality of pre-stored algorithms for calculating Singh’s tire wear rate, with each algorithm calibrated for the applicable tire position.
• after obtaining data of the at least one in-operational measurement of at least one property of the at least one tire of the vehicle.
See at least:
“Acquiring, from a vehicle bus of a motor vehicle . . . driving-related quantities . . . related to driving of the motor vehicle.” Ciaravola, paragraph [0044].
“Computing, based on the acquired driving-related quantities and a predefined vehicle dynamics model related to the motor vehicle, second frictional-energy-related quantities.” Ciaravola, paragraph [0045].
“Estimating . . . tread wear experienced by the tires of the motor vehicle during driving on the basis of the second frictional-energy-related quantities and of the calibrated TWM.” Ciaravola, paragraphs [0046] and [0047].
Ciaravola expressly teaches an operational sequence in which measurement data are first acquired and the model-based computations are then performed using the acquired data. Singh supplies the particular in-operational tire-property measurements required by Claim 21, including real-time sensed tire conditions such as tread stiffness. Singh, paragraphs [0044] and [0045].
Neither reference expressly states that selection among the stored models occurs after the tire-property data are obtained. In the combined system, however, selection of the applicable calibrated TWM would be an antecedent part of initiating Ciaravola’s post-acquisition model computation. Once the current operational data have been received, Singh’s processor would retrieve the pre-stored TWM corresponding to the known tire type and wheel position and apply that TWM to the current data.
The claim requires that selection occur after the measurement data are obtained, but does not require that the measurement itself determine which algorithm is selected. Selecting the applicable algorithm when the operational data become available would have been a predictable on-demand processing sequence. It would use the same configuration information, the same measurement inputs, and the same calibrated model and would produce the same expected tire wear result, while avoiding retrieval or loading of the model before data are available for processing.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to store Ciaravola’s plurality of previously calibrated, position-specific tread wear models as executable algorithms in Singh’s processor-accessible storage and, after obtaining Singh’s real-time in-operational tire-property measurement data, select and apply the algorithm corresponding to the tire and mounting position being monitored to calculate the tire wear rate.
Singh teaches a processor-based method that obtains real-time tire-property data, stores tire and vehicle predictor information, and calculates tire wear rate using a mathematical model. Singh, paragraphs [0030], [0031], and [0044] through [0048]. Ciaravola complements Singh by teaching a plurality of previously calibrated tread wear models specifically corresponding to different tire mounting positions and by using those calibrated models after operational data acquisition to estimate tire wear. Ciaravola, paragraphs [0042], [0044] through [0047], and [0060].
The teachings are technically compatible because Singh already identifies wheel position as a tire wear predictor and provides processor-accessible storage, while Ciaravola provides different calibrated models for those respective wheel positions. Storing and selecting the applicable calibrated model would predictably improve accuracy by applying position-specific calibration values rather than a single generic calibration to every tire position. Selecting the applicable model after the current operational data are received would also follow Ciaravola’s acquire-then-compute workflow and permit on-demand execution when the data are available.
The proposed combination would not change the operating principle of either reference. It would use Singh’s processor, storage, real-time measurements, and tire wear rate output together with Ciaravola’s position-specific calibrated models according to their established functions, yielding the predictable result of a more accurate and efficiently implemented tire wear rate calculation.
Regarding Claim 26,
Disclosure by Singh
Singh teaches:
• A computer-implemented method
See at least:
“Aspects of the tire wear estimation system 50 preferably are executed on a processor that is accessible through the vehicle CAN bus.” Singh, paragraph [0030].
Singh expressly teaches processor execution of the tire wear estimation operations.
• for calculating a tire wear rate of a vehicle,
See at least:
“The system 50 instead utilizes a tire wear estimation model that receives multiple input parameters to generate a high-accuracy estimation of the rate of tire wear.” Singh, paragraph [0029].
Singh expressly teaches model-based calculation of an estimated tire wear rate for a tire supporting a vehicle.
• the method comprising:
See at least:
“The present invention also includes a method of estimating the wear rate of a tire 12.” Singh, paragraph [0049].
Singh expressly characterizes the disclosed tire wear estimation operations as a method.
• Training a data-driven mathematical tire wear model
See at least:
“The tire wear estimation system 50 generates the estimated wear rate 60 through model fitting.... For example, a Multiple Regression Linear (MLR) Model may be used.” Singh, paragraph [0040].
“The model fitting is done using stepwise regression, in turn using a forward selection technique, with p-value criteria.” Singh, paragraph [0041].
Singh expressly teaches fitting a mathematical regression model through supervised statistical learning. Model fitting determines the model relationship and selected predictors from data and therefore constitutes training a data-driven mathematical tire wear model.
• transmitting technical data of at least one tire of a vehicle;
See at least:
“The tire sensor 24 may also transmit certain selected predictors 52, such as the ambient temperature 78 and tire identification data, to the server 110.... Still other selected predictors 52 for estimation of the wear rate 60, such as tread compound data 82 and tread structure data 84, may be sent from a third source 118 to the server 110.” Singh, paragraph [0047].
Singh expressly teaches transmitting tire identification, tread-compound, and tread-structure data to a remote server for tire wear estimation.
• transmitting technical data of the vehicle;
See at least:
“Sensors on the tire 12 and/or the vehicle 10 are a first source 114 that measure real-time predictors 102, which are wirelessly transmitted by means known in the art 112 to the server 110.” Singh, paragraph [0047].
Singh expressly teaches transmitting predictors measured by vehicle sensors. Singh further identifies wheel position and drivetrain type as vehicle-effect predictors in paragraphs [0031] through [0033]. In the disclosed remote-server implementation, transmission of selected vehicle predictors to the server is at least implicit because the server receives the predictor set used for the model calculation.
• transmitting telematics information of the vehicle; and
See at least:
“Other selected predictors 52 for estimation of the wear rate 60, such as location, weather, and road condition data, may be transmitted from a second source 116 to the server 110.” Singh, paragraph [0047].
Singh expressly teaches transmitting location, weather, and road-condition information for use in tire wear estimation.
• obtaining a calculated tire wear rate
See at least:
“The predicted wear state 104 includes the above-described wear rate 60.... The predicted wear state 104 is wirelessly transmitted by means known in the art 112 to a device 120 for display to a user or a technician, such as a smartphone.” Singh, paragraphs [0046] and [0047].
Singh expressly teaches that device 120 receives a transmitted predicted wear state containing the previously calculated wear rate 60. The receiving device therefore obtains a calculated tire wear rate.
• based at least in part on the transmitted technical data of the at least one tire of the vehicle,
See at least:
“The tire sensor 24 may also transmit... tire identification data.... [T]read compound data 82 and tread structure data 84 may be sent... to the server 110. On the server 110, the predictors 52 are input into the model 86 for estimation of the wear rate 60.” Singh, paragraph [0047].
Singh expressly uses the transmitted tire technical data as model predictors for calculating wear rate 60.
• the transmitted technical data of the vehicle,
See at least:
“Sensors on the tire 12 and/or the vehicle 10... measure real-time predictors 102, which are wirelessly transmitted... to the server 110.... On the server 110, the predictors 52 are input into the model 86 for estimation of the wear rate 60.” Singh, paragraph [0047].
Singh expressly uses the transmitted vehicle-sensor predictors in the server-based wear calculation.
• and the transmitted telematics information of the vehicle,
See at least:
“Other selected predictors 52 for estimation of the wear rate 60, such as location, weather, and road condition data, may be transmitted... to the server 110.” Singh, paragraph [0047].
Singh expressly identifies the transmitted telematics data as predictors for estimating wear rate 60.
• wherein the calculated tire wear rate is calculated
See at least:
“On the server 110, the predictors 52 are input into the model 86 for estimation of the wear rate 60.” Singh, paragraph [0047].
Singh expressly teaches calculation of the wear rate by the server before transmission of the resulting predicted wear state.
• according to the data-driven mathematical tire wear model.
See at least:
“All of the predictors 52 are input into a model 86 to generate the estimated wear rate 60.... For example, a Multiple Regression Linear (MLR) Model may be used.” Singh, paragraph [0040].
Singh expressly calculates the wear rate according to a fitted mathematical regression model.
Claim Limitations Not Explicitly Taught by Singh
Singh does not explicitly teach:
• based on a plurality of calculated tire wear rates,
• wherein the data-driven mathematical model is continuously adapted
• to data obtained from a plurality of vehicles,
• wherein the obtained data comprise technical data of at least one tire of each of the vehicles,
• technical data of each of the vehicles,
• and telematics information of each of the vehicles;
Disclosure and Obviousness Based on Ciaravola
Ciaravola teaches or renders obvious:
• based on a plurality of calculated tire wear rates,
See at least:
“Preferably, the ANN is trained... based on a given database (e.g., a wear fleet database) including tire-usage-related statistical data and corresponding RTM-related statistical data.” Ciaravola, paragraph [0117].
“Training the ANN... includ[es] applying to the ANN, for each used tire, the recorded tire-usage-related quantities associated with said used tire as inputs and the respective second correction factor CF2 as output.” Ciaravola, paragraph [0123].
Ciaravola expressly teaches supervised training using multiple respective fleet records paired with corresponding calculated quantitative outputs. Ciaravola uses CF2 rather than tire wear rate as the supervised output. Singh, however, expressly defines tire wear rate as the dependent output of its supervised mathematical model and identifies different wear rates for different tires and vehicle configurations in paragraph [0033]. It would have been obvious, when applying Ciaravola’s fleet-based supervised-training arrangement to Singh’s tire wear-rate model, to pair each fleet input record with the corresponding calculated tire wear rate because supervised fitting requires labeled outputs representing the quantity the model is intended to estimate.
• wherein the data-driven mathematical model is continuously adapted
See at least:
“Preferably, the ANN is trained... based on... a wear fleet database.” Ciaravola, paragraph [0117].
“The ANN training... may be conveniently performed so as to obtain different ANN specifically trained for different tire models and/or different vehicle models.” Ciaravola, paragraph [0144].
Ciaravola does not expressly use the phrase “continuously adapted.” Under the reasonable interpretation that continuous adaptation encompasses repeated or incremental updating as additional fleet records become available, it would have been obvious to update the trainable model with newly accumulated records of the same type used for its original supervised training. The new records would provide additional examples of tire, vehicle, route, and operating conditions and would predictably maintain or improve estimation accuracy. If the application specification instead requires uninterrupted real-time online learning, the present combination should be reconsidered.
• to data obtained from a plurality of vehicles,
See at least:
“Preferably, the ANN is trained... based on... a wear fleet database.” Ciaravola, paragraph [0117].
“The ANN training... may be... performed so as to obtain different ANN specifically trained for different... vehicle models.” Ciaravola, paragraph [0144].
Ciaravola teaches fleet-based training directed to different vehicle models. Populating that database with respective records obtained from multiple vehicles would have been the predictable implementation for representing the disclosed vehicle-model and operating-condition variations.
• wherein the obtained data comprise technical data of at least one tire of each of the vehicles,
See at least:
“The tire-usage-related statistical data are indicative of recorded tire-usage-related quantities... such as... tire inflation pressure, associated with used tires.” Ciaravola, paragraph [0118].
“Each pair of RTM amounts is related to a respective used tire and corresponds to respective recorded tire-usage-related quantities associated with said respective used tire.” Ciaravola, paragraph [0119].
Ciaravola expressly teaches respective tire-associated records containing tire pressure and remaining-tread information. Maintaining the selected tire variables for at least one tire associated with each contributing fleet vehicle would have been the predictable use of a consistent supervised-training feature set.
• technical data of each of the vehicles,
See at least:
“The tire-usage-related statistical data are indicative of recorded tire-usage-related quantities (e.g., vehicle and route parameters... such as... vehicle wheel alignment and tire inflation pressure).” Ciaravola, paragraph [0118].
Ciaravola expressly includes vehicle parameters and wheel-alignment information in the fleet records. Recording the selected vehicle variables for each labeled vehicle observation would provide the corresponding feature values required for consistent supervised training.
• and telematics information of each of the vehicles;
See at least:
“The tire-usage-related statistical data are indicative of recorded tire-usage-related quantities... such as route severity expressed as Root Mean Square (RMS) of transversal and longitudinal accelerations.” Ciaravola, paragraph [0118].
Ciaravola expressly includes route and acceleration data in the fleet records. These vehicle-operation data constitute telematics information. Recording the selected telematics variables for each labeled vehicle observation would have been the predictable implementation of Ciaravola’s fleet-training arrangement.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to train and repeatedly update Singh’s data-driven mathematical tire wear-rate model using Ciaravola’s fleet-based supervised-training arrangement, with respective tire, vehicle, and telematics input records paired with corresponding calculated tire wear rates.
Singh teaches a server-based supervised mathematical model that calculates tire wear rate from transmitted tire, vehicle, and telematics predictors. Ciaravola teaches supervised training from respective fleet records containing corresponding tire, vehicle, route, and operating inputs and calculated outputs. The input categories are technically compatible, and using Singh’s intended tire wear-rate output as the label for Ciaravola’s fleet-training records would apply the disclosed supervised-learning process according to its established function.
The combination would predictably improve the statistical reliability and representative coverage of Singh’s model across different tires, vehicles, routes, and operating conditions without changing Singh’s principle of operation.
After combining Singh and Ciaravola, all limitations of Claim 26 are taught or rendered obvious under the stated interpretation of “continuously adapted.”
Regarding Claim 27,
The combination of Singh and Ciaravola establishes the method of Claim 26, which is the basis for Claim 27.
Disclosure by Singh
Singh teaches:
• further comprising:
See at least:
“The second embodiment of the wear estimation system 100 incorporates the first embodiment... and adds certain real-time predictors 102.” Singh, paragraph [0044].
Singh expressly supplements the underlying method with additional real-time predictor operations.
• transmitting data of at least one in-operational measurement
See at least:
“Sensors on the tire 12 and/or the vehicle 10 are a first source 114 that measure real-time predictors 102, which are wirelessly transmitted... to the server 110.” Singh, paragraph [0047].
Singh expressly teaches measuring real-time predictors during operation and wirelessly transmitting the resulting measurement data.
• of at least one property of the at least one tire of the vehicle;
See at least:
“Such real-time measurements include changes in the physical attributes or characteristics of the tire, such as the stiffness of the tread 16.” Singh, paragraph [0045].
Singh expressly identifies tread stiffness as a measured physical property of the vehicle tire.
• wherein the calculated tire wear rate is further calculated
See at least:
“When the first embodiment... is integrated with the real-time predictors 102, a predicted wear state 104 is calculated.” Singh, paragraph [0046].
Singh expressly performs a further calculation that integrates the original wear-rate estimate with the real-time predictors.
• based at least in part on the at least one in-operational measurement
See at least:
“The predictors 52 are input into the model 86 for estimation of the wear rate 60, which is integrated with the real-time predictors 102 to yield the predicted wear state 104.” Singh, paragraph [0047].
Singh expressly bases the refined tire-wear determination at least partly on the transmitted real-time measurement data.
• of the at least one property of the at least one tire of the vehicle.
See at least:
“The second embodiment... adds the predictors 102 of real-time measurements of sensed conditions of the tire 12 to the estimation of the wear rate 60.” Singh, paragraph [0048].
Singh expressly uses the measured tire condition in the wear-rate estimation.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to use Ciaravola’s fleet-trained model in Singh’s server-based method while retaining Singh’s transmission and integration of real-time measurements of sensed tire properties.
Ciaravola’s fleet training improves the representative accuracy of the underlying model, while Singh’s transmitted real-time measurement supplies current tire-condition information for operational refinement. The teachings perform complementary functions and would predictably produce a more accurate tire wear determination.
After combining Singh and Ciaravola, all limitations of Claim 27 are taught or rendered obvious.
Regarding Claim 28,
The combination of Singh and Ciaravola establishes the method of Claim 20, which is the basis for Claim 28.
Claim Limitations Not Explicitly Taught by Singh
Singh does not explicitly teach:
• estimating a residual tread depth and/or a remaining mileage of the tire and/or a remaining time before change
• according to a stored minimum tread depth threshold,
• based on the calculated tire wear rate.
Disclosure and Obviousness Based on Ciaravola
Ciaravola teaches or renders obvious:
• estimating a residual tread depth and/or a remaining mileage of the tire and/or a remaining time before change
See at least:
“Estimating RTM... includes computing a remaining tread depth based on the tread wear value and an initial tread depth.” Ciaravola, paragraph [0062].
The limitation is written in the alternative through “and/or.” Ciaravola expressly teaches the residual-tread-depth alternative by computing remaining tread depth.
• according to a stored minimum tread depth threshold,
See at least:
“Estimating RTM... includes detecting an approaching end-of-life condition... if the remaining tread depth reaches a predefined threshold.” Ciaravola, paragraph [0063].
Ciaravola expressly compares remaining tread depth with a predefined minimum threshold. Singh teaches processor-accessible storage in paragraph [0030]. It would have been obvious to retain Ciaravola’s predefined threshold in that storage because the processor must access the threshold to perform the automated comparison.
• based on the calculated tire wear rate.
See at least:
“Estimating tread wear... includes computing... a tread wear value indicative of a reduction in tread depth due to a distance driven by the motor vehicle, and estimating RTM... includes computing a remaining tread depth based on the tread wear value and an initial tread depth.” Ciaravola, paragraph [0062].
“Tread wear and RTM estimation... is carried out every N kilometers/miles driven by the motor vehicle.” Ciaravola, paragraph [0064].
Ciaravola expressly derives remaining depth from cumulative tread wear over a driven-distance interval. Singh supplies a calculated tire wear rate. It would have been obvious to apply Singh’s rate over Ciaravola’s known distance interval to determine cumulative tread loss and subtract that loss from the initial tread depth. This is the ordinary use of a rate to calculate accumulated change over a known interval and predictably yields Ciaravola’s remaining tread depth.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to apply Singh’s calculated tire wear rate over Ciaravola’s known driven-distance interval, estimate the resulting remaining tread depth, and compare that depth with a predefined threshold stored in processor-accessible memory.
The modification would convert Singh’s wear-rate output into actionable remaining-tread information and predictably improve safety by identifying a tire approaching its minimum acceptable tread depth.
After combining Singh and Ciaravola, all limitations of Claim 28 are taught or rendered obvious through the expressly claimed residual-tread-depth alternative.
Regarding Claim 29,
The combination of Singh and Ciaravola establishes the method of Claim 28, which is the basis for Claim 29.
Disclosure by Singh
Singh teaches:
• reporting at least one of the calculated tire wear rate, the estimated residual tread depth, the remaining mileage of the tire, and the remaining time before change according to the stored minimum tread depth threshold.
See at least:
“All of the predictors 52 are input into a model 86 to generate the estimated wear rate 60 for a given tire 12.” Singh, paragraph [0040].
“Once the estimated wear rate 60 is generated, it is communicated from the tire wear estimation system 50 to the vehicle operating systems, such as braking and stability control systems, through the vehicle CAN bus.” Singh, paragraph [0042].
The limitation requires reporting at least one of the listed alternatives. Singh expressly teaches that model 86 calculates estimated wear rate 60 and that the resulting wear rate is communicated to vehicle operating systems through the vehicle CAN bus. Communicating the generated wear-rate result constitutes reporting the calculated tire wear rate. Because the calculated tire wear rate is one of the expressly recited alternatives, Singh satisfies the limitation without requiring separate reporting of residual tread depth, remaining mileage, or remaining time before change.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to retain Singh’s reporting of the calculated tire wear rate when incorporating Ciaravola’s remaining-tread-depth and minimum-threshold evaluation into Singh’s tire wear estimation method.
Singh expressly calculates tire wear rate 60 and communicates that result to vehicle operating systems through an existing CAN-bus architecture. Ciaravola complements Singh by using calculated tire wear information to determine remaining tread depth and identify an approaching end-of-life condition. Retaining Singh’s reporting operation in the combined method would require no change to the underlying calculations and would predictably make the calculated tire wear rate available for vehicle control, safety, maintenance, or notification functions.
The combination would therefore use Singh’s reporting functionality and Ciaravola’s remaining-tread evaluation according to their established functions and would predictably improve the practical utility of the combined tire wear monitoring method.
Regarding Claim 30,
The combination of Singh and Ciaravola establishes the method of Claim 29, which is the basis for Claim 30.
Disclosure by Singh
Singh teaches:
• wherein the control system is arranged in the vehicle.
See at least:
“Once the estimated wear rate 60 is generated, it is communicated from the tire wear estimation system 50 to the vehicle operating systems, such as braking and stability control systems, through the vehicle CAN bus.” Singh, paragraph [0042].
Singh expressly teaches communicating the calculated tire wear rate to braking and stability control systems through the vehicle CAN bus. The braking and stability control systems are vehicle operating systems arranged in the vehicle. Singh therefore teaches the additional limitation of Claim 30.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to report the tire wear information generated by the combined method to Singh’s in-vehicle braking or stability control system through the vehicle CAN bus.
Singh expressly provides an in-vehicle control architecture for receiving the calculated tire wear rate. Ciaravola supplies the remaining-tread-depth and threshold evaluation incorporated into the method of Claims 28 and 29. Retaining Singh’s in-vehicle control system would use the existing CAN-bus communication arrangement according to its established function and would predictably permit immediate vehicle use of the reported tire-condition information for control, safety, maintenance, or driver-notification purposes.
Regarding Claim 31,
The combination of Singh and Ciaravola establishes the method of Claim 29, which is the basis for Claim 31.
Claim Limitations Not Explicitly Taught by Singh
Singh does not explicitly teach:
• wherein the control system is arranged outside the vehicle,
• and configured to collect,
• from a plurality of vehicles,
• at least the calculated tire wear rate, the estimated residual tread depth, the remaining mileage of the tire, and the remaining time before change according to the stored minimum tread-depth threshold.
Disclosure and Obviousness Based on Ciaravola
Ciaravola renders obvious:
• wherein the control system is arranged outside the vehicle,
See at least:
“The processing device/system 32 is implemented/carried out by means of a cloud computing system 32A that is wirelessly and remotely connected to the acquisition device 31.” Ciaravola, paragraph [0077].
Ciaravola expressly teaches implementing the tire-wear processing and control functionality through a remotely connected cloud computing system arranged outside the vehicle.
• and configured to collect, from a plurality of vehicles,
See at least:
“The cloud computing system 32A may be advantageously used to provide many motor vehicles 4 and, hence, many users 5 with a tread wear monitoring service.” Ciaravola, paragraph [0080].
Ciaravola expressly teaches a common cloud system serving many vehicles. Because the cloud system is wirelessly connected to the respective vehicle acquisition devices and performs the tire-wear monitoring service using received vehicle data, Ciaravola teaches collecting tire-wear monitoring information from a plurality of vehicles.
• at least the calculated tire wear rate, the estimated residual tread depth, the remaining mileage of the tire, and the remaining time before change according to the stored minimum tread-depth threshold.
See at least:
“Once the estimated wear rate 60 is generated, it is communicated from the tire wear estimation system 50 to the vehicle operating systems, such as braking and stability control systems, through the vehicle CAN bus.” Singh, paragraph [0042].
“Estimating tread wear . . . includes computing, based on the second frictional-energy-related quantities and the calibrated TWM, a tread wear value indicative of a reduction in tread depth due to a distance driven by the motor vehicle, and estimating RTM . . . includes computing a remaining tread depth based on the tread wear value and an initial tread depth.” Ciaravola, paragraph [0062].
“Estimating RTM . . . includes detecting an approaching end-of-life condition for the tires of the motor vehicle if the remaining tread depth reaches a predefined threshold.” Ciaravola, paragraph [0063].
“Tread wear and RTM estimation . . . is carried out every N kilometers/miles driven by the motor vehicle.” Ciaravola, paragraph [0064].
Singh expressly teaches calculating and electronically communicating tire wear rate. Ciaravola expressly teaches calculating remaining tread depth from accumulated tread wear and comparing the remaining depth with a predefined minimum threshold. Ciaravola further relates the wear calculation to kilometers or miles driven.
Ciaravola does not expressly identify “remaining mileage” and “remaining time before change” as outputs of its principal embodiment. Those outputs nevertheless would have been obvious to a person of ordinary skill from the disclosed quantities. Once the current remaining tread depth, minimum tread-depth threshold, and tire wear rate are known, the remaining distance before the threshold is reached is predictably determined from the tread-depth margin and the rate of tread loss per unit distance. The remaining time before change is predictably determined from that remaining distance and the vehicle’s known or historically determined distance traveled per unit time. These calculations merely express the same predicted threshold crossing disclosed by Ciaravola in distance and time units and require no change to the underlying tread-wear model.
It further would have been obvious to transmit and collect all four related outputs in Ciaravola’s cloud system. The calculated tire wear rate, remaining tread depth, remaining mileage, and remaining time concern the same tire-condition determination, are generated from the same wear and usage data, and collectively provide the information needed for centralized fleet maintenance and replacement scheduling. Sending the complete output record over the existing wireless connection would have been a predictable use of the disclosed communication and cloud-processing architecture and would not have changed the calculations or their expected results.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to implement the control system receiving the reported tire-wear information of Claim 29 as Ciaravola’s remote cloud computing system and configure that system to collect, from multiple vehicles, the calculated tire wear rate, remaining tread depth, remaining mileage before the minimum tread-depth threshold is reached, and corresponding remaining time before tire change.
Singh teaches calculation and electronic reporting of tire wear rate. Ciaravola teaches calculation of remaining tread depth, comparison with a predefined end-of-life threshold, distance-based tire-wear monitoring, and a remote cloud system providing tire-wear monitoring services to many vehicles. The teachings are technically compatible because the four claimed metrics are related outputs derived from the same tire-wear, remaining-depth, threshold, and vehicle-usage data.
A person of ordinary skill would have been motivated to derive and collect the complete metric set to permit the cloud system to identify both the present condition of each tire and the predicted distance and time remaining before replacement. Centralizing those outputs would predictably improve fleet maintenance scheduling, permit comparison among vehicles, and provide timely tire-replacement information while using the existing wireless and cloud architecture according to its established function.
Regarding Claim 32,
The combination of Singh and Ciaravola establishes the method of Claim 20, which is the basis for Claim 32.
Disclosure by Singh
Singh teaches:
• wherein the technical data of the at least one tire of a vehicle include at least one of
• the tire manufacturer, the tire model, the tire pattern, the tire specification, the tire size, the tire mounting position, retread information, and batch number of the tire.
See at least:
“The dimensional tire effects 68 in turn include the tire rim size 70, the tire width 72, and the tire outer diameter 74.” Singh, paragraph [0036].
“One vehicle effect 54 is a wheel position 56 on the vehicle 10.” Singh, paragraph [0031].
The limitation requires at least one listed item. Singh expressly teaches tire-size information and tire mounting position. Either disclosure independently satisfies the alternative limitation.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to retain Singh’s tire-size or tire-mounting-position data as an input to the combined fleet-trained model because both references recognize that tire configuration and mounting position materially affect tire wear.
After combining Singh and Ciaravola, all limitations of Claim 32 are taught or rendered obvious.
Regarding Claim 33,
The combination of Singh and Ciaravola establishes the method of Claim 21, which is the basis for Claim 33.
Disclosure by Singh
Singh teaches:
• wherein performing the in-operational measurement of the at least one property of the at least one tire of the vehicle includes
See at least:
“The second embodiment of the wear estimation system 100 adds predictors 102 that include real-time measurements of sensed conditions of the tire 12.” Singh, paragraph [0044].
“Such real-time measurements include changes in the physical attributes or characteristics of the tire, such as the stiffness of the tread 16.” Singh, paragraph [0045].
Singh expressly teaches performing real time measurements of physical properties of the monitored tire during vehicle operation. Singh does not expressly identify residual tread depth as the particular real time property measured in its exemplary embodiment.
Claim Limitations Not Explicitly Taught by Singh
Singh does not explicitly teach:
• measuring the residual tread depth in operation,
• and associating the measured residual tread depth
• with an odometer reading
• corresponding to a distance traveled of the vehicle.
Disclosure and Obviousness Based on Ciaravola
Ciaravola teaches or renders obvious:
• measuring the residual tread depth in operation,
See at least:
“The estimated tread wear and RTM may be advantageously corrected based on tread depth measurements performed by means of a tread depth measuring tool.” Ciaravola, paragraph [0065].
“One approach to the monitoring and/or measurement of tread wear has been through the use of wear sensors disposed in the tire tread, which has been referred to as a direct method or approach.” Singh, paragraph [0003].
Ciaravola expressly teaches measuring the actual current tread depth of a tire and using the measured value to correct or reset the ongoing tread wear and remaining tread estimation. The current tread depth is the residual tread depth because it represents the tread depth remaining after operational wear.
Ciaravola performs the exemplary measurement with a tread depth measuring tool and does not expressly require that the measurement occur while the tire is operating. Singh, however, acknowledges that direct measurement of tread wear using sensors disposed in the tire tread was known. Singh further teaches tire mounted sensors that obtain and transmit real time tire property measurements during operation. Singh, paragraphs [0028] and [0044] through [0047].
A person of ordinary skill would have found it obvious to use the known direct tread wear sensor as the measurement source for Ciaravola’s corrective tread depth value. The direct sensor and the external measuring tool address the same physical characteristic, namely the extent of tread remaining after wear. The sensor would provide the current tread wear state electronically during operation, and the known initial tread depth or calibrated sensor position would permit that measured wear state to be represented as the residual tread depth used by Ciaravola’s correction process.
The modification is supported by a recognized design incentive stated in Ciaravola. Ciaravola explains that drivers may fail to check tire tread depth because they lack a measuring tool or forget to have the tires periodically checked at a tire shop. Ciaravola, paragraph [0003]. Obtaining the correction measurement during operation would eliminate dependence on the driver and tire shop visit, reduce the interval during which estimation error can accumulate, and provide a more current residual tread depth for the ongoing wear calculation.
• and associating the measured residual tread depth
See at least:
“Receiving, by one or more processors, data associated with one or more tread depth measurements . . . descriptive of a tread depth of at least one tread of at least one tire of the vehicle.” Ciaravola, paragraph [0006].
“Associating, by the one or more processors, a respective time value or a distance value with each of the one or more tread depth measurements.” Ciaravola, paragraph [0007].
Ciaravola expressly identifies, in its discussion of known tread life monitoring systems, processor based association of each tread depth measurement with a corresponding distance value. This state of the art disclosure is relied upon because it directly addresses the measurement to distance association not repeated in Ciaravola’s preferred correction embodiment.
Ciaravola’s detailed embodiment is consistent with that known association. It teaches correcting or resetting the wear estimate using a physical tread depth measurement and performing the ongoing estimation according to kilometers or miles driven. Ciaravola, paragraphs [0064] and [0065]. A person of ordinary skill would therefore associate the correction measurement with the distance point at which it was obtained so that the processor could continue the distance based calculation from the corrected tread depth.
• with an odometer reading
See at least:
“Associating, by the one or more processors, a respective time value or a distance value with each of the one or more tread depth measurements.” Ciaravola, paragraph [0007].
“Preferably, tread wear and RTM estimation . . . is carried out every N kilometers/miles driven by the motor vehicle.” Ciaravola, paragraph [0064].
“An acquisition device 31 . . . [is] coupled to a vehicle bus 40 . . . to acquire the driving-related quantities.” Ciaravola, paragraphs [0067] through [0069].
Ciaravola expressly teaches associating the tread depth measurement with a distance value, performing the wear estimation according to kilometers or miles driven, and acquiring vehicle operating information from the vehicle bus. Ciaravola does not expressly identify the distance value as an odometer reading.
Using the contemporaneous odometer reading as the distance value would have been a predictable implementation. The odometer provides the cumulative vehicle distance at the time the tread depth measurement is obtained. Using that value would directly anchor the correction measurement to the vehicle’s distance history without requiring the processor to reconstruct cumulative mileage from separate speed or wheel rotation samples.
This implementation would also preserve the accuracy of the corrected wear rate. A tread depth value without its corresponding mileage would not identify the distance interval over which the observed tread loss occurred. Associating the measurement with the current odometer reading provides the processor with both endpoints needed to relate physical tread loss to accumulated vehicle travel.
• corresponding to a distance traveled of the vehicle.
See at least:
“Estimating tread wear . . . includes computing . . . a tread wear value indicative of a reduction in tread depth due to a distance driven by the motor vehicle.” Ciaravola, paragraph [0062].
“Preferably, tread wear and RTM estimation . . . is carried out every N kilometers/miles driven by the motor vehicle.” Ciaravola, paragraph [0064].
Ciaravola expressly relates tread depth reduction and the periodic wear calculation to the distance driven by the vehicle. The contemporaneous odometer reading therefore corresponds to the cumulative distance traveled when the residual tread depth measurement is obtained.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to implement Ciaravola’s corrective residual tread depth measurement using the known direct tread wear sensor and real time tire sensing architecture described by Singh and to associate each operational residual tread depth measurement with the contemporaneous odometer reading representing the distance traveled by the vehicle when the measurement is obtained.
Ciaravola teaches using an actual tread depth measurement to correct or reset a tire wear estimate that is updated according to kilometers or miles driven. Ciaravola also identifies the known practice of associating each tread depth measurement with a corresponding distance value. Singh acknowledges direct tread wear measurement using sensors disposed in the tire tread and teaches the real time acquisition and electronic transmission of tire property measurements during vehicle operation.
A person of ordinary skill would have had reason to automate Ciaravola’s corrective measurement because Ciaravola expressly identifies missed or delayed tire shop measurements as a safety and reliability problem. Obtaining residual tread depth during operation would provide correction data without waiting for a driver initiated inspection, reduce accumulated estimation error between inspections, and allow the model to remain aligned with the tire’s current physical condition.
Associating the operational measurement with the odometer reading would identify the precise mileage at which the corrected tread depth applies. This would permit the processor to relate tread loss to the correct distance interval and continue the wear calculation from an accurate depth and mileage reference point.
The combination would use the known direct wear sensor, vehicle distance information, and processor based correction according to their established functions. The direct sensor would supply the same tread condition value used by Ciaravola’s correction process, and the odometer would supply the distance value already contemplated by Ciaravola’s distance based monitoring. A person of ordinary skill therefore would have had a reasonable expectation of success.
Regarding Claim 34,
The combination of Singh and Ciaravola establishes the method of Claim 20, which is the basis for Claim 34.
Disclosure by Singh
Singh teaches:
• wherein the telematics information of the vehicle include at least one of
• the vehicle usage, tire pressure, tractor load, region, country, longitudinal acceleration, lateral acceleration, speed, GPS coordinates, odometer, type of road, load, tire inflation pressure, gear shifts, engine RPMs, wheel speed, throttle/brake pedal position, tire temperature, external temperature, steering wheel angle.
See at least:
“Tire pressure as sensed by the sensor 24 may be used as a predictor 52.... The roughness of the road driven by the vehicle 10 may impact tire wear, and may thus be employed as a predictor 52.” Singh, paragraph [0039].
“A convenient indicator of weather effects 76 is an ambient temperature 78.” Singh, paragraph [0037].
The limitation requires at least one listed item. Singh expressly teaches tire pressure, type or condition of road, and external ambient temperature. Any one of these disclosures independently satisfies the alternative limitation.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to retain Singh’s tire-pressure, road-condition, or ambient-temperature telematics predictor in the combined model because each is a recognized contributor to tire wear and predictably improves estimation accuracy.
After combining Singh and Ciaravola, all limitations of Claim 34 are taught or rendered obvious.
Regarding Claim 35,
Disclosure by Singh
Singh discloses:
• A system for monitoring a tire wear rate of a vehicle,
See at least:
“The system 50 instead utilizes a tire wear estimation model that receives multiple input parameters to generate a high-accuracy estimation of the rate of tire wear.” Singh, paragraph [0029].
Singh expressly discloses a system that monitors and estimates tire wear rate.
• the system comprising one or more computing devices configured to:
See at least:
“Aspects of the tire wear estimation system 50 preferably are executed on a processor....” Singh, paragraph [0030].
“The second embodiment... may be implemented using a cloud-based server 110.” Singh, paragraph [0047].
Singh expressly discloses processor and server computing devices configured to perform the tire wear estimation operations.
• Train a data-driven mathematical tire wear model
See at least:
“The tire wear estimation system 50 generates the estimated wear rate 60 through model fitting.... For example, a Multiple Regression Linear (MLR) Model may be used.” Singh, paragraph [0040].
“The model fitting is done using stepwise regression.” Singh, paragraph [0041].
Singh expressly discloses computing devices that fit and thereby train a data-driven mathematical regression model.
• obtain technical data of at least one tire of a vehicle;
See at least:
“Each tire 12 preferably is equipped with a sensor or transducer 24... for... detecting certain real-time tire parameters, such as tire pressure and temperature.... [and] tire identification.” Singh, paragraph [0028].
Singh expressly discloses obtaining technical tire information.
• obtain technical data of the vehicle;
See at least:
“One vehicle effect 54 is a wheel position 56.” Singh, paragraph [0031].
“Another vehicle effect 54 is the vehicle drivetrain type 58.” Singh, paragraph [0032].
Singh expressly discloses obtaining vehicle configuration information.
• obtain telematics information of the vehicle; and
See at least:
“The route and driver effects 62 in turn include route severity 64 and driver severity 66.” Singh, paragraph [0034].
Singh expressly discloses obtaining vehicle-use and route information.
• calculate a tire wear rate
See at least:
“All of the predictors 52 are input into a model 86 to generate the estimated wear rate 60.” Singh, paragraph [0040].
Singh expressly discloses calculating tire wear rate 60.
• based at least in part on the obtained technical data of the at least one tire of the vehicle,
See at least:
“The dimensional tire effects 68 comprise one of the predictors 52 to be input into the tire wear estimation system 50.” Singh, paragraph [0036].
Singh expressly uses tire technical data as model inputs.
• the obtained technical data of the vehicle
See at least:
“The wheel position 56 is one of the predictors 52 to be input into the tire wear estimation system 50.” Singh, paragraph [0031].
Singh expressly uses vehicle technical data as model inputs.
• and the obtained vehicle telematics information of the vehicle
See at least:
“The route and driver effects 62 may be sensed by the sensor 24, may be included in the tire ID data, and/or may be stored in the above-described storage medium.” Singh, paragraph [0035].
Singh expressly uses route and driver information in the tire wear estimation.
• according to a data-driven mathematical tire wear model.
See at least:
“All of the predictors 52 are input into a model 86 to generate the estimated wear rate 60.... A Multiple Regression Linear (MLR) Model may be used.” Singh, paragraph [0040].
Singh expressly calculates the tire wear rate according to a fitted mathematical model.
Claim Limitations Not Explicitly Disclosed by Singh
Singh does not explicitly disclose:
• based on a plurality of calculated tire wear rates,
• wherein the data-driven mathematical model is continuously adapted
• to data obtained from a plurality of vehicles,
• wherein the obtained data comprise technical data of at least one tire of each of the vehicles,
• technical data of each of the vehicles,
• and telematics information of each of the vehicles;
Disclosure and Obviousness Based on Ciaravola
Ciaravola discloses or renders obvious the remaining limitations for the reasons set forth above for Claim 26:
“The ANN is trained... based on... a wear fleet database.” Ciaravola, paragraph [0117].
“Training the ANN... includ[es] applying... for each used tire, the recorded tire-usage-related quantities... as inputs and the respective second correction factor CF2 as output.” Ciaravola, paragraph [0123].
“The tire-usage-related statistical data... [include] vehicle and route parameters... vehicle wheel alignment and tire inflation pressure.” Ciaravola, paragraph [0118].
Ciaravola discloses fleet-based supervised training from respective tire, vehicle, and telematics records. Applying that architecture to Singh’s intended tire-wear-rate output would have made training on corresponding calculated tire wear rates obvious. Repeatedly updating the trainable model as additional fleet records became available would predictably maintain or improve accuracy under the stated interpretation of continuous adaptation.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to configure Singh’s processor and cloud-server system to train and repeatedly update its tire wear-rate model using Ciaravola’s fleet-based supervised-training records containing corresponding tire, vehicle, and telematics inputs.
The teachings use compatible input categories and perform complementary functions. The combination would predictably improve the accuracy, reliability, and representative coverage of Singh’s system across different tires, vehicles, routes, and operating conditions.
After combining Singh and Ciaravola, all limitations of Claim 35 are disclosed or rendered obvious under the stated interpretation of “continuously adapted.”
Regarding Claim 36,
The combination of Singh and Ciaravola establishes the system of Claim 35, which is the basis for Claim 36.
Claim Limitations Not Explicitly Disclosed by Singh
Singh does not explicitly disclose:
• estimate a residual tread depth and/or a remaining mileage of the tire and/or a remaining time before change
• according to a stored minimum tread-depth threshold,
• based on the calculated tire wear rate.
Disclosure and Obviousness Based on Ciaravola
Ciaravola discloses or renders obvious:
• estimate a residual tread depth and/or a remaining mileage of the tire and/or a remaining time before change
See at least:
“Estimating RTM... includes computing a remaining tread depth based on the tread wear value and an initial tread depth.” Ciaravola, paragraph [0062].
The limitation is written in the alternative through “and/or.” Ciaravola expressly discloses the residual-tread-depth alternative.
• according to a stored minimum tread-depth threshold,
See at least:
“Estimating RTM... includes detecting an approaching end-of-life condition... if the remaining tread depth reaches a predefined threshold.” Ciaravola, paragraph [0063].
Ciaravola expressly discloses comparison with a predefined minimum tread-depth threshold. Storing the processor-accessible threshold would have been the predictable implementation of the automated comparison.
• based on the calculated tire wear rate.
See at least:
“Estimating tread wear... includes computing... a tread wear value indicative of a reduction in tread depth due to a distance driven... and estimating RTM... includes computing a remaining tread depth based on the tread wear value and an initial tread depth.” Ciaravola, paragraph [0062].
“Tread wear and RTM estimation... is carried out every N kilometers/miles driven by the motor vehicle.” Ciaravola, paragraph [0064].
Applying Singh’s calculated wear rate over Ciaravola’s known distance interval would produce the accumulated tread loss used to determine remaining tread depth. This is the predictable use of the wear-rate output according to its ordinary function.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to configure Singh’s computing devices to apply the calculated tire wear rate over a driven-distance interval, determine Ciaravola’s remaining tread depth, and compare that depth with a predefined threshold stored in processor-accessible memory.
The modification predictably provides actionable tire-safety and replacement information.
After combining Singh and Ciaravola, all limitations of Claim 36 are disclosed or rendered obvious through the residual-tread-depth alternative.
Regarding Claim 37,
The combination of Singh and Ciaravola establishes the system of Claim 36, which is the basis for Claim 37.
Disclosure by Singh
Singh discloses:
• report at least one of the calculated tire wear rate, the estimated residual tread depth, the remaining mileage of the tire, and the remaining time before change according to a configured minimum tread depth
• to a control system.
See at least:
“Once the estimated wear rate 60 is generated, it is communicated from the tire wear estimation system 50 to the vehicle operating systems, such as braking and stability control systems, through the vehicle CAN bus.” Singh, paragraph [0042].
The limitation requires reporting at least one listed output. Singh expressly reports calculated tire wear rate 60 to vehicle braking or stability control systems. The calculated-tire-wear-rate alternative independently satisfies the limitation.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to report the tire wear rate generated by the combined wear and remaining-tread system to Singh’s vehicle operating control system so the tire-condition information can support braking, stability, maintenance, or notification functions.
After combining Singh and Ciaravola, all limitations of Claim 37 are disclosed or rendered obvious.
Regarding Claim 38,
The combination of Singh and Ciaravola establishes the system of Claim 37, which is the basis for Claim 38.
Disclosure by Singh
Singh discloses:
• wherein the control system is arranged in the vehicle.
See at least:
“Once the estimated wear rate 60 is generated, it is communicated... to the vehicle operating systems, such as braking and stability control systems, through the vehicle CAN bus.” Singh, paragraph [0042].
Singh expressly discloses vehicle braking and stability control systems connected through the vehicle CAN bus. Those operating control systems are arranged in the vehicle.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to provide the tire-wear information generated by the combined system to Singh’s in-vehicle braking or stability control system.
This arrangement preserves Singh’s expressly disclosed CAN-bus architecture and provides immediate local use of the tire-condition information. After combining Singh and Ciaravola, all limitations of Claim 38 are disclosed or rendered obvious.
Regarding Claim 39,
The combination of Singh and Ciaravola establishes the system of Claim 37, which is the basis for Claim 39.
Claim Limitations Not Explicitly Disclosed by Singh
Singh does not explicitly disclose the following additional limitations:
• wherein the control system is arranged outside the vehicle,
• and configured to collect,
• from a plurality of vehicles,
• at least the calculated tire wear rate, the estimated residual tread depth, the remaining mileage of the tire, and the remaining time before change
• according to a stored minimum tread-depth threshold.
Disclosure and Obviousness Based on Ciaravola
Ciaravola discloses or renders obvious:
• wherein the control system is arranged outside the vehicle,
See at least:
“The processing device/system 32 is implemented/carried out by means of a cloud computing system 32A that is wirelessly and remotely connected to the acquisition device 31.” Ciaravola, paragraph [0077].
Ciaravola expressly discloses implementing the tire-wear processing and control functionality through a remotely connected cloud computing system. The cloud computing system is physically arranged outside the vehicle and receives vehicle-associated information through the acquisition device.
• and configured to collect,
See at least:
“The processing device/system 32 is implemented/carried out by means of a cloud computing system 32A that is wirelessly and remotely connected to the acquisition device 31.” Ciaravola, paragraph [0077].
Ciaravola at least implicitly discloses that the cloud computing system is configured to collect vehicle-associated tire-wear information. The cloud system is remotely connected to the vehicle acquisition device and requires the information received through that connection to perform the disclosed tread-wear monitoring and remaining-tread-material processing.
• from a plurality of vehicles,
See at least:
“The cloud computing system 32A may be advantageously used to provide many motor vehicles 4 and, hence, many users 5 with a tread wear monitoring service.” Ciaravola, paragraph [0080].
Ciaravola expressly discloses a common cloud computing system providing the monitoring service to many vehicles. In combination with the remote connection of paragraph [0077], the cloud system receives and aggregates tire-wear monitoring information associated with a plurality of vehicles.
• at least the calculated tire wear rate, the estimated residual tread depth, the remaining mileage of the tire, and the remaining time before change
See at least:
“Estimating tread wear . . . includes computing, based on the second frictional-energy-related quantities and the calibrated TWM, a tread wear value indicative of a reduction in tread depth due to a distance driven by the motor vehicle, and estimating RTM . . . includes computing a remaining tread depth based on the tread wear value and an initial tread depth.” Ciaravola, paragraph [0062].
“Determining, by the one or more processors, an estimated time or an estimated distance at which the projected tread depth is expected to equal or pass a tread depth threshold based at least in part on the model.” Ciaravola, paragraph [0009].
“Providing, by the one or more processors, the estimated time or the estimated distance to a notification system.” Ciaravola, paragraph [0010].
The parent system established by Singh and Ciaravola calculates and reports the tire wear rate. Ciaravola expressly calculates remaining tread depth from calculated tread wear and initial tread depth. Ciaravola further acknowledges as known in the tire tread-life art determining and reporting an estimated distance or estimated time at which projected tread depth reaches a tread-depth threshold.
Ciaravola does not expressly state that one cloud record contains all four claimed values. It would nevertheless have been obvious to a person of ordinary skill to include the calculated tire wear rate, remaining tread depth, remaining distance, and remaining time in the tire-condition record collected for each vehicle. The values concern the same tire, are derived from the same wear, depth, threshold, distance, and usage information, and collectively describe the tire’s present condition and predicted replacement point.
The remaining mileage is predictably determined from the difference between the current remaining tread depth and the minimum tread-depth threshold, together with the tread-loss rate per unit distance. The remaining time is predictably determined from the remaining mileage and the vehicle’s distance traveled per unit time, available from the vehicle-usage or telematics history of the inherited system. These calculations apply known quantities using their standard mathematical relationships and do not require a different tire-wear model or an additional sensing architecture.
• according to a stored minimum tread-depth threshold.
See at least:
“Estimating RTM . . . conveniently includes detecting an approaching end-of-life condition for the tires of the motor vehicle if the remaining tread depth reaches a predefined threshold.” Ciaravola, paragraph [0063].
Ciaravola expressly discloses evaluating remaining tread depth according to a predefined minimum tread-depth threshold. Although Ciaravola does not expressly use the word “stored,” retaining the predefined threshold in memory accessible to the cloud processing system would have been an obvious implementation. The processor must have continuing access to the threshold to perform the disclosed automated comparison for each monitored tire and vehicle. Storing the threshold would permit consistent and repeatable threshold evaluation without requiring the value to be reentered for each calculation.
Motivation to Combine Singh and Ciaravola
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh and Ciaravola before them, to implement the control system receiving the reported tire-wear information of Claim 37 as Ciaravola’s remote cloud computing system and configure the cloud system to collect, from a plurality of vehicles, the calculated tire wear rate, estimated residual tread depth, remaining mileage, and remaining time before tire change according to a stored minimum tread-depth threshold.
Singh discloses calculating and electronically reporting tire wear rate. Ciaravola discloses calculating remaining tread depth, comparing that depth with a predefined end-of-life threshold, remotely processing vehicle tire-wear information in a cloud computing system, and providing the monitoring service to many vehicles. Ciaravola further acknowledges the known determination and reporting of estimated time or distance to a tread-depth threshold.
The teachings are technically compatible because the four claimed values are related outputs generated from the same tire-wear, remaining-depth, threshold, distance, and vehicle-usage information. Configuring the cloud system to collect the complete set would require only including the related output values in the vehicle tire-condition record transmitted or generated through the existing cloud architecture.
A person of ordinary skill would have been motivated to collect the complete set to provide both the current tire condition and the predicted distance and time remaining before replacement. Centralized collection would predictably improve fleet maintenance scheduling, permit comparison and prioritization among vehicles, support timely tire replacement, and reduce the need for separate monitoring systems or calculations. The modification would use the calculations, communications, and cloud-processing components of Singh and Ciaravola according to their established functions and would have had a reasonable expectation of success.
Claims 23 is rejected under 35 U.S.C. 103 as being unpatentable over Singh, in view of Ciaravola, in view of Khosla (US 6718259 B1), and in view of Ross (WO 2007121517 A1).
Regarding Claim 23,
The combination of Singh and Ciaravola establishes the method of Claim 22, including the plurality of pre-stored algorithms for calculating tire wear rate and the tire wear rate calculated based at least in part on the obtained in-operational tire-property measurement, which is the basis for Claim 23.
Disclosure by Singh
Singh teaches:
• based at least in part on the obtained data of the at least one in-operational measurement
See at least:
“On the server 110, the predictors 52 are input into the model 86 for estimation of the wear rate 60, which is integrated with the real-time predictors 102 to yield the predicted wear state 104.” Singh, paragraph [0047].
“The second embodiment of the wear estimation system 100 provides additional refinement and accuracy, as it adds the predictors 102 of real-time measurements of sensed conditions of the tire 12 to the estimation of the wear rate 60.” Singh, paragraph [0048].
Singh expressly teaches using data obtained from real-time sensed tire conditions in the tire-wear determination. In the method established for Claim 22, that measurement-based tire wear rate supplies the benchmark against which the respective candidate algorithm outputs are compared.
• of the at least one property of the at least one tire of the vehicle.
See at least:
“Such real-time measurements include changes in the physical attributes or characteristics of the tire, such as the stiffness of the tread 16.” Singh, paragraph [0045].
Singh expressly teaches that the in-operational measurement concerns a physical property of tire 12 supporting vehicle 10, including tread stiffness.
Claim Limitations Not Explicitly Taught by Singh
Singh does not explicitly teach:
• wherein selecting one of a plurality of pre-stored algorithms for calculating a tire wear rate comprises:
• running a plurality of algorithms for calculating the tire wear rate;
• choosing an algorithm of the plurality of algorithms
• which yields a calculated value for the tire wear rate
• that is closest to the tire wear rate calculated
Although the combination of Singh and Ciaravola establishes the plurality of pre-stored tire-wear algorithms, that combination does not explicitly teach executing the plurality and selecting one according to the claimed closest-value criterion.
Disclosure and Obviousness Based on Khosla
Khosla renders obvious:
• wherein selecting one of a plurality of pre-stored algorithms for calculating a tire wear rate comprises:
See at least:
“The invention dynamically selects the Kalman filter, from the filter bank, which is best suited to existing road conditions.” Khosla, col. 8, ll. 6-8.
Khosla expressly teaches a multi-model selection procedure in which a processor evaluates a bank of model-specific filters and dynamically selects the candidate best suited to the current measured condition. Applying that selection procedure to the plurality of pre-stored tire-wear algorithms established by Singh and Ciaravola would provide the general selection framework recited by Claim 23.
• running a plurality of algorithms for calculating the tire wear rate;
See at least:
“The adaptive Kalman filter protocol accordingly implements 3 parallel Kalman filters (one for each model) operating on the same measurement sequence simultaneously.” Khosla, col. 9, ll. 43-46.
Khosla expressly teaches simultaneously executing a respective computational filter for each candidate model using the same measurement sequence. Applied to Ciaravola’s plurality of pre-stored tire-wear algorithms, each algorithm would process the same current tire, vehicle, telematics, and in-operational measurement data and generate a respective candidate tire-wear-rate value.
• choosing an algorithm of the plurality of algorithms
See at least:
“The invention dynamically selects the Kalman filter, from the filter bank, which is best suited to existing road conditions.” Khosla, col 8, ll. 6-8.
“One of the models still ‘best’ describes the upcoming road geometry amongst all three models (i.e., has the highest weight factor).” Khosla, page 10.
Khosla expressly identifies and dynamically selects the candidate filter that best corresponds to the current measured operating condition. Khosla also uses the weighted candidate outputs to form a fused final estimate. That separate fusion step does not negate Khosla’s express identification and selection of the best-suited candidate. Claim 23 requires choosing an algorithm but does not require that the remaining outputs be discarded or that only the chosen output be used thereafter.
• which yields a calculated value for the tire wear rate
See at least:
“The outputs of each Kalman filter after the measurement update step, utilizing the latest available measurement vector Y at time k, are the estimated state vector X and error covariance matrix P.” Khosla, col. 9, ll. 54-57.
Khosla expressly teaches that each independently executed candidate filter produces its own calculated estimate. Applied to the plurality of tire-wear algorithms established by Singh and Ciaravola, each candidate algorithm would yield a respective calculated value for the common output quantity, tire wear rate.
Motivation to Combine Singh, Ciaravola, and Khosla
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh, Ciaravola, and Khosla before them, to execute Ciaravola’s plurality of pre-stored tire-wear algorithms using Khosla’s parallel multi-model architecture and identify the algorithm best corresponding to Singh’s current in-operational tire measurement.
Singh teaches tire-wear determination using current tire, vehicle, telematics, and real-time tire-condition data. Ciaravola supplies multiple previously prepared tire-wear models. Khosla teaches simultaneously executing candidate models on the same measurement data and identifying the candidate best suited to the observed condition.
Applying Khosla’s architecture would permit direct evaluation of the tire-wear algorithms under identical operating conditions and would predictably improve selection accuracy where different algorithms provide different performance for different tires, vehicle configurations, mounting positions, and operating conditions.
Claim Limitations Not Explicitly Disclosed by the Combination of Singh, Ciaravola, and Khosla
After combining the teachings of Singh, Ciaravola, and Khosla, the following limitation is not explicitly disclosed:
• that is closest to the tire wear rate calculated
Khosla identifies the candidate model having the highest posterior probability or weight but does not expressly require choosing the candidate whose scalar tire-wear-rate output has the smallest numerical difference from the measurement-based tire wear rate.
Disclosure and Obviousness Based on Ross
Ross teaches or renders obvious:
• that is closest to the tire wear rate calculated
See at least:
“The idea is to hypothesise r system models, one for each of the r values, to estimate state of each model using a Kalman filter, and to determine the probability that each model is correct from the statistical properties of the innovation sequence generated by the filter for that model.” Ross, paragraph [0064].
“Over time the most accurate model is assigned the largest probability while the less accurate models are assigned lower probabilities.” Ross, paragraph [0064].
“We are now looking to determine the most likely model Mj, given Z(ti).” Ross, paragraph [0065].
“The residual, residual covariance, and the prior model probability are then used to determine the current model probability.” Ross, paragraph [0074].
Ross expressly teaches evaluating each candidate model from the discrepancy between its prediction and the measurement information and using the resulting innovation or residual, residual covariance, and prior model probability to identify the candidate most consistent with the observations.
Ross does not expressly teach selecting the candidate having the smallest raw numerical difference between two scalar tire-wear-rate values. Ross’s maximum-posterior-probability determination is not necessarily identical to selecting the smallest raw residual because Ross also considers residual covariance and prior probability. The precise closest-value rule is therefore not attributed to Ross as an express disclosure.
The Singh and Ciaravola combination supplies the measurement-based tire wear rate serving as the benchmark, while Khosla supplies the respective candidate tire-wear-rate outputs. Where each candidate algorithm and the benchmark produce the same scalar quantity, a person of ordinary skill would have found it obvious to define the residual for candidate algorithm j as the difference between its calculated tire wear rate and the measurement-based tire wear rate and to choose the algorithm having the smallest absolute residual or squared residual.
Absolute-residual and squared-residual comparison produce the same ordering for scalar residual magnitudes. Using either criterion would be a direct minimum-error implementation of Ross’s disclosed objective of identifying the candidate model most consistent with the measurement information. It would also provide a computationally simple selection rule when the design objective is to choose one tire-wear algorithm according to numerical agreement with the current measurement-based benchmark.
Motivation to Combine Singh, Ciaravola, Khosla, and Ross
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh, Ciaravola, Khosla, and Ross before them, to run Ciaravola’s plurality of pre-stored tire-wear algorithms in parallel using Khosla’s multi-model architecture, compare each candidate tire-wear-rate value with the tire wear rate calculated from Singh’s in-operational tire-property measurement, and choose the algorithm having the smallest absolute or squared difference from that measurement-based value.
Khosla teaches parallel execution and identification of the candidate model best matching current vehicle measurements. Ross teaches evaluating candidate models through measurement residuals to identify the model most consistent with the observed data. Because the candidate algorithms and the measurement-based benchmark produce the same scalar tire-wear-rate quantity, selecting the minimum residual would have been a direct and predictable implementation of those model-evaluation teachings.
A person of ordinary skill would have been motivated to use the minimum-difference criterion to reduce disagreement between the selected algorithm and the tire’s current measured condition, thereby improving algorithm-selection accuracy. The modification would require only conventional subtraction, magnitude comparison, and minimum-value selection and would have had a reasonable expectation of success.
Khosla concerns automotive multi-model estimation. Ross is reasonably pertinent to the particular problem addressed by Claim 23 because Ross teaches determining, from operational measurements, which of multiple candidate models most accurately represents the physical system being monitored.
Claims 24 is rejected under 35 U.S.C. 103 as being unpatentable over Singh, in view of Ciaravola, and in view of Khosla.
Regarding Claim 24,
The combination of Singh and Ciaravola establishes the method of Claim 21, which is the basis for Claim 24.
Disclosure by Singh
Singh teaches:
• wherein calculating the tire wear rate further comprises:
See at least:
“The second embodiment of the wear estimation system 100 incorporates the first embodiment of the wear estimation system 50 as described above, and adds certain real-time predictors 102.” Singh, paragraph [0044].
Singh expressly teaches further processing within the tire wear calculation by adding real-time tire-condition predictors to the underlying wear estimation method.
• calculating tire wear rate
See at least:
“All of the predictors 52 are input into a model 86 to generate the estimated wear rate 60 for a given tire 12.” Singh, paragraph [0040].
Singh expressly teaches calculating tire wear rate 60 by processing predictor data with mathematical model 86.
Claim Limitations Not Explicitly Taught by Singh
Singh does not explicitly teach:
• according to a self-tuning mathematical model
• and wherein the self-tuning model is tuned
• based on data
• of the at least one in-operational measurement
• of the at least one tire of the vehicle.
Singh teaches fitting a tire wear model and integrating real-time tire-condition measurements with the resulting wear-rate estimate. Singh does not expressly teach automatically tuning parameters of the tire wear model based on the in-operational tire measurement.
Disclosure by Ciaravola
Ciaravola teaches using operational measurements as inputs to a previously calibrated tire wear model:
See at least:
“In the tread wear monitoring step 2, the given parameters . . . are known from the tread wear model calibration step 1, and also the driving-related quantities . . . are ‘known’ from the measurements on board the vehicle.” Ciaravola, paragraph [0060].
Ciaravola expressly teaches applying onboard operational measurements to a calibrated tire wear model. Ciaravola does not teach that those operational measurements alter the calibrated model parameters. Rather, the parameters used during monitoring remain those previously determined during calibration.
Claim Limitations Not Explicitly Disclosed by the Combination of Singh and Ciaravola
After combining Singh and Ciaravola, the following limitations remain not explicitly disclosed:
• according to a self-tuning mathematical model
• and wherein the self-tuning model is tuned
• based on data
• of the at least one in-operational measurement
• of the at least one tire of the vehicle.
Disclosure and Obviousness Based on Khosla
Khosla teaches or renders obvious:
• according to a self-tuning mathematical model
See at least:
“Another aspect of the invention adaptively tunes process model parameters such that the filter responds to road geometry changes quickly during sharp changes and provides stable and low noise estimates otherwise.” Khosla, col. 4, ll. 43-46)
Khosla expressly teaches a processor-executed mathematical estimator that adaptively tunes process-model parameters. The estimator is self-tuning because its operative model configuration is automatically adjusted during operation rather than remaining fixed. This is a direct correspondence to a self-tuning mathematical model, not merely an update of the estimated physical state.
• and wherein the self-tuning model is tuned
See at least:
“The adaptive Kalman filter protocol accordingly implements 3 parallel Kalman filters (one for each model) operating on the same measurement sequence simultaneously.” Khosla, col. 9, ll. 43-46).
“Since the weight factors are updated after each measurement step, the filter adapts to the changing road geometry and is self-learning in that respect.” Khosla, col. 10, ll. 7-10).
Khosla expressly teaches automatic tuning through a bank of candidate process models having different process-model parameter settings and through measurement-driven updating of the relative model weights. Updating the weights changes which parameterized model predominates in the estimator and thereby adjusts the effective mathematical model applied to the current condition. This measurement-driven adjustment implements Khosla’s expressly disclosed adaptive tuning of process-model parameters.
• based on data
See at least:
“The outputs of each Kalman filter after the measurement update step, utilizing the latest available measurement vector Y at time k, are the estimated state vector X and error covariance matrix P.” Khosla, col. 9, ll. 54-57).
Khosla expressly teaches that the adaptive operation uses the latest available measurement data. The measurement-derived filter outputs are used to calculate the model probabilities and weights that determine the effective model configuration.
• of the at least one in-operational measurement
See at least:
“A sensor measurement sequence y₁, y₂, . . . yₖ is provided to a bank of Kalman filters, all of which are identical, except for the road model used in each.” Khosla, page col. 46-49).
“Since the weight factors are updated after each measurement step, the filter adapts to the changing road geometry and is self-learning in that respect.” Khosla, page 10.
Khosla expressly teaches tuning from successive sensor measurements obtained while the vehicle operates. Each new operational measurement updates the probabilities assigned to the differently parameterized models and thereby tunes the effective model to the current measured condition.
• of the at least one tire of the vehicle.
See at least:
“Such real-time measurements include changes in the physical attributes or characteristics of the tire, such as the stiffness of the tread 16.” Singh, paragraph [0045].
“Sensors on the tire 12 and/or the vehicle 10 are a first source 114 that measure real-time predictors 102, which are wirelessly transmitted . . . to the server 110.” Singh, paragraph [0047].
Singh expressly teaches that the in-operational measurement data concern a physical property of tire 12 of vehicle 10. Khosla supplies the automatic measurement-driven tuning technique. It would have been obvious to use Singh’s real-time tire-property measurement as the operational measurement data that tune the mathematical tire wear model.
Motivation to Combine Singh, Ciaravola, and Khosla
Therefore, given the teachings as a whole, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, having Singh, Ciaravola, and Khosla before them, to configure the mathematical tire wear model established by Singh and Ciaravola to adaptively tune its process-model parameters or effective parameterized-model weighting based on Singh’s in-operational measurement of the monitored tire.
Singh teaches calculating tire wear rate and using real-time measurements of sensed tire properties to improve the wear determination. Ciaravola confirms that onboard tire and vehicle measurements are suitable inputs to a calibrated tire wear model. Khosla teaches automatically adapting an automotive mathematical estimator by tuning process-model parameters and updating the relative weights of differently parameterized models after each operational measurement.
A person of ordinary skill would have applied Khosla’s measurement-driven parameter adaptation to Singh’s tire wear model so that the same current tire-condition measurements used in the wear determination also adjust the operative model to the tire’s changing physical condition. The modification would predictably improve responsiveness and estimation accuracy as the tire and its operating conditions change.
Khosla is reasonably pertinent to the problem addressed by Claim 24 because it concerns maintaining the accuracy of a vehicle-related mathematical estimator by automatically adapting its process model in response to current operational sensor measurements. Applying that established adaptive-estimation technique to the tire wear model would have required only conventional implementation of measurement updates, model-probability calculations, and parameterized-model selection or weighting and would have had a reasonable expectation of success.
Response to Arguments
Rejections Under 35 U.S.C. § 112
Applicant’s amendments and arguments concerning the rejections under 35 U.S.C. § 112(a), 35 U.S.C. § 112(b), and 35 U.S.C. § 112(d) have been fully considered and are persuasive.
The amendments and accompanying remarks adequately resolve the previously identified issues concerning written description and enablement under 35 U.S.C. § 112(a), definiteness under 35 U.S.C. § 112(b), and claim dependency under 35 U.S.C. § 112(d).
Accordingly, all rejections under 35 U.S.C. § 112 are withdrawn.
Rejections Under 35 U.S.C. § 103
Applicant’s arguments and amendments have been fully considered but are not persuasive.
As an initial matter, the rejection does not rely on Singh’s lookup table as constituting the claimed training architecture, nor does it rely on Singh alone for the complete fleet based training arrangement. Singh teaches fitting a supervised mathematical model whose output is tire wear rate. Ciaravola supplies the complementary supervised training arrangement using a wear fleet database containing respective tire usage records and corresponding calculated outputs. The rejection rests on those collective teachings.
Claim 20
Applicant argues that the cited references fail to teach or suggest:
“training a data-driven mathematical tire wear model;”
“obtaining technical data of at least one tire of a vehicle;” and
“calculating a tire wear rate based at least in part on the obtained technical data of the at least one tire of the vehicle.”
Examiner disagrees for the following reasons.
Training a Data Driven Mathematical Tire Wear Model
Applicant contends that Singh merely retrieves fixed parameters from a lookup table and therefore does not teach training a data driven mathematical tire wear model. This argument does not account for Singh’s separate model fitting disclosure.
Singh teaches:
“All of the predictors 52 are input into a model 86 to generate the estimated wear rate 60 for a given tire 12. The tire wear estimation system 50 generates the estimated wear rate 60 through model fitting, and any appropriate model may be selected. For example, a Multiple Regression Linear (MLR) Model may be used. By way of background, linear regression is a simple approach to supervised learning.” Singh, paragraph [0040].
Singh further teaches:
“The model fitting is done using stepwise regression, in turn using a forward selection technique, with p-value criteria.” Singh, paragraph [0041].
Thus, Singh does not merely retrieve a tire wear rate from a fixed lookup relationship. Singh fits a supervised mathematical model and statistically selects predictors for inclusion in that model.
Model fitting through stepwise regression constitutes training because it determines, from data, the operative mathematical relationship between the input predictors and the output tire wear rate. The forward selection procedure evaluates candidate variables under defined statistical criteria and incorporates variables that meaningfully contribute to the model. Singh, paragraphs [0040] and [0041].
Applicant’s reliance on Singh’s lookup table or database disclosure does not establish otherwise. The lookup table or database supplies certain stored predictor information. Singh separately fits model 86 using supervised regression. The source of some model inputs does not convert the trained mathematical model into a mere retrieval tool.
Applicant also relies on Application paragraph [0009] and characterizes the claimed invention as requiring big data, machine learning, large data volumes, supervised experiments, and evaluation of multiple algorithms. Claim 20, however, does not require:
a neural network;
a particular machine learning architecture;
a specified volume of data;
a particular experimental protocol; or
evaluation of a plurality of candidate algorithms.
Those unrecited features cannot be imported from the specification into Claim 20. Singh’s supervised regression fitting falls within the broad claim language “training a data-driven mathematical tire wear model.”
Claim 20 does recite that the model is continuously adapted. The claim does not, however, require uninterrupted, record by record, real time online learning, nor does it specify an adaptation frequency or retraining protocol. Under the broadest reasonable interpretation consistent with the specification, the limitation encompasses repeated or recurring adaptation as additional fleet records become available.
Ciaravola supplies the relevant fleet based supervised training arrangement. Ciaravola teaches:
“Preferably, the ANN is trained . . . based on a given database, e.g., a wear fleet database, including tire-usage-related statistical data and corresponding RTM-related statistical data.” Ciaravola, paragraph [0117].
Ciaravola further teaches:
“Training the ANN . . . includ[es] applying to the ANN, for each used tire, the recorded tire-usage-related quantities associated with said used tire as inputs and the respective second correction factor CF2 as output.” Ciaravola, paragraph [0123].
Ciaravola therefore expressly teaches supervised machine learning using multiple respective tire records from a fleet database. Singh supplies the intended dependent output of the combined model, namely tire wear rate. It would have been obvious to apply Ciaravola’s fleet based supervised training arrangement to Singh’s tire wear rate model and to pair each fleet input record with its corresponding calculated tire wear rate because supervised fitting uses labeled outputs representing the quantity the model is intended to estimate.
A person of ordinary skill would have had a reasonable expectation of success. Both references use supervised mathematical models and overlapping categories of tire, vehicle, route, and operating information. Applying Ciaravola’s fleet training structure would not change Singh’s model inputs, intended output, or underlying principle of operation. It would predictably improve representative coverage and estimation reliability across different tires, vehicles, routes, and operating conditions.
Applicant’s asserted distinction between a lookup table and a trained model therefore does not correspond to the applied disclosure. Singh separately teaches supervised model fitting, and Ciaravola teaches the complementary fleet based supervised training arrangement.
Obtaining Technical Data of at Least One Tire of a Vehicle
Applicant next argues that Application paragraph [0041] defines “technical data of at least one tire” exclusively as tire manufacturer, tire model, tire pattern, tire specification, tire size, tire mounting position, retread information, and batch number.
That position is not supported by the claim language or the cited specification passage.
An applicant acts as its own lexicographer only when the specification clearly sets forth a special definition and demonstrates a deliberate intent to depart from the term’s ordinary meaning. The quoted passage states that the technical data include the listed examples. The open ended term “include” does not clearly make the list exclusive or exclude other technical information describing the tire.
Claim 20 also does not recite the asserted list. In contrast, dependent Claim 32 expressly adds that the technical tire data include at least one of the listed categories. The claim structure further supports reading Claim 20 more broadly than the specific examples subsequently added by Claim 32.
Even assuming, for the sake of argument, that the listed categories were required by Claim 20, Singh teaches several of those categories and does not disclose only data representing a tire’s current operating state.
Singh teaches tire identification data:
“The sensor 24 preferably also includes a tire identification for each specific tire 12, and transmits measured parameters and tire ID data to a remote processor.” Singh, paragraph [0028].
Singh teaches tire dimensional information:
“The dimensional tire effects 68 in turn include the tire rim size 70, the tire width 72, and the tire outer diameter 74.” Singh, paragraph [0036].
Singh states that those dimensional tire effects are predictors input into the tire wear estimation system and may be included in tire identification data or stored in the processor accessible storage medium. Singh, paragraph [0036].
Singh also teaches tire construction and identification information:
“The physical tire effects 80 in turn include the compound used for the tread 16, which may be indicated by the treadcap code 82, and the tread structure, which may be indicated by the tire mold code 84.” Singh, paragraph [0038].
Singh states that those physical tire effects are predictors input into the tire wear estimation system and may be included in tire identification data or stored in the storage medium. Singh, paragraph [0038].
Further, Singh identifies wheel position 56 as a predictor used by the tire wear estimation system. Singh, paragraph [0031]. Wheel position corresponds to the mounting position of the monitored tire on the vehicle.
Accordingly, Singh teaches, among other things:
tire dimensional or size information;
tire mounting position;
tread compound information;
tread structure information;
treadcap and mold codes; and
tire identification information.
These are technical and identifying characteristics of the tire. They are not limited to pressure, temperature, or another transient driving condition. Applicant’s focus on the passage concerning real time pressure and temperature does not account for Singh’s additional disclosures in paragraphs [0031], [0036], and [0038].
Applicant’s asserted lexicographic interpretation therefore does not distinguish Claim 20 from Singh. Even under Applicant’s proposed interpretation, Singh teaches qualifying technical tire data.
Calculating the Tire Wear Rate Based on the Technical Tire Data
Applicant’s third argument depends on the premise that Singh does not obtain technical data of the tire. As explained above, that premise is incorrect.
Singh expressly identifies the dimensional tire effects as predictors input into the tire wear estimation system. Singh, paragraph [0036]. Singh likewise identifies tread compound and tread structure as physical tire effect predictors input into the system. Singh, paragraph [0038].
Singh then states:
“All of the predictors 52 are input into a model 86 to generate the estimated wear rate 60 for a given tire 12.” Singh, paragraph [0040].
Thus, Singh expressly teaches calculating tire wear rate 60 based at least in part on technical data of the tire. The technical tire data are identified as model predictors, and the predictors are supplied to model 86 to generate the estimated wear rate.
Singh’s remote server embodiment confirms the same relationship:
“The tire sensor 24 may also transmit certain selected predictors 52, such as the ambient temperature 78 and tire identification data, to the server 110. . . . Still other selected predictors 52 for estimation of the wear rate 60, such as tread compound data 82 and tread structure data 84, may be sent from a third source 118 to the server 110. On the server 110, the predictors 52 are input into the model 86 for estimation of the wear rate 60.” Singh, paragraph [0047].
Singh therefore teaches both obtaining and using technical tire data in the calculation of tire wear rate.
Claim 26
Applicant traverses independent Claim 26 for substantially the same reasons asserted for Claim 20. Those arguments are unpersuasive for the reasons stated above.
Claim 26 additionally recites transmitting technical tire data, vehicle data, and telematics information and obtaining the resulting calculated tire wear rate. Singh teaches transmitting tire identification data, tread compound data, tread structure data, real time predictors, location data, weather data, and road condition data to server 110. Singh, paragraph [0047].
Singh further teaches that the server inputs the transmitted predictors into model 86 to calculate wear rate 60 and that the predicted wear state containing wear rate 60 is transmitted to a receiving device. Singh, paragraphs [0046] and [0047].
Ciaravola supplies the complementary fleet based supervised training arrangement discussed above. Applicant has not identified a separate technical deficiency in the transmitting and receiving architecture of Claim 26.
The rejection of Claim 26 is therefore maintained.
Claim 35
Applicant traverses independent system Claim 35 for substantially the same reasons asserted for Claim 20. Those arguments likewise do not overcome the rejection.
Singh teaches that aspects of tire wear estimation system 50 are executed on a processor accessible through the vehicle CAN bus. Singh, paragraph [0030]. Singh further teaches implementation using cloud based server 110, which receives the tire, vehicle, and operational predictors and calculates estimated wear rate 60 using model 86. Singh, paragraph [0047].
Singh therefore discloses one or more computing devices configured to obtain the recited categories of data and calculate tire wear rate according to a data driven mathematical model.
Singh’s supervised model fitting disclosures in paragraphs [0040] and [0041], together with Ciaravola’s fleet based ANN training disclosures in paragraphs [0117] through [0123], teach or render obvious the model training functionality recited by Claim 35.
The rejection of Claim 35 is therefore maintained.
Dependent Claims
Applicant states generally that the dependent claims are patentable for the same reasons asserted for independent Claims 20, 26, and 35. Because the asserted deficiencies in the independent claims have not been established, those same arguments do not overcome the dependent claim rejections.
Applicant has not presented separate substantive arguments identifying a deficiency in the applied teachings concerning the additional limitations of Claims 21 through 25, 27 through 34, or 36 through 39. Applicant also has not substantively addressed the additional teachings of Khosla and Ross applied to Claim 23, Khosla applied to Claim 24, or Stalnaker applied to Claim 25.
Accordingly, the dependent claim rejections are maintained for the reasons stated in the rejections.
Prima Facie Case of Obviousness
The Office agrees that the Examiner bears the initial burden of establishing a prima facie case of obviousness. See In re Oetiker, 977 F.2d 1443, 1445, 24 USPQ2d 1443, 1444 (Fed. Cir. 1992); In re Piasecki, 745 F.2d 1468, 1472, 223 USPQ 785, 788 (Fed. Cir. 1984).
That burden has been met.
The rejection identifies the teachings corresponding to the claimed subject matter, identifies the features supplied through the combination, and explains why a person of ordinary skill would have applied Ciaravola’s fleet based supervised training arrangement to Singh’s tire wear rate model. The proposed combination uses technically compatible data and modeling techniques, retains the established function of each reference, and would have had a reasonable expectation of improving model accuracy and representative coverage.
Applicant’s arguments focus principally on Singh’s lookup table disclosure and do not address Singh’s separate supervised model fitting disclosures in paragraphs [0040] and [0041] or the collective teachings of Singh and Ciaravola.
Where a rejection is based on a combination of references, nonobviousness cannot be established by attacking the references individually without addressing what their combined teachings would have suggested to one of ordinary skill. See In re Keller, 642 F.2d 413, 425, 208 USPQ 871, 881 (CCPA 1981).
The cases cited by Applicant do not require withdrawal where, as here, the complete claim has been compared with the collective teachings of the references and the rejection provides articulated reasoning with rational underpinning for the proposed combination.
Examiner Conclusion
Applicant’s arguments have been fully considered but do not overcome the rejections.
Singh teaches fitting a supervised mathematical tire wear model, obtaining technical tire data, and calculating tire wear rate using those technical tire data. Ciaravola teaches the complementary use of supervised machine learning with a wear fleet database containing respective tire usage records and calculated outputs.
Accordingly:
the rejections of Claims 20 through 22, 26 through 28, and 30 through 39 under 35 U.S.C. § 103 over Singh in view of Ciaravola are maintained;
the rejection of Claim 23 over Singh and Ciaravola, further in view of Khosla and Ross, is maintained;
the rejection of Claim 24 over Singh and Ciaravola, further in view of Khosla, is maintained.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to OLUWABUSAYO ADEBANJO AWORUNSE whose telephone number is (571)272-4311. The examiner can normally be reached M - F (8:30AM - 5PM).
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, Jelani Smith can be reached at (571) 270-3969. 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.
/OLUWABUSAYO ADEBANJO AWORUNSE/Examiner, Art Unit 3662
/JELANI A SMITH/Supervisory Patent Examiner, Art Unit 3662
Prosecution Insights