DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Introductory Remarks
In response to communications filed on 10 August 2026, claims 1 and 4-8 are amended per Applicant's request. Claims 10-11 are cancelled. No claims were withdrawn. No new claims were added. Therefore, claims 1-9 and 12-14 are presently pending in the application, of which claim 1 is presented in independent form.
The previously raised 101 rejection of the pending claims is withdrawn in view of the amendments to the claims.
The previously raised 103 rejection of the pending claims is withdrawn in view of the amendments to the claims. A new ground(s) of rejection has been issued.
Response to Arguments
Applicant’s arguments filed 10 August 2026 with respect to the rejection of the claims under 35 U.S.C. 101 (see Remarks, p. 7-10) have been fully considered. However, for the Office’s own reasons, the claims have been found to be patent eligible, and the 101 rejection has been withdrawn.
Applicant’s arguments filed 10 August 2026 with respect to the rejection of the claims under 35 U.S.C. 103 have been fully considered but are not persuasive.
Applicant argues the prior art do not disclose “normalizing input data to a reference signal” (see Remarks, p. 11-14). This is unpersuasive, as Hawkins was found to disclose this feature. Applicant’s arguments only address one of multiple paragraph citations of Hawkins instead of looking at the rejection in its totality. For example, Hawkins, [0048], explicitly discloses recognizing and predicting “the same or similar transitions in the input signal based on the learned transitions”. The preprocessing disclosed in IBR-Hawkins, [0092], [0098], and [0114], disclose the particular step involved, i.e., preprocessing the data, such as multiplying by a scalar value, applying a function or transform to the data, etc., to change the range of data values, is, by definition, normalization. Normalization does not occur in a vacuum; it is used such that comparisons between different data is capable of being performed. Normalization, for example, often has its range between 0 and 1. Thus, it is clear that when Hawkins, [0048] describes the input signal being recognized and predicted “based on” the learned transitions, this means that they both share the same range of values, e.g., the input signal cannot have a range of -1 to 1 and the learned signal having a range of 0 to 100, otherwise it would be inoperable to carry out Hawkins’ intended invention. Indeed, the fact that the recognition and prediction of an input signal is “based on” learned transitions supports Hawkins implicitly disclosing the input signal is “normalized” to learned transitions (i.e., reference signal) such that the system properly operates in recognizing and predicting the same or similar transitions in the input signal based on the learned transitions, e.g., changing the range of data values which “normalizes” the input data in order to be able to perform the comparison to the learned transitions (i.e., reference signal).
The 103 rejection has been modified to incorporate this expanded explanation solely for the purpose of clarifying the rejection; however, the basis for the 103 rejection remains the same (i.e., no additional paragraphs were cited).
Applicant’s argument that Yu does not disclose “time-normalization of the input data in a time window onto the reference signal” (see Remarks, p. 14) is unpersuasive. Applicant argues that “Yu segments a signal and the segments are superimposed within the very same signal”. However, this argument completely ignores why superimposition occurred in the first place. Again, it is for comparison against other signals (hence, “identify[ing] multiple cyclic components at different time scales”, where the identified cyclic pattern may include any number of cyclic components). The reference signal was already disclosed by Hawkins; the 103 rejection has been modified solely for the purpose of clarifying this. Thus, Applicant would be attacking the references individually when the rejection is based on the combination of references.
Applicant’s argument that Nguyen does not disclose “combining the frequency segments of the input data associated with different time segments according to the time-normalization of the first sub-step” (see Remarks, p. 15-16) is unpersuasive. Nguyen’s disclosure overlaps with Yu in terms of taking into consideration the frequency data over some time period. Yu discloses the time-normalization aspect, as argued above. Therefore, Applicant is attacking the references individually when the rejection is based on the combination of references.
Applicant’s argument that the prior art do not disclose “determining an inadmissible deviation of a system behavior of a steering system or a braking system in a vehicle” (see Remarks, p. 16-17) is unpersuasive.
With respect to Hawkins, Applicant argues that “Hawkins”—and similarly IBR-Hawkins (see Remarks, p. 17)—“merely suggests applications to vehicle operation or automatic vehicle navigation…but clearly fails to teach detecting an inadmissible deviation of a system behavior of a steering system or a braking system in a vehicle” (see Remarks, p. 16-17). This is unpersuasive. Prior art still qualifies as (analogous) prior art if: (1) the reference is from the same field of endeavor as the claimed invention (even if it addresses a different problem); or (2) the reference is reasonably pertinent to the problem faced by the inventor (even if it is not in the same field of endeavor as the claimed invention).
In this case, Hawkins is reasonably pertinent to the problem faced by the inventor, i.e., disclosing most of the claimed steps. Hawkins even further discloses that the claimed steps can be applied to the same field of endeavor as the claimed invention, e.g., Hawkins disclosing possible applications to vehicle operation or automatic vehicle navigation. Note that the claimed “steering” and “braking” are more narrow forms of Hawkins’ “vehicle operations”, to which Hawkins’ disclosure could apply. Therefore, Hawkins, as the very least, qualifies as analogous prior art, with obvious suggestions to apply this more specifically to steering and braking, thereby making the combination with Nguyen (who discloses more specific vehicle operations) even more obvious.
Applicant further argues that “[Combining Hawkins] with Noda fails to arrive at determining an inadmissible deviation of a system behavior of a steering system or a braking system in a vehicle” (see Remarks, p. 17). However, Noda was not cited with respect to this claimed feature; therefore, Applicant is arguing against a reference not being used in the rejection for this specific feature.
Therefore, contrary to Applicant’s arguments, the combination of the prior art arrive at the claimed invention.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 4, 6-9, and 12-14 are rejected under 35 U.S.C. 103 as being unpatentable over Hawkins et al. (“Hawkins”) (US 2014/0067734 A1, incorporating by reference Hawkins et al. (“IBR-Hawkins”) (App. No. 13/218,170, published as US 2013/0054495 A1) at [0042]), in view of Noda et al. (“Noda”) (US 2018/0231969 A1), in further view of Yu et al. (“Yu”) (US 9,471,544 B1), in further view of Nguyen (“Nguyen”) (US 2019/0311289 A1).
Regarding claim 1: Hawkins teaches A method for determining an inadmissible deviation of a system behavior of a [system] from a standard value range using a neural network, the method comprising:
in a learning phase, (i) supplying the neural network with input data and output data of the [system], … and (iii) training, using the input data and the output data, the neural network to predict the system behavior of the [system] (Hawkins, [0048], where the sequence processor 314 learns and stores transitions between spatial patterns represented as sparse vector 342. See also Hawkins, [0066], where processing node 300 incorporates spatial patterns and temporal sequences (i.e., “input data of the technical device”) associated with the anomaly (i.e., “output data”) in its learning (i.e., “training”), such that predictions of whether an anomaly has occurred can be made when the same or similar spatial patterns and temporal sequences are later encountered.
See Hawkins, [0046-0057], with respect to the “neural network”, in particular the sequence processor may learn, store, and detect temporal sequences via the use of cells activated by select signals at certain time steps, with connections between cells, where using these learned transitions, the sequence processor 314 recognizes and predicts the same or similar transitions in the input signal by monitoring the activation states of its cells. Although Hawkins does not appear to explicitly utilize the phrase “neural network”, one of ordinary skill in the art would have recognized that Hawkins describes a neural network as disclosed in, e.g., the cited portions1,2);
in a prediction phase, which follows the learning phase, (i) supplying the neural network with the input data of the [system], (ii) normalizing the input data supplied to the neural network to the data of the reference signal, (iii) computing, in the neural network output comparison data based on the input data supplied in the prediction phase, the output comparison data corresponding to a predicted system behavior of the [system], and (iv) detecting the inadmissible deviation of the [system] in response to a difference between the output comparison data and the output data of the [system] lying outside the … range (Hawkins, [0048], where the sequence processor 314 recognizes and predicts the same or similar transitions in the input signal based on the learned transitions (i.e., “reference signal”). See also Hawkins, [0066], where the system learned patterns and sequences and produces predictions of whether an anomaly has occurred when the same or similar spatial patterns and temporal sequences are later encountered.
See Hawkins, [0052-0060], where cells send prediction output 404 as SP output 324 to anomaly detector 308. Anomaly detector 308 compares prediction output 404 (i.e., “output comparison data”) with subsequent sparse vector 342 (the actual value or state) (i.e., “the output data of the technical device”) to detect an anomaly.
See IBR-Hawkins, [0092], [0098], and [0114], where data may be preprocessed, such as converting integer values to floating point values, multiplying by a scalar value, applying a function or transform to the data (e.g., a linear, logarithmic, or dampening function, or a Fourier transform) to change the range of data values (i.e., “(ii) normalizing the input data supplied to the monitoring algorithm to the data of the reference signal”, i.e., implicitly to the “learned transitions” in order to perform the step described by Hawkins, [0048] in which same or similar transitions in the input signal “based on the learned transitions” are recognized and predicted)); and
in response to detecting the inadmissible deviation of the [system], at least one of outputting a warning signal or deactivating a function of the [system] … (Hawkins, [0068-0069], where state information associated with anomalies are flagged after further analysis, and issues anomaly signal 352 when sequence processor 314 is placed in a state associated with the flagged anomalies. User interface device 344 alerts the user of the flagged anomaly after receiving anomaly signal 352 from the state monitor 521).
Although Hawkins does not appear to explicitly state that the anomaly signal is a “warning” signal as claimed, the claimed invention does not distinguish over the prior art because the differences in the claim limitations and the prior art’s disclosure are only found in the nonfunctional descriptive material and are not functionally involved in the steps recited. The claimed steps would have been performed the same regardless of the specific type of data involved (i.e., a warning signal as claimed, an anomaly signal as disclosed in the prior art, or some other type of signal alerting that there is some sort of deviation in the data). Thus, this descriptive material will not distinguish the claimed invention from the prior art in terms of patentability. See In re Gulack, 703 F.2d 1381, 1385, 217 USPQ2d 401, 404 (Fed. Cir. 1983); In re Lowry, 32 F.3d 1579, 32 USPQ2d 1031 (Fed. Cir. 1994).
Therefore, it would have been obvious to a person of ordinary skill in the art to have referred to Hawkins’ teachings in making the claimed invention, because such data does not functionally relate to the steps in the method claimed and because the subjective interpretation of the data does not patentably distinguish the claimed invention over the prior art.
Hawkins does not appear to explicitly teach that the system pertaining to a steering system or a braking system in a vehicle; [wherein the learning phase includes] (ii) normalizing the input data supplied to the neural network to data of a reference signal, the reference signal defining a reference system behavior of the steering system or the braking system corresponding to a defined driving maneuver of the vehicle; [wherein the range is] a standard value range; [and] wherein the normalizing the input data in both of the learning phase and the prediction phase includes (i) in a first sub-step, performing time-normalization of the input data in a time window onto the reference signal, (ii) in a second sub-step, determining frequency segments of the input data by transforming the input data for time segments of the time window into a frequency domain, and (iii) in a third sub-step, combining the frequency segments of the input data associated with different time segments according to the time-normalization of the first sub-step.
Noda teaches [wherein the learning phase includes] (ii) normalizing the input data supplied to the neural network to data of a reference signal, the reference signal defining a reference system behavior of the [system] (Noda, [0096-0097], where the learning means 141 normalizes the feature points extracted in a previous step to convert into a feature vector, and clusters each feature vector to learn clusters. The learning means then stores the cluster center and the cluster radius r of each cluster in the learning result storage unit, completing a series of learning processing (and later, e.g., in Noda, [0102], when the diagnosis means will normalize the diagnosis target data to convert into a feature vector, and calculates an abnormality measure u based on a cluster)); [and]
[wherein the range is] a standard value range (Noda, [0083], where when the abnormality measure u>1, the diagnosis target data is present outside the cluster (outside the normal range (i.e., “standard value range”), and thus the diagnosis unit diagnoses the mechanical facility 2 as “abnormality predictor is present”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins and Noda (hereinafter “Hawkins as modified”) with the motivation of normalizing the learning data in order to make the learned data consistent with the predicted data (which Hawkins discloses normalizing), thereby obtaining consistent (and therefore more accurate) predictions, and (2) comparing the values against a standard value range with the motivation of enabling greater flexibility in the values that would be considered acceptable.
Hawkins as modified does not appear to explicitly teach the system pertaining to a steering system or a braking system in a vehicle; that the monitoring algorithm is a neural network; the reference system behavior pertaining to the steering system or the braking system corresponding to a defined driving maneuver of the vehicle; [and] wherein the normalizing the input data in both of the learning phase and the prediction phase includes (i) in a first sub-step, performing time-normalization of the input data in a time window onto the reference signal, (ii) in a second sub-step, determining frequency segments of the input data by transforming the input data for time segments of the time window into a frequency domain, and (iii) in a third sub-step, combining the frequency segments of the input data associated with different time segments according to the time-normalization of the first sub-step.
Yu teaches in a first sub-step, performing time-normalization of the input data in a time window onto the reference signal (Yu, [6:34-45], where the system identifies a period of interest and segments the signal based on the identified period. The resulting segments may then be superimposed, thus building a point-by-point model of the cyclic pattern. Furthermore, the processor 106 may identify multiple cyclic components are different time scales by repeating the above analysis using different values for the period, and the identified cyclic pattern may include any number of cyclic components. See Hawkins above with respect to the “reference signal”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins as modified and Yu (hereinafter “Hawkins as modified”) with the motivation of enabling comparisons to be performed even at different time scales.
Hawkins as modified does not appear to explicitly teach the system pertaining to a steering system or a braking system in a vehicle; that the monitoring algorithm is a neural network; the reference system behavior pertaining to the steering system or the braking system corresponding to a defined driving maneuver of the vehicle; [and] wherein the normalizing the input data in both of the learning phase and the prediction phase includes (ii) in a second sub-step, determining frequency segments of the input data by transforming the input data for time segments of the time window into a frequency domain, and (iii) in a third sub-step, combining the frequency segments of the input data associated with different time segments according to the time-normalization of the first sub-step.
Nguyen teaches the system pertaining to a steering system or a braking system in a vehicle; the reference system behavior pertaining to the steering system or the braking system corresponding to a defined driving maneuver of the vehicle (Nguyen, [0017], where the system compares the telematics data with multiple instances of known driving behavior information (i.e., “reference system behavior”) to recognize safe and unsafe driving behavior, etc. (i.e., “defined driving maneuver of the vehicle”), which may include hard acceleration, braking, or cornering (i.e., “a steering system or a braking system in a vehicle”)); [and]
wherein the normalizing the input data in both of the learning phase and the prediction phase includes … (ii) in a second sub-step, determining frequency segments of the input data by transforming the input data for time segments of the time window into a frequency domain (Nguyen, [0072], where the time domain signal is partitioned into overlapping short frames, and the Fourier transform is applied independently on each frame, e.g., using short time Fourier transform. See Yu, [11:37-49], where applying a Fourier transform, such as a Fast Fourier Transform (FFT) transforms the signal component in the time domain to a representation in a frequency domain), and
(iii) in a third sub-step, combining the frequency segments of the input data associated with different time segments according to the time-normalization of the first sub-step (Nguyen, [0073], where on each frame (i.e., “frequency segment”), the disclosed technology computes spectral energy, spectral centroid and spectral variance, and aggregates over different frames using statistical extraction. See Yu above with respect to the “time-normalization of the first sub-step”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins as modified and Nguyen (hereinafter “Hawkins as modified”) with the motivation of enabling a telemetry-based insurance model, thereby enabling an insurance company, for example, to tailor their insurance plan for the driver (see, e.g., Nguyen, [0003] and [0017]), and attempting to ensure that the sampling rate of sensors are high enough to capture various ranges of information pertaining to different classifications, e.g., a vehicle moving or being idle (Nguyen, [0072]), and such that the behavior of sensors can be described at different time scales.3
Regarding claim 4: Hawkins as modified teaches The method as claimed in claim 1, wherein the input data supplied to the neural network are in time-discrete form (Noda, [0068], [0114], and [0133-0134], where the sensor data acquired by the diagnosis target data acquisition unit includes detection values of a sensor and an elapsed time from the start time t11, e.g., there are elapsed times from the start time of the operation process being indicative of a “time-discrete form” as claimed. See Hawkins in claim 1 above with respect to the “neural network”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins as modified and Noda with the motivation of enabling the system to quickly identify anomalies at any point in time (as opposed to, e.g., a time window), and thus quickly raise any necessary alarms/notifications faster.
Regarding claim 6: Hawkins as modified teaches The method as claimed in claim 1, wherein the transforming the input data for the time window into the frequency domain, which is performed in the second sub-step, is carried out using a short-time Fourier transform (Nguyen, [0072], where the Fourier transform is applied independently on each frame, e.g., using short time Fourier transform).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins as modified and Nguyen with the motivation of capturing rapidly changing data signals (see, e.g., Nguyen, [0072], where a vehicle may be accelerating/braking).
Regarding claim 7: Hawkins as modified teaches The method as claimed in claim 1, wherein the output data of the steering system or the braking system is transformed into the frequency domain and compared in the frequency domain with the output comparison data computed in the neural network (Hawkins, [0060], where anomaly detector compares prediction output 404 with subsequent sparse vector 342 (the actual value or state) to detect an anomaly. See Hawkins in claim 1 above with respect to the “neural network”. See Yu, [11:37-49], where the system computes a fast Fourier transform (FFT) of the ith signal component, where the FFT transforms the signal component in the time domain to a representation in a frequency domain. See Nguyen in claim 1 above with respect to the “steering system or the braking system”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins as modified and Yu with the motivation of easily identifying and isolating certain frequency components of interest.4
Regarding claim 8: Hawkins as modified teaches The method as claimed in claim 1, wherein the output comparison data computed in the neural network is transformed into a time domain and compared in the time domain with the output data of the steering system or the braking system (Hawkins, [0060], where anomaly detector compares prediction output 404 with subsequent sparse vector 342 (the actual value or state) to detect an anomaly. See Hawkins in claim 1 above with respect to the “neural network”. See Yu, [11:23-36], where the signal is decomposed (i.e., “transformed”)5 into multiple signal components, e.g., by breaking the signal down into signal components in the time domain (i.e., “transformed into a time domain”). See Nguyen in claim 1 above with respect to the “steering system or the braking system”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins as modified and Yu with the motivation of preserving instantaneous frequency changes in the signal and phase information (Yu, [11:23-36]), as well as enabling potential integration with (typical) systems that identify faults based on an analysis of data in the time domain6, i.e., greater convenience for integrating with other tools.
Regarding claim 9: Hawkins as modified teaches The method as claimed in claim 1, wherein the reference signal is formed from a plurality of preceding values of the input data (Yu, [Claim 4], where the historical probability distribution is generated based on previously received samples; a likelihood is computed for each sample point in the signal based on at least in part on the historical probability distribution; selecting a likelihood threshold; and comparing that likelihood to the likelihood threshold. As a result, because the historical probability distribution is compared to the likelihood threshold (i.e., “reference signal”), this indicates that the likelihood threshold is “formed from a plurality of preceding items of the input data”, as claimed).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins as modified and Yu with the motivation of determining past trends to predict an anomaly, as past trends are good references for learning from.
Regarding claim 12: Hawkins as modified teaches The method as claimed in claim 1, wherein a control unit in a vehicle is configured to perform the method (Hawkins, [Claim 11], where the disclosed system may be implemented as a non-transitory computer-readable storage medium storing instructions, the instructions when executed by a processor cause the processor to implement the disclosed steps).
Although Hawkins as modified does not appear to explicitly state that the processor pertains to “a control unit in a vehicle” for performing the method, Hawkins’ disclosed processor is analogous to the claimed “control unit in a vehicle” as it is reasonably pertinent to the problem faced by the inventor. Therefore, one of ordinary skill in the art would have found it obvious to have modified Hawkins to incorporate additional types of control units (other than a processor within a system, as disclosed by Hawkins) with the motivation of broadening the types of applications of anomaly signal detection, including within the realm of vehicles.
Regarding claim 13: Hawkins as modified teaches The method as claimed in claim 1, wherein a computer program product includes program code configured to carry out the method (Hawkins, [Claim 11], where the disclosed system may be implemented as a non-transitory computer-readable storage medium storing instructions, the instructions when executed by a processor cause the processor to implement the disclosed steps).
Regarding claim 14: Hawkins as modified teaches The method as claimed in claim 13, wherein a non-transitory machine-readable storage medium is configured to store the computer program product (Hawkins, [Claim 11], where the disclosed system may be implemented as a non-transitory computer-readable storage medium storing instructions, the instructions when executed by a processor cause the processor to implement the disclosed steps).
Claim 2 is rejected under 35 U.S.C. 103 as being unpatentable over Hawkins et al. (“Hawkins”) (US 2014/0067734 A1, incorporating by reference Hawkins et al. (“IBR-Hawkins”) (App. No. 13/218,170, published as US 2013/0054495 A1) at [0042]), in view of Noda et al. (“Noda”) (US 2018/0231969 A1), in further view of Yu et al. (“Yu”) (US 9,471,544 B1), in further view of Nguyen (“Nguyen”) (US 2019/0311289 A1), in further view of Mezic et al. (“Mezic”) (US 2016/0203036 A1).
Regarding claim 2: Hawkins as modified teaches The method as claimed in claim 1, but does not appear to explicitly teach wherein the normalizing the input data in both of the learning phase and the prediction phase comprises: harmonizing a number of values of the input data supplied to the neural network with a number of values of the data of the reference signal.
Mezic teaches harmonizing a number of values of the input data supplied to the neural network with a number of values of the data of the reference signal (Mezic, [0039-0042], where the system maps information to a standard format, where the mapping of the provided information into the standard format or language allows the feature detector to determine which indicator functions are to be applied to any given time-series dataset. Note that the “information” comprises different features, i.e., “values”, which are, e.g., mapped into various columns that break up the provided phrase into discrete pieces of information using standard language. The information generated from the feature detector 141 can be stored for later retrieval by the machine learning feedback system 144, which uses heuristics such as artificial neural network to perform the disclosed fault detection (Mezic, [0008] and [0055]) (i.e., “supplied to the neural network”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins as modified and Mezic with the motivation of determining which indicator functions are to be (specifically) applied to any given time-series dataset for detecting anomalies in a particular type of subsystem (Mezic, [0042]), thus resulting in improved accuracies in detection (since, e.g., components/sensors may be different and have variability; thus, having different indicator functions allows for improved/more accurate detections of anomalies depending on the specific type of component/sensor being monitored).
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over Hawkins et al. (“Hawkins”) (US 2014/0067734 A1, incorporating by reference Hawkins et al. (“IBR-Hawkins”) (App. No. 13/218,170, published as US 2013/0054495 A1) at [0042]), in view of Noda et al. (“Noda”) (US 2018/0231969 A1), in further view of Yu et al. (“Yu”) (US 9,471,544 B1), in further view of Nguyen (“Nguyen”) (US 2019/0311289 A1), in further view of Saini et al. (“Saini”) (US 2018/0225320 A1).
Regarding claim 3: Hawkins as modified teaches The method as claimed in claim 1, but does not appear to explicitly teach wherein, when a number of values of the input data and a number of values of the data of the reference signal are equal, but the input data is skewed with respect to the data of the reference signal, the normalizing the input data in both of the learning phase and the prediction phase comprises: mapping the input data onto the data of the reference signal.
Saini teaches wherein, when a number of values of the input data and a number of values of the data of the reference signal are equal, but the input data is skewed with respect to the data of the reference signal, the normalizing the input data in both of the learning phase and the prediction phase comprises: mapping the input data onto the data of the reference signal (Saini, [0060], where if the data of the interest is non-normal, a transformation may be applied to the data set of interest to normalize the data set, e.g., by applying a Box-Cox transformation which provides a data set having a normal or approximately normal distribution (i.e., “the data of the reference signal”). See Yu, [11:23-36], where each signal component has the same length as the signal, and the superposition of all the signal components results in the signal (i.e., “a number of items of the input data and a number of items of the reference signal are equal”).
See Hawkins as modified above with respect to the “normalizing the input data in both the learning phase and the prediction phase”, e.g., more specifically, IBR-Hawkins, [0092], [0098], and [0114], and Noda, [0096-0097] and [0102]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins as modified and Saini with the motivation of accurately/correctly identifying data points (as skewness can result in incorrect identification of anomalous/non-anomalous data points) (Saini, [0008] and [0060]).
Furthermore, although Hawkins as modified and Saini do not appear to explicitly state that the transformation is performed “when” the number of items of the input data and the reference signal are equal, one of ordinary skill in the art would have found it obvious to have performed this transformation of the skew only under such circumstances with the motivation of ensuring that the data is comparable (i.e., thus enabling a clean transformation).
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Hawkins et al. (“Hawkins”) (US 2014/0067734 A1, incorporating by reference Hawkins et al. (“IBR-Hawkins”) (App. No. 13/218,170, published as US 2013/0054495 A1) at [0042]), in view of Noda et al. (“Noda”) (US 2018/0231969 A1), in further view of Yu et al. (“Yu”) (US 9,471,544 B1), in further view of Nguyen (“Nguyen”) (US 2019/0311289 A1), in further view of Peng et al. (“Peng”) (US 2008/0201397 A1).
Regarding claim 5: Hawkins as modified teaches The method as claimed in claim 1, but does not appear to explicitly teach wherein the time-normalization of the input data onto the reference signal, which is performed in the first sub-step, is carried out using dynamic time warping.
Peng teaches wherein the time-normalization of the input data onto the reference signal, which is performed in the first sub-step, is carried out using dynamic time warping (Peng, [0005] and [0019], where the disclosed system applies dynamic time warping (DTW) to the time-series data).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Hawkins as modified and Peng with the motivation of achieving the optimal alignment of highly correlated data and the approximate time shifts between them (Peng, [0033]) and efficiently minimizing the effects of shifting and distortion in time by allowing “elastic” transformation of time series in order to detect similar shapes with different phases.7
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to IRENE BAKER whose telephone number is (408)918-7601. The examiner can normally be reached M-F 8-5PM PT.
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, Boris Gorney can be reached at (571) 270-5626. 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.
/IRENE BAKER/Primary Examiner, Art Unit 2154
27 August 2026
1 See, e.g., Thaler. US 5,852,815 A, which pertains to an artificial neural network with references to activation cells relating to, e.g., various layers of the neural network.
2 Ross et al. US 2017/0103316 A1, which, like Thaler above, pertains to a neural network with references to cells making up those neural networks.
3 Mezic et al. US 2016/0203036 A1 at [0045] (“Once the new time-series are generated, the spectral analyzer 142 can perform a spectral analysis (e.g., …a discrete Fourier transform…) of each of the new time-series to generate a spectral response for each of the new time-series. Performance of the spectral analysis may result in the conversion of the data from the time domain to the frequency domain such that the behavior of the sensors 115 (e.g., whether the data points at different time instances result in a true or false condition) can be described at different time-scales…”).
4 Jardine et al. “A review on machinery diagnostics and prognostics implementing condition-based maintenance”. Mechanical Systems and Signal Processing 20 (2006) 1483-1510. Published 2005. URL Link: < https://www.sciencedirect.com/topics/engineering/frequency-domain-analysis>. Accessed Jun 2025. [3.1.2. Frequency-domain analysis on p. 1487] (“The advantage of frequency-domain analysis over time-domain analysis is its ability to easily identify and isolate certain frequency components of interest”).
5 See, e.g., Rajagopal et al. US 2002/0150298 A1 at [0017] (“…the unified signal transform may be operable to decompose the signal into generalized basis functions…”), thus demonstrating that Yu’s disclosure of a “decomposition” is a type of transform.
6 Mezic et al. at [0062].
7 Senin. “Dynamic Time Warping Algorithm Review”. Published 2008. URL Link: <https://csdl.ics.hawaii.edu/techreports/2008/08-04/08-04.pdf>. See page 2 under section “DTW Algorithm”.
Prosecution Insights